MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Hemisphere for Ornament and Crime
MUFF WIGGLER Forum Index -> Eurorack Modules  
Author Hemisphere for Ornament and Crime
chysn
Hemisphere is alternate alternate firmware for Ornament and Crime. It's a dual applet framework that focuses on modular building blocks with visual interfaces.

If you'd care to participate in a public beta release, you can download a hex file at http://www.beigemaze.com/hemisphere.html

The beta codebase keeps most of the original O_C apps intact. The exception is Meta-Q, which had to be disabled to make room for the new software.

The beta release includes 14 applets including: dual quantizer, dual clock divider/multiplier, 5-step sequencer, dual binary trigger sequencer, sample and hold with random, dual calculator (arithmetical operations), dual logic (with CV control), threshold logic neuron, gated/sequential switch, probability brancher, skewed LFO, dual ADSR, 8-bit audio looper, and gated VCA.

More info, including links to videos, is here:

http://butmostlycrime.blogspot.com/2018/06/introducing-hemisphere.html
http://butmostlycrime.blogspot.com/2018/07/hemisphere-beta-release-pfa q.html

If there's any interest, support can be sought at this forum post.

Note that this project is an open-source initiative and is unaffiliated with the Ornament and Crime development team.
Zymos
woah
FarmerPhilip
I saw a YouTube video introducing this a few days ago: very interesting, especially how some of the modes can share inputs. Looking forward to this a lot!
Godphaser
Wow!

Quote:
threshold logic neuron


Could you explain what's this please? love
rayultine
seriously chysn, how did you come up with all of this from scratch in three weeks? By 2019, our ornament and crimes are going to be cooking us breakfast and 3D printing its babies.

edit:
sans meta-Q, is there enough space left to also squeeze in Darkest Timeline?

edit:
another crazy idea: in Lo-Fi Tape, if length of loop could be CV controlled, then you are basically replicating the 2hp Freez module with 8-bit audio resolution. And since the o_c inputs are DC coupled, you could also loop CV, right? holy shit.
ipassenger
Awesome stuff, probably no time to help beta but great work! SlayerBadger!
roya
Been using it in the last few days, great great stuff!!
chysn
Godphaser wrote:

Quote:
threshold logic neuron

Could you explain what's this please? love


It's a three-input/single-output logic gate. Each of three inputs ("dendrites") is given a weight (in the case of Hemisphere, a weight of between -9 and 9). The output ("axon") is given a threshold (in Hemisphere, between -27 and 27). When the sum of the weights of high inputs exceeds the threshhold, the output goes high.

The neuron can be used to reproduce common logic gates; for example:

Input 1: w=3
Input 2: w=3
Input 3: w=3
Output: t=8

This reproduces and AND gate, because all inputs need to go high (for a sum of 9) to exceed the threshold of 8.

Input 1: w=3
Input 2: w=3
Input 3: w=3
Output: t=2

This reproduces an OR gate, because only one input needs to go high to exceed the threshold of 2.

The values can also be negative. So

Input 1: w=-3
Input 2: w=-3
Output: t=-5

This is a two-input NAND gate, because the output is high (because 0>-5) unless both inputs 1 and 2 are high, which brings the sum below the threshold.

But you're not limited to reproducing standard Boolean operations. You can create your own logical operations. For example, you can do this:

Input 1: w=2
Input 2: w=2
Input 3: w=5
Output: t=3, and mult the output into Input 3

This creates a type of state memory: when Inputs 1 and 2 go high, the output goes high (2+2>3), and the output is sent back to Input 3, which forces the output to remain high, whatever happens to Inputs 1 and 2 later.

A variation of that theme is

Input 1: w=2
Input 2: w=-2
Input 3: w=2
Output: t=1, and mult the output into Input 3

Now, Input 1 triggers the axon, which feeds back into itself. Input 1 can go low again and the output stays high. But a high signal at Input 2 will reset the memory unless Input 1 is still on.

My newest O_C project is an artificial neural network full app with four* of these things linked together (or functioning separately); so I'll have a more complete treatment at some point.

Oh, yeah... what can this actually be used for in a patch? Complex but deterministic clocking, or patching specific responses to controller events, or generative patches. If you like regular Boolean logic, you'll probably have a blast with threshold logic.

* Turned out to be six in the end
chysn
rayultine wrote:
seriously chysn, how did you come up with all of this from scratch in three weeks?


I'd have to credit the straightforward nature of the underlying framework, and a good set of tools within the framework. Patrick and Tim were both generous with their time, so it was an easy system to learn.

Quote:
sans meta-Q, is there enough space left to also squeeze in Darkest Timeline?


The Hemisphere beta takes up Meta-Q's space and a little more. The Teensy Loader says that it's at 100%. Obviously it's rounding up, but...

Eventually, I plan to distribute Hemisphere (with many more applets), Darkest Timeline, the in-progress Neural Network, and some other stuff (including a handful of current O_C apps) as Hemisphere Suite. Right now, I'm interested in finding out if I'm on the right track in terms of a friendly UI and stable, useful, applets.

Quote:
another crazy idea: in Lo-Fi Tape, if length of loop could be CV controlled, then you are basically replicating the 2hp Freez module with 8-bit audio resolution.


Sure. With two CV inputs, and one of them locked into audio in, I was having way more fun modulating sound-on-sound, so I connected length to the knob. It was sort of a judgment call. With such a short amount of time, modulating time parameters isn't as enjoyable as it is on, say, Phonogene*.

Quote:
And since the o_c inputs are DC coupled, you could also loop CV, right? holy shit.


Huh, yeah, I guess so. I was so caught up in the audio experiment that I never considered playing with CV. But it is basically a really coarse 2048-step sequencer without clock control. Good catch!

* That's right, I still have a Phonogene. Shut up.
chysn
roya wrote:
Been using it in the last few days, great great stuff!!


Thanks for trying it!
uebl
That's.. awesome! Now I only need to figure out whether this can be compiled to flip the screen, controls and inputs/outputs so it can be used upside-down in a skiff..
chysn
uebl wrote:
That's.. awesome! Now I only need to figure out whether this can be compiled to flip the screen, controls and inputs/outputs so it can be used upside-down in a skiff..


Hemisphere itself works as a normal Ornament and Crime app, so the FLIP_180 compile option should work fine. If it doesn’t, let me know!
tnktni
Wow, intrigued. Excited to try this out.
shellfritsch
hey chysn, thanks for your hard work on this. the limitations of the applets make the interface waaay more approachable. plus i've never really needed a quad quantizer. dual is fine for me smile

i just tried out the lofi looper and i'm getting a loud click on the end of the loop - even when starting with an empty buffer. i can tell it's at the end because when i turn the encoder to change the loop length it goes away. also is it possible to have monitoring of the input? could "cv b" act as a pitch control? or maybe start point shift?

with the clock div/mult - based on the help page it seems that the cv inputs are unused? can they be used to control div/mult amount?

finally - is there a way to disable the screensaver, or at least make the time before activating waaaay longer. i find it very distracting to play with a setting, look away at something else, then look up to see it on something else. or i'll just be staring at the help page, trying to commit the controls to memory, and the screensaver will pop up while i'm reading.

again - really appreciate your efforts!
mdoudoroff
applause

So happy to see new stuff for this amazing little platform.

I wonder if it’s technically possible to have a web site where you check off the apps you want and then download a firmware package containing that combo?
chysn
Shellfritsch, thanks for the input!

shellfritsch wrote:
i just tried out the lofi looper and i'm getting a loud click on the end of the loop - even when starting with an empty buffer. i can tell it's at the end because when i turn the encoder to change the loop length it goes away.


I'm working on this issue. I've got a mistake handling the buffer somewhere. For now, as you mentioned, the workaround is to change the end point a little bit.

Quote:
also is it possible to have monitoring of the input?


I kind of like the current display, but I think I could add an output level overlay without the screen getting too busy.

Quote:
could "cv b" act as a pitch control?


No, unfortunately not. We're dealing with such low sample rates here that dividing by anything drives the sample rate so low that the sound goes away, and multiplying by anything immediately results in a chirpy mess. Pitch-wise, I'm really threading a needle here.

Quote:
or maybe start point shift?


I'm pretty convinced that sound-on-sound level is the right function of CV B. Similar to the pitch issue, the buffer is so small that there's not enough "wiggle" room to make time modulation very interesting.

Quote:
with the clock div/mult - based on the help page it seems that the cv inputs are unused? can they be used to control div/mult amount?


Yeah, I think that's a good idea.

Quote:
finally - is there a way to disable the screensaver, or at least make the time before activating waaaay longer. i find it very distracting to play with a setting, look away at something else, then look up to see it on something else. or i'll just be staring at the help page, trying to commit the controls to memory, and the screensaver will pop up while i'm reading.


We talked about the screensaver a bit in the DIY topic. The consensus seems to be that the OLED isn't in any immediate danger. The screensaver is part of the O_C's framework, so I treated it like this:

(1) The only thing that's removed during the screensaver for applets is the header and blinking cursors. I think the header is where burn-in is most likely to happen. Also, there needs to be some visual indicator that a screensaver is in use, because the first button push while in screensaver mode is never delegated to the app; it just turns off the screensaver.

(2) But some people, like me, leave their synths on all day. So Hemisphere blanks the screen after 30 minutes of inactivity.

But, I forgot about the help screens. For beta 2, I'll make it so that help screens stay active in screensaver mode.

Note that you can change screensaver time to up to 120 seconds in the O_C's calibration menu.

Thanks again!
shellfritsch
ah i didn't know you can set the screensaver timeout already. i'm totally new to O_C

regarding input monitoring - i was referring to hearing the input, not seeing. not sure if this is possible. if not i can always patch my own monitoring.

regarding cv b on the looper - i didn't realize that was already set to SOS. makes sense.
chysn
mdoudoroff wrote:

I wonder if it’s technically possible to have a web site where you check off the apps you want and then download a firmware package containing that combo?


Technically, it’s trivial. But as an actual project, it’s kind of that perfect storm of “time consuming” and “uninteresting” that makes it unlikely to get my interest. But anyone who wants to take it on has my full support. smile
chysn
shellfritsch wrote:
regarding input monitoring - i was referring to hearing the input, not seeing. not sure if this is possible. if not i can always patch my own monitoring.

regarding cv b on the looper - i didn't realize that was already set to SOS. makes sense.


Oh! Yes, you can monitor by patching CV to input B, and you can balance between hearing the input and the loop. Note that you're hearing the input through the same 8-bit/2kHz system that you're recording through; there's not really a way to pass a signal through unscathed.

The end-of-cycle click issue turned out to be a buffer over-read. It's corrected for the next beta release, which will be on Sunday, July 8th.
Godphaser
Quote:
It's a three-input/single-output logic gate. Each of three inputs ("dendrites") is given a weight (in the case of Hemisphere, a weight of between -9 and 9). The output ("axon") is given a threshold (in Hemisphere, between -27 and 27). When the sum of the weights of high inputs exceeds the threshhold, the output goes high.


Super nice! Thanks a lot for your answer.

It's kinda like blackjack but for CV.
chysn
Godphaser wrote:
It's kinda like blackjack but for CV.


Ha! Pretty much.
Eurtrude
First, thanks a lot for this release, it's like a dream that comes true for me smile
Some superb ideas in there too !

I just installed the thing but barely had time to play with the apps.
Will there be some possibilities for routing the CV inputs ? Like control the segments of enveloppes, or lfo rate, etc. ?

About ADSR enveloppes, is there a bug ? the release is full or middle and there is almost no control over it. Besides, the resolution of steps for the other segments is a bit too low (I mean the changes are too big from one step to the next, not very precise).

I'll test more the others apps when I'll have some time, is there a place to write about it other than here ?

Thanks again for your work, it's freaking awesome smile
chysn
Eurtrude wrote:
Will there be some possibilities for routing the CV inputs ? Like control the segments of enveloppes, or lfo rate, etc. ?


Yeah, CV control over attack and release in the ADSR and over rate and skew for the LFO are already in Sunday's release. And I can increase the resolution, too. There are only 100 steps in each segment.

Quote:
I'll test more the others apps when I'll have some time, is there a place to write about it other than here ?


I'd say, this is probably the best place for now. You can always feel free to PM me, too.
BaloErets
Wow, this is really the bomb SlayerBadger! With 2 O_c's in my rack, I gave Hemisphere a good play today. Everything just works. Ton's of potential having access 2 4 of these "mini modules". The Calculations alone is just so useful!

Thank you so much for sharing with us we're not worthy
Eurtrude
chysn wrote:
Eurtrude wrote:
Will there be some possibilities for routing the CV inputs ? Like control the segments of enveloppes, or lfo rate, etc. ?


Yeah, CV control over attack and release in the ADSR and over rate and skew for the LFO are already in Sunday's release. And I can increase the resolution, too. There are only 100 steps in each segment.

You mean it's already there ? I haven't tried, just saw in the menu that CV indicates Nothing in those apps.

It would be really great to increase the resolution (well, at least to my point of view), I always liked the 258 steps from the Piqued app.

Anyway, this is amazing and honnestly a game changer for the O_c we're not worthy

Quote:
I'll test more the others apps when I'll have some time, is there a place to write about it other than here ?


I'd say, this is probably the best place for now. You can always feel free to PM me, too.


Ok smile
Mikro93
I'm still fairly new to o_C, but then this happened:

chysn wrote:

My newest O_C project is an artificial neural network full app with four of these things linked together (or functioning separately); so I'll have a more complete treatment at some point.


Congrats for the Hemisphere as is already, and I'm so looking forward to see what's next smile

Cheers!
fever606
chysn wrote:
The beta codebase keeps most of the original O_C apps intact.

Potentially dumb question, but does Hemisphere just present itself as another O_C app, meaning that I can bail out of it and load up CopierMaschine for all of my LFSR needs? (At least until I can get my hands dirty enough to figure out how to roll my own DU-RDT replacement as an applet for Hemisphere seriously, i just don't get it)
Eurtrude
fever606 wrote:
chysn wrote:
The beta codebase keeps most of the original O_C apps intact.

Potentially dumb question, but does Hemisphere just present itself as another O_C app, meaning that I can bail out of it and load up CopierMaschine for all of my LFSR needs? (At least until I can get my hands dirty enough to figure out how to roll my own DU-RDT replacement as an applet for Hemisphere seriously, i just don't get it)


Yep, exactly ! It only replaces Meta-Q
fever606
Eurtrude wrote:
Yep, exactly ! It only replaces Meta-Q

Wooo! This is fun! Thanks!
tnktni
Eurtrude wrote:
fever606 wrote:
chysn wrote:
The beta codebase keeps most of the original O_C apps intact.

Potentially dumb question, but does Hemisphere just present itself as another O_C app, meaning that I can bail out of it and load up CopierMaschine for all of my LFSR needs? (At least until I can get my hands dirty enough to figure out how to roll my own DU-RDT replacement as an applet for Hemisphere seriously, i just don't get it)


Yep, exactly ! It only replaces Meta-Q


Oh wow OK, now I definitely have to try it out.
chysn
I've added instructional documentation for Hemisphere in general (as opposed to the applets) at GitHub:

https://github.com/Chysn/O_C-Hemisphere/wiki/Hemisphere-General-Instru ctions

There are some refinements for the upcoming (July 8) Beta 2, and these are noted in the text.
antofthesky
Are the help screen contents for each applet online anywhere, instead of just on the module? Would be nice to read.
prphnc
Dreams become reality!
Thank you chysn!!!
It's peanut butter jelly time!
chysn
antofthesky wrote:
Are the help screen contents for each applet online anywhere, instead of just on the module? Would be nice to read.


Not yet, but that's coming. The closest thing to applet documentation at the moment is the YouTube playlist at

https://www.youtube.com/playlist?list=PLC5d4vp670NsuEUQybeCPYOvoDJb9fn Bi
chysn
Hemisphere Beta 2 is now available at http://www.beigemaze.com/hemisphere.html

Release notes:

* Selection now applies only to the selected hemisphere; the opposite hemisphere functions as normal
* LoFiTape: Fixed buffer over-read, which caused clicking for full-length loops
* Screensaver won't interrupt help screens
* ADSR: Added bipolar CV control of attack and release
* ADSR: Increased value resolution for all segments
* ADSR: Sustain is now silent at lowest level
* SkewLFO: Added bipolar CV control of rate and skew
* SkewLFO: Increased value resolution for rate and skew
* Added Turing Machine based on original Tom Whitwell kit (16-bit digital shift register)
* Removed Gated VCA to make room for the above
* Removed unnecessary screensaver animations to save space
* Help screen won't be duplicated if both hemispheres are assigned the same applet

The video for the Turing Machine is here https://www.youtube.com/watch?v=RNWCq8mW5wE
SavageMessiah
Nice, I think most of the O_C apps are pretty unpleasant to use, UI-wise, but these seem simple enough to get around that, and two at once too! Once this is out of beta I'll definitely be giving it a try.
BaloErets
Great stuff chysn! The Turing Machine addition is fantastic! Haven't got the chance to play with it yet, but looking forwards to it.

Quick question; I didn't hear mention what the CV inputs do. Are we to assume that are CV for Scale and Probability like the MTM version, or are the currently not programmed to do anything?
chysn
BaloErets wrote:
Great stuff chysn! The Turing Machine addition is fantastic! Haven't got the chance to play with it yet, but looking forwards to it.

Quick question; I didn't hear mention what the CV inputs do. Are we to assume that are CV for Scale and Probability like the MTM version, or are the currently not programmed to do anything?


The CV inputs don't do anything. I did the Turing Machine mostly to gain some insight into how a Turing Machine works, so this is probably the purest-possible implementation.

There's sort of a bad news/good news situation. The bad news is that I'm officially out of memory for new features. There are about 500 bytes left on Ornament and Crime when Beta 2 is installed. So adding features to anything is largely out of the question. I'm not sure what I'm going to do when the O_C 1.3.5 update comes out. I'll merge it in somehow, but it'll be an interesting weekend.

The good news is that I'm working on Hemisphere Suite now, which is a version of Hemisphere that isn't going to be bound by keeping O_C apps around. Here I have enough space to go crazy:

https://www.youtube.com/watch?v=myTENbq0_g4

and I have some fun Turing Machine type ideas.
AnalogAssailant
Will you be adding CV control to clock div/multiplier applet?
chysn
AnalogAssailant wrote:
Will you be adding CV control to clock div/multiplier applet?


I'm trying to find the space for that. My initial coding of it took me about a hundred bytes over the limit, so I'm working on shaving things.
pines
Thanks Chysn! I loaded up beta 2 on my second O&C, and it really expands the functionality. The logic operations are especially welcome. Looking forward to the “Official” release. The suite sounds really interesting.
roya
I’ll be more than happy with a second Oc dedicated to hemisphere! I really enjoy it so far...

I may write a couple “apps” I need soon too... wink
Kroaton
This might be a weird question but could the USB in on the Teensy be used with the O_c and Hemisphere?

A serial PureData input app in the Hemisphere Suite would make any O_c an extender for TerminalTedium and computers running PD.
chysn
Kroaton wrote:
This might be a weird question but could the USB in on the Teensy be used with the O_c and Hemisphere?


Yeah, sure... the whole thing runs on the Teensy, so you can do whatever with the hardware on board. The native firmware optionally uses the USB port to send serial terminal debugging info, but you can do whatever you want to with it.

I'll probably get interested in USB applications at some point. I just got a flush-mount USB jack and little button to install into the back of my case's rear panel, so I can program my O_C without keeping it hanging out of my rack all the time. That will inevitably lead to me wanting to use the Teensy's USB for other stuff.
chysn
All right, folks... I can put clock division/multiplication under CV control if I get rid of the Sample and Hold applet. Does anybody care? This isn't, like, somebody's only sample and hold, right? It was basically the first proof-of-concept applet, and it seems like the least useful thing.

I feel like Ken Mattingly here trying to find 200 bytes, except nobody's going to die if I fail.

EDIT: I'm moving S&H into the Calculator. This is a nice solution. Because the Hemisphere S&H is, in fact, my system's only S&H.
AnalogAssailant
Quote:
I can put clock division/multiplication under CV control if I get rid of the Sample and Hold applet. Does anybody care?



My 2 cents..

I just had another look at your S&H video. It's a neat little app but i think i would prefer cv control over the div/mult, but others may disagree.

I suppose you could always re-introduce another (more fleshed out?) s&h applet with the second firmware path you have planned, the firmware that will see more original o_C apps removed.
mrerdat
This rules. Really enjoying the Turing machine into dual quantizer combo!
chysn
AnalogAssailant wrote:
I suppose you could always re-introduce another (more fleshed out?) s&h applet with the second firmware path you have planned, the firmware that will see more original o_C apps removed.


Perhaps. But no functionality is given up by putting S&H in the Calculator. In fact, it's even better, since it can be used in conjunction with another operation. I'm changing the Rand up a bit, too, so it plays more nicely with S&H. I'll have to do another video for Calculator.
roya
What if we had a way to easily add/remove applets with #define somewhere in the code when building the firmware? Also, maybe we can do the same for the core apps to have more room for applets?
chysn
roya wrote:
What if we had a way to easily add/remove applets with #define somewhere in the code when building the firmware? Also, maybe we can do the same for the core apps to have more room for applets?


If you're compiling it yourself, it's just a matter of removing or commenting out DECLARE_APPLET() lines from hemisphere_config.h (for Hemisphere). This is pretty straightforward, as there are no dependencies between applet files.

For the main O_C apps, see OC_apps.ino. Again, you can comment the DECLARE_APPS() lines. The apps' files will still be compiled, but the linker seems smart enough to exclude unreachable references, so there's nothing else you really need to do.

As for sizes of the apps: Copier Maschine (~12K), Harrington (~4K), Automatonnetz (~4K), Quantermain (~12K), Meta-Q (~19K), Quadraturia (~3K), Low-rents (~2K), Piqued (~6K), Sequins (~21K), D. Ping Pong (~2K), V. Sweet (~2K), Acid Curds (~13K). Each Hemisphere applet takes about 2K.

Note that the sizes above were determined with the pretty naive methodology of commenting the DECLARE_APP() line out and observing the difference in the size of the compiled binary.
Shledge
It's already fairly easy - comment it out in apps.ino and it won't compile with the rest.

You can do the same for applets - go to hemisphere_config.h and add/remove from there.
kamindustries
This looks amazing! Apologies if this is redundant, but is there a list somewhere of all the applets currently available for Hemisphere, maybe with a 1-liner description of each? Can't seem to find it anywhere.
REVIVER
kamindustries wrote:
This looks amazing! Apologies if this is redundant, but is there a list somewhere of all the applets currently available for Hemisphere, maybe with a 1-liner description of each? Can't seem to find it anywhere.


https://butmostlycrime.blogspot.com/2018/07/hemisphere-beta-release-pf aq.html
chysn
Beta 3 is now available at http://www.beigemaze.com/hemisphere.html

Release notes:

* Clock Divider/Multiplier: Division/Multiplication value can now be under CV control (positive CV sets division, negative CV sets multiplication).

* Sample and Hold/Random: Removed applet

* Calculator: Added S&H functionality (set a channel's operation to "S&H" and send a clock to that channel's digital input to sample the corresponding CV input. The held value will be at that channel's output).

* Calculator: Change Random operation to clocked random (set a channel's operation to "Rnd". If the corresponding digital input is not clocked, random voltages will be sent from that channel's output. If a clock is sent, then the random signal will be sampled and held at each clock. Turning the encoder in either direction will turn off clocked operation).

Unless there are any actual bugs found, this will be the final Hemisphere beta release.
AnalogAssailant
Awesome chysn, thanks for doing this.

This, plus your other firmware coming down the pipe, has prompted me to order two more o_C modules.
Shledge
Clock divider seems to mute itself if I try and divide by CV - works fine if I'm multiplying a clock.
chysn
Shledge wrote:
Clock divider seems to mute itself if I try and divide by CV - works fine if I'm multiplying a clock.


So, you change the clock division value via CV and it stops emitting clocks?
Shledge
Pretty much, but only on 1/1 and the divisors. It kicks back in if I change it via encoder.
chysn
Shledge wrote:
Pretty much, but only on 1/1 and the divisors. It kicks back in if I change it via encoder.


Ok, thanks! Expect a fix tomorrow.
rayultine
I installed Hemispheres! I'm in love!

Can confirm that recording CV signals into Lofi Tape is glitchy awesomeness. There's no way pitch control or one-shot play could be programmed in, is there?
All sorts of audio->CV anarchy would ensue

Can also confirm that I do not need to buy a Plog anymore.
chysn
rayultine wrote:
I installed Hemispheres! I'm in love!

Can confirm that recording CV signals into Lofi Tape is glitchy awesomeness. There's no way pitch control or one-shot play could be programmed in, is there?
All sorts of audio->CV anarchy would ensue

Can also confirm that I do not need to buy a Plog anymore.


Thanks for trying it! I appreciate it.

Lofi tape was originally a one-shot thing (with digital 1 firing it), but it felt way better for me as a looper. Digital 1 now acts as a gated pause, which behaves a lot like a one-shot mode if you provide long gates with short pauses between them... or put a short gate through an inverter.

The O_C's interrupt fires every 60µs, which seems fast, but for audio it's downright pokey. There's really no place to go on pitch adjustment. One factor up or down, and it's inaudible or a squeaky mess. And there's no floating-point arithmetic on the M4 processor. This thing just wasn't built to be a sampler smile

I am planning an audio-rate CV sequencer for Hemisphere Suite, though, so keep an eye out for that.
Shledge
Same - it pretty much turns an o_C into two separate modules. Each applet is basic in functionality, but I actually prefer that - no nonsense, straight to the point, just like my doepfer modules.
chysn
Okay, this release should fix the clock divider under CV control. I wasn't able to shake it while testing. This is still beta 3, but the file in the archive will be called Hemisphere_beta3A.hex:

www.beigemaze.com/downloads/Hemisphere_beta3.zip
rayultine
Shledge wrote:
Same - it pretty much turns an o_C into two separate modules. Each applet is basic in functionality, but I actually prefer that - no nonsense, straight to the point, just like my doepfer modules.


Also, allows really fun self-patching opportunities!
chysn
I'd like to get some feedback on this, if you guys don't mind.

Right now, there's a CV Forwarding mode that routes the two left-hemisphere CV ins to the right hemisphere. I'm thinking that this is not, in practice, a useful feature.

What seems more useful is clock forwarding. I've been playing with this. Like CV forwarding, clock forwarding is activated by long-pressing the Down button. When you release it, the right hemisphere ignores its own incoming clock signals and clocks to the left hemisphere's incoming clock. The awesome thing is that you can hold the Down button until you want to switch over the clock, to time things just right.

I don't want to include both CV and clock forwarding. It's just too awkward, and this whole project is designed to avoid awkward interface elements. But I've just been having way more damn fun with clock forwarding than I ever did with CV forwarding, especially in Hemisphere Suite where I have the Turing Machine and the Euclidean drummer going at the same time.

What do you guys think?
rayultine
Uhh... I haven't played with that feature. Seems like it kinda needs a video to explain how both features work, and why CV forwarding isn't as much fun as you thought.

But yeah, I guess everytime I try to think of a reason where CV forwarding would be useful I come to the realization that I'm thinking of clocks and triggers rather than modulating CV. When you conceived the idea, which app combinations did you think would be most relevant?

Quote:
Lofi tape was originally a one-shot thing (with digital 1 firing it), but it felt way better for me as a looper. Digital 1 now acts as a gated pause, which behaves a lot like a one-shot mode if you provide long gates with short pauses between them... or put a short gate through an inverter.


That's where the allure of menus reveals itself. It would be wonderful to be able to switch that behavior to economize modules and patch cables razz

Quote:

There's really no place to go on pitch adjustment. One factor up or down, and it's inaudible or a squeaky mess.


Would it still be fun to play with fine-grained adjustments? Are you sure that "squeaky mess" doesn't sound awesome/weird to someone else's ears? lolz
BaloErets
chysn wrote:
I'd like to get some feedback on this, if you guys don't mind.

Right now, there's a CV Forwarding mode that routes the two left-hemisphere CV ins to the right hemisphere. I'm thinking that this is not, in practice, a useful feature.

What seems more useful is clock forwarding. I've been playing with this. Like CV forwarding, clock forwarding is activated by long-pressing the Down button. When you release it, the right hemisphere ignores its own incoming clock signals and clocks to the left hemisphere's incoming clock. The awesome thing is that you can hold the Down button until you want to switch over the clock, to time things just right.

I don't want to include both CV and clock forwarding. It's just too awkward, and this whole project is designed to avoid awkward interface elements. But I've just been having way more damn fun with clock forwarding than I ever did with CV forwarding, especially in Hemisphere Suite where I have the Turing Machine and the Euclidean drummer going at the same time.

What do you guys think?


I think you nailed it. Clock forwarding is so much more useful. If there was ever a way to be able to divide/multiply the clock being forwarded, that would be bananas nanners It's peanut butter jelly time! Rockin' Banana!

Thanks again for bringing this to the community. Really nice to finally see some addons to O_c pop up. What a great module with incredible devs like yourself supporting it! we're not worthy
AnalogAssailant
Quote:
What seems more useful is clock forwarding


Yeah thinking about it sharing the clock is much more likely to be done than sharing cv inputs.

Having said that I am the kind of person that would keep it all and omit nothing, but i get where you're coming from and so if i have to choose to keep one then it would be clock forwarding.
chysn
rayultine wrote:
Uhh... I haven't played with that feature. Seems like it kinda needs a video to explain how both features work, and why CV forwarding isn't as much fun as you thought.


Fair enough. There is a brief demonstration in the introductory video at about the 3-minute mark:https://www.youtube.com/watch?v=JnGpu2jqeg8&t=3m0s. The CV data from CV inputs 1 and 2 are just passed to inputs 3 and 4.

Quote:
When you conceived the idea, which app combinations did you think would be most relevant?


I think I had the Calculator in mind, but it's such a narrow use case. Forwarding a single clock input makes so much more sense.

Quote:
Would it still be fun to play with fine-grained adjustments? Are you sure that "squeaky mess" doesn't sound awesome/weird to someone else's ears? lolz


It certainly could somebody's cup of tea, but there are plenty of ways to get that sound. Mainly my point is that there are no fine-grained adjustments. The sample rate must be the O_C's interrupt speed over some integer divisor, and this divisor has a narrow audible zone that makes pitch modulation less useful than the other options.
chysn
UPDATE: Beta 4 is now available at http://www.beigemaze.com/hemisphere.html

The only differences involve Master Clock Forwarding.

* CV Forwarding is replaced with Master Clock Forwarding.

Master Clock Forwarding works like this: Long-press the Down button for about 2 seconds. When you release it, you will activate Master Clock Forwarding. A clock icon will appear at the top of the display when Master Clock Forwarding is on.

When Master Clock Forwarding is on, an applet in the right hemisphere looking for a clock (or trigger) at its first input will look instead at the left hemisphere's first input. This allows you to clock two applets with a single clock signal patched to Ornament and Crime. Since Master Clock Forwarding is not active until the button is released, you may time activation and deactivation in time to your music.

To deactivate Master Clock Forwarding, long-press the Down button again.

Master Clock Forwarding applies when the the right hemisphere applet is looking for a clock. It does nothing when the applet is looking for a gated signal. The following applets will use the Master Clock:

Brancher
Dual Calculator (S&H or clocked random, channel 1 only)
Clock Divider
Dual Quantizer (channel 1 only)
Sequence5
Skewed LFO (reset on Master Clock)
Switch (sequential channel 1)
Turing Machine
Trigger Sequencer

* The Switch has been updated so that channel 1 is the sequential channel and channel 2 is the gated channel. This allows Switch to use the Master Clock.
SlyFrank
A recent post said this can take the place of a Plog. I love o_C, and have built a few, so I have one that I could load up with this cool-looking firmware (Meta-Q is one of my favorite o_C apps, so that is one of the apps I would like least to see be jettisoned, but no issue as I have an extra o_C...).

I have been weighing my options for a logic module for a while, but haven't pulled the trigger. This would be great if I could just load this firmware and have a full-featured logic module for 'free.'

I watched the YT video on the logic app but didn't get much detail out of it. Since I have an extra o_C module lying around this could resurrect it!

1) How many inputs are there for the logic - two or three?

2) Is it a dual logic module, like the Plog?

2) Can the logic 'flavors' be CV'd, like Plog?

None of these are deal-breakers, just trying to wrap my head around the logic app. Awesome work on this!

Thanks!
chysn
Remember that every Hemisphere applet splits the O_C exactly in half. So each one has access to two digital ins, two CV ins, two outs, one enocder, its corresponding button, and a 64x64 display.

So the Logic applet is configured as follows:

* The two digital inputs are the logical operands, in the form of high or low signals, like a gate
* The encoder is used to select a logical operator (and, or, xor, nand, nor, xnor) for each channel. The channel can also be set to “CV” to put the operator selection under CV control
* When the operator type is set to “CV,” the CV inputs set the operator for the corresponding channel
* The encoder button is used to move the selection cursor between the two channels
* The two CV outputs output the result of the corresponding operator on the two operands

It’s sort of a dual logic module in the sense that two logical operations are performed and output. But both operations use the same two inputs. Of course, you can set both Hemispheres to use the Logic applet for quad logic with two independent pairs of inputs.
BaloErets
SlyFrank wrote:
A recent post said this can take the place of a Plog. I love o_C, and have built a few, so I have one that I could load up with this cool-looking firmware (Meta-Q is one of my favorite o_C apps, so that is one of the apps I would like least to see be jettisoned, but no issue as I have an extra o_C...).

I have been weighing my options for a logic module for a while, but haven't pulled the trigger. This would be great if I could just load this firmware and have a full-featured logic module for 'free.'

I watched the YT video on the logic app but didn't get much detail out of it. Since I have an extra o_C module lying around this could resurrect it!

1) How many inputs are there for the logic - two or three?

2) Is it a dual logic module, like the Plog?

2) Can the logic 'flavors' be CV'd, like Plog?

None of these are deal-breakers, just trying to wrap my head around the logic app. Awesome work on this!

Thanks!


The Logic applet in Hemisphere is indeed amazing, and can recreate many of the functions of Plog, but doesn't replace it. I think in most situations, it really depends on your needs. Plog is a Logic workhorse. What you would be loosing is;

- Third input per logic operation
- built in attenuation/offset for CV over logic operations
- Toggle Flip Flop
- Data Flip Flop

There's also some nice normalization in Plog between the channels which can be configured to your liking via jumpers on the back.

I think the beauty of Hemisphere is that it packs a whole bunch of basic functions into a very clear and easy to use package. Depending on people's needs, basic functionality can often cover the task at hand. But I don't see it replacing dedicated modules anymore than thinking it was meant to do so thumbs up
Shledge
Flip flops can be done with the clock divider applet, I believe.
chysn
antofthesky wrote:
Are the help screen contents for each applet online anywhere, instead of just on the module? Would be nice to read.


These are finished, and available on the GitHub wiki: https://github.com/Chysn/O_C-HemisphereSuite/wiki
sushiluv
Github has Hemisphere Suite Apps listed, does that mean it is allready availabe? Can´t find a hex for it, thanks.
rayultine
I'm a nerd that likes to print manuals and user guides. I made a pdf of chysn's applet descriptions:

www (dot) sendspace.com/file/ibjd9o

Don't get a virus from sendspace ads please.
chysn
sushiluv wrote:
Github has Hemisphere Suite Apps listed, does that mean it is allready availabe? Can´t find a hex for it, thanks.


The ones listed are available in the repository, but I don't have a hex file yet. Let me get a few things squared away and I'll make the hex file for Hemisphere Suite available early next week.

rayultine wrote:
I'm a nerd that likes to print manuals and user guides. I made a pdf of chysn's applet descriptions:

www (dot) sendspace.com/file/ibjd9o


Thanks!
BaloErets
Shledge wrote:
Flip flops can be done with the clock divider applet, I believe.


Perhaps, but if you're using the clock divider app, then you only have 1 channel of logic, which would bring it even further away from the full functionality of a Plog.
chysn
The Plog discussion is interesting, because I only have 184HP, and I very nearly bought a Plog instead of an Ornament and Crime. I don't think I would have regretted going that way.

In other news, I finished a new applet for Hemisphere Suite called "Lower Renz," which is a single-generator version of Low-Rents. I underestimated what a fun algorithm that is. I don't understand what it's doing, but it's fun.
naturarerum
just flashed this on my uo_c! watching this thread very closely smile)
chysn
naturarerum wrote:
just flashed this on my uo_c! watching this thread very closely smile)


Thanks for trying it!
Shledge
BaloErets wrote:
Shledge wrote:
Flip flops can be done with the clock divider applet, I believe.


Perhaps, but if you're using the clock divider app, then you only have 1 channel of logic, which would bring it even further away from the full functionality of a Plog.


You can also do it with the logic neuron applet - I believe you can make many forms of logic on that alone.
dooj88
chysn wrote:
I feel like Ken Mattingly here trying to find 200 bytes, except nobody's going to die if I fail.


lolspew took me a second to remember where i knew that name.

enjoying watching the progress on this. don't have much wiggling time, but keeping this app my pocket for a rainy day
chysn
Hey, I have a request. If anybody is using Hemisphere Beta 4, could you please see if you can access and use the Quadraturia app? I have a report that Quadraturia is freezing as soon as it's started, but I cannot reproduce this. I know that one of them is a µO_C, but I'm not sure what difference that makes.
SkoPe
chysn Just checked and it also freezes when selecting Quadraturia app.
chysn
SkoPe wrote:
chysn Just checked and it also freezes when selecting Quadraturia app.


OK, thanks! Does it freeze with any other apps, or just Quadraturia? Also, what kind of O_C do you have?
Roni
chysn wrote:
I know that one of them is a µO_C, but I'm not sure what difference that makes.


I think the only differences are the MCP6004 is replaced by 2 x MCP6002 and of course Jim's "Either Way" power. Everything else is identical.
SkoPe
chysn
With Beta 3 and Hemisphere app loaded it was impossible to change to Harrington1200, would freeze immediately...didn’t try any other apps.

Now in Beta 4 freezes when selecting Quadraturia, other apps can be loaded no problem.

Mine was built with mxmxmx rev 2d PCB

Thanks a lot for sharing this wonderful work! Having a great fun! This is fun!
chysn
To quote then-Admiral Kirk, this is damn peculiar. My best guess at the moment is that it's a memory issue. I'll try pulling out the biggest memory hog (LoFi Tape) and see if that improves stability. SkoPe, I appreciate you taking the time to check this!
chysn
Beta 5 is available at http://beigemaze.com/hemisphere.html

* Added reset to clock divider
* Made some display updates for line segment editors (ADSR, LFO, etc.)
* Fixed sequencer reset timing
REVIVER
Will the Lorenz applet be added in an upcoming beta? Very interested in that one.

screaming goo yo

EDIT:

Hopefully not repeating any requests from other users (have only grazed this thread otherwise)....

Not sure if it's possible, but having a way to join the two trigger sequencers in the one applet to make 16 steps would be nice.

Any way of joining 2x of the 5 Step Sequencer applets to create a 10 step sequence in a similar sense? Obviously, Master Clock Forwarding will clock them both in time, but to have one side forward to the next sequencer after the last step and so on would be amazing.

EDIT EDIT: Ah, I see in the Beta 4 notes (missed that one) which states that only Channel one on the Dual Calc Random is clocked. Got it! thumbs up
rayultine
REVIVER wrote:

Not sure if it's possible, but having a way to join the two trigger sequencers in the one applet to make 16 steps would be nice.

Any way of joining 2x of the 5 Step Sequencer applets to create a 10 step sequence in a similar sense? Obviously, Master Clock Forwarding will clock them both in time, but to have one side forward to the next sequencer after the last step and so on would be amazing.


Yah! It would be amazing to replicate the Sputnik 5-step or be able to do some Cartesian sequencing.
chysn
REVIVER wrote:
Will the Lorenz applet be added in an upcoming beta? Very interested in that one.


LowerRenz is in Hemisphere Suite, which excludes all of the native O_C apps. I learned that the Lorenz modulation was awesome, and didn't want to totally lose it when I'm running only Hemisphere. Right now, I think there are 20 or 21 applets in Hemisphere Suite. You can see the complete list at https://github.com/Chysn/O_C-HemisphereSuite/wiki I'm planning on making a hex file available on Friday, July 20th.

Quote:
Not sure if it's possible, but having a way to join the two trigger sequencers in the one applet to make 16 steps would be nice.


That one is just 2x8. Since they share a clock, they could be joined with a 1/8 clock divider and a Doepfer A-152 or other switch. It's likely that I'll do a 1x16 version of it, though. I kind of like that 4-bit editing scheme.

Quote:
Any way of joining 2x of the 5 Step Sequencer applets to create a 10 step sequence in a similar sense? Obviously, Master Clock Forwarding will clock them both in time, but to have one side forward to the next sequencer after the last step and so on would be amazing.


It's sort of the same story as above; you can double anything with a clock divider and a switch!

Quote:
I also noticed that in beta 5, if the Dual Calculator applet is on the right, the left Rand will pick up the clock from the left applet in Master Lock Forwarding as expected. However, if you open another Rand on the right of the Dual Calc applet, it doesn't pick up the clock. Is this by design? It'll pick it up if I mult it and then patch to the second digital input...


This is by design; master clock forwarding only passes on Trigger 1. I went back and forth on that idea, but the single-clock forward provides way more flexibility than forwarding both inputs. For example, I think it's important to have independent control over resets, or to be able to take a channel entirely off the grid.

Thanks for trying it, and for offering your thoughts! I really do appreciate it.
chysn
rayultine wrote:
or be able to do some Cartesian sequencing.


What is Cartesian sequencing?
REVIVER
chysn wrote:

LowerRenz is in Hemisphere Suite, which excludes all of the native O_C apps. I learned that the Lorenz modulation was awesome, and didn't want to totally lose it when I'm running only Hemisphere. Right now, I think there are 20 or 21 applets in Hemisphere Suite. You can see the complete list at https://github.com/Chysn/O_C-HemisphereSuite/wiki I'm planning on making a hex file available on Friday, July 20th.

Thanks for trying it, and for offering your thoughts! I really do appreciate it.


Thank you for developing this! I'm finding it incredibly useful in a lot of situations. Looking forward to the suite.
rayultine
chysn wrote:
rayultine wrote:
or be able to do some Cartesian sequencing.


What is Cartesian sequencing?


A 5x5 xy grid - sorta like the Make Noise Rene. Hey, I'm just using their marketing term razz
chysn
rayultine wrote:
chysn wrote:
rayultine wrote:
or be able to do some Cartesian sequencing.


What is Cartesian sequencing?


A 5x5 xy grid - sorta like the Make Noise Rene. Hey, I'm just using their marketing term razz


Sorry, I don't know much about Rene. Each cell holds a specific voltage? Do you address the cell with CV X and Y, or clocked X and Y, or some other way? Now that you say it's a grid, the term Cartesian makes sense.
Mashmore
EDIT Nevermind. Found it!

It is probably staring me in the face... but where do i snag the suite? Ill dig more on GitHub in the meantime
Timmy
rayultine wrote:
chysn wrote:
rayultine wrote:
or be able to do some Cartesian sequencing.


What is Cartesian sequencing?


A 5x5 xy grid - sorta like the Make Noise Rene. Hey, I'm just using their marketing term razz


Or sorta like the Automatonnetz app in O&C...
rayultine
err yeah of course! But I always thought Automatonnetz was more for chord voicing, and since I'm a lout when it comes to music theory...
chysn
Mashmore wrote:
It is probably staring me in the face... but where do i snag the suite? Ill dig more on GitHub in the meantime


https://github.com/Chysn/O_C-HemisphereSuite
Shledge
I downloaded the suite and was able to fit most of the normal apps back in with it. This really should be put on the main o_C wiki at this point as an alt firmware.

I think I found another bug on the clock divider app - I notice it won't change by CV consistently if I use a gate (eg. use a gate to instantly change from 1/1 to 2/1 for ratcheting), unless I slew it slightly. Other types of CV like LFOs etc. work fine.

I also noticed the ADSR is again not snappy, but I fixed that by multiplying the time value for attack from 200 to 1600. Arse-smackingly fast now. cool
Mashmore
chysn wrote:
Mashmore wrote:
It is probably staring me in the face... but where do i snag the suite? Ill dig more on GitHub in the meantime


https://github.com/Chysn/O_C-HemisphereSuite


Got it up and running and gave it a quick spin. Thanks so much! Loving it. Going to put it three it’s paces more today. Keep up the amazing work. nanners
chysn
Shledge wrote:
I downloaded the suite and was able to fit most of the normal apps back in with it.


Sure, Hemisphere Suite doesn't really use that much memory right now. The reason that I removed everything is that I know I'll eventually use all the remaining space; and as a "product," I don't want to have to slice things out as I scale up.

Of course, for those with the know-how to move stuff in and out at will, you've got the best of all worlds!

Quote:
I think I found another bug on the clock divider app - I notice it won't change by CV consistently if I use a gate (eg. use a gate to instantly change from 1/1 to 2/1 for ratcheting), unless I slew it slightly. Other types of CV like LFOs etc. work fine.


This is as designed. The issue here is that a gate goes immediately from 5V to 0V. O_C has no way to detect whether a cable is actually plugged in, so it can't differentiate whether a 0V reading means "nothing is plugged into me" or "I'm receiving 0V from a CV source," so the software needs to pick one of those options. In this case, since we also want to be able to control the setting from the panel, 0V is interpreted as "nothing is plugged in." Otherwise, a reading of 0V would always override the panel setting.

The Hemisphere framework provides a simulated CV "center detent" for this kind of thing. So to set 1/1 via CV, you need to provide slightly more (or less) than 0V. If you want to use a gate, you'll have to provide some offset to it.

Quote:
I also noticed the ADSR is again not snappy, but I fixed that by multiplying the time value for attack from 200 to 1600. Arse-smackingly fast now. cool


Thanks! I must have introduced an issue when adding modulation to attack. I'll look into this.
REVIVER
Just loaded Hemisphere Suite...haven't had a chance to play with it just yet, but I noticed a funky little quirk at the main menu where if you scroll down to References and then continue to scroll (i.e. turn the right knob clockwise once again), the full menu shifts down to the lower half of the screen.

Obviously not an issue, but thought it was worth mentioning.

Very intrigued re: "Darkest Timeline"... eek!
chysn
REVIVER wrote:
Very intrigued re: "Darkest Timeline"... eek!


It's a sequencer with kind of a strange feature set. It's programmed with a CV source rather than directly from the panel. It has 32 steps, with an index point so that you can start a sequence of any length from any point. It has an alternate timeline output, which effectively makes it 1x32 step or 2x16 step or, in record mode, a CV delay. It has trigger outputs with probability and complementary probability. It's not complicated, but it's a lot of fun. I'll probably update it soon to use the quantizer.

https://www.youtube.com/watch?v=sMHv4XcWU8s
https://www.youtube.com/watch?v=F1LWZ8wfUaI
Shledge
chysn wrote:


This is as designed. The issue here is that a gate goes immediately from 5V to 0V. O_C has no way to detect whether a cable is actually plugged in, so it can't differentiate whether a 0V reading means "nothing is plugged into me" or "I'm receiving 0V from a CV source," so the software needs to pick one of those options. In this case, since we also want to be able to control the setting from the panel, 0V is interpreted as "nothing is plugged in." Otherwise, a reading of 0V would always override the panel setting.

The Hemisphere framework provides a simulated CV "center detent" for this kind of thing. So to set 1/1 via CV, you need to provide slightly more (or less) than 0V. If you want to use a gate, you'll have to provide some offset to it.


That makes sense, luckily I've loads of offsets. smile
REVIVER
In the Scope applet, is it expected to see a bit of a CV offset on the numeric CV monitor with nothing plugged into the CV inputs?

Mine seems to be fluctuating rapidly between 11 and 15 on the readout.

Possibly a CV offset introduced somewhere else? Perhaps a build issue?

hmmm.....
chysn
REVIVER wrote:
In the Scope applet, is it expected to see a bit of a CV offset on the numeric CV monitor with nothing plugged into the CV inputs?

Mine seems to be fluctuating rapidly between 11 and 15 on the readout.

Possibly a CV offset introduced somewhere else? Perhaps a build issue?

hmmm.....


This is normal. There's always just a little bit of noise hanging out in the ADC. It's like .15% of the full range.
REVIVER
chysn wrote:

This is normal. There's always just a little bit of noise hanging out in the ADC. It's like .15% of the full range.


Got it! Thank you for the clarification. Appreciated.

Guinness ftw!
Mashmore
Bug report I do believe.

I have a/b set to lowrenz. A couple lfo's into cv1/2.
When i go to select an algorithm for outs c/d when i get to lowrenz it freezes.

(EDIT) It also freezes when i scroll past the MIDI In applet as well.

(EDIT)(EDIT) I was getting it to freeze by scrolling any app really. So out of curiosity I unplugged all cables from the O/c and now it does not behave this way.
So I got the bug this way.

i had nothing patched in to trig1-4.
I had an lfo in cv1 and an lfo into cv 2

I had ABC patched into plaits and nothing patched from D yet.

Hope this helps! nanners
Shledge
Mashmore wrote:
Bug report I do believe.

I have a/b set to lowrenz. A couple lfo's into cv1/2.
When i go to select an algorithm for outs c/d when i get to lowrenz it freezes.

(EDIT) It also freezes when i scroll past the MIDI In applet as well.

It appears to be a freeze on the display portion only. I seem to still be getting cv out of the O_c when this happens


What version of teensyduino and arduino IDE are you using?
Mashmore
Shledge wrote:
Mashmore wrote:
Bug report I do believe.

I have a/b set to lowrenz. A couple lfo's into cv1/2.
When i go to select an algorithm for outs c/d when i get to lowrenz it freezes.

(EDIT) It also freezes when i scroll past the MIDI In applet as well.

It appears to be a freeze on the display portion only. I seem to still be getting cv out of the O_c when this happens


What version of teensyduino and arduino IDE are you using?


Aurdrino 1.8.1 Teensydruino 1.35
Teensyloader app is 1.4 not sure if that matters.
chysn
Mashmore wrote:
Shledge wrote:
Mashmore wrote:
Bug report I do believe.

I have a/b set to lowrenz. A couple lfo's into cv1/2.
When i go to select an algorithm for outs c/d when i get to lowrenz it freezes.

(EDIT) It also freezes when i scroll past the MIDI In applet as well.

It appears to be a freeze on the display portion only. I seem to still be getting cv out of the O_c when this happens


What version of teensyduino and arduino IDE are you using?


Aurdrino 1.8.1 Teensydruino 1.35
Teensyloader app is 1.4 not sure if that matters.


I'm not able to reproduce this, but I will have a hex file available within the next few days--after I take care of a few things--and I'll be interested in the stability of that.

Note that I identified a similar issue in a revision that I pushed yesterday, and it's since been corrected. I had finally strained the processor a bit too far in the Game of Life, and the display would lock up and O_C required a restart. Scrolling through applets quickly was a reliable way to reproduce that crash. So if your build was too "old," you might have been caught up in that whole mess. Keep in mind that Hemisphere Suite is pre-beta and I'm still doing a lot of fooling around.
Mashmore
chysn wrote:
Mashmore wrote:
Shledge wrote:
Mashmore wrote:
Bug report I do believe.

I have a/b set to lowrenz. A couple lfo's into cv1/2.
When i go to select an algorithm for outs c/d when i get to lowrenz it freezes.

(EDIT) It also freezes when i scroll past the MIDI In applet as well.

It appears to be a freeze on the display portion only. I seem to still be getting cv out of the O_c when this happens


What version of teensyduino and arduino IDE are you using?


Aurdrino 1.8.1 Teensydruino 1.35
Teensyloader app is 1.4 not sure if that matters.




I'm not able to reproduce this, but I will have a hex file available within the next few days--after I take care of a few things--and I'll be interested in the stability of that.

Note that I identified a similar issue in a revision that I pushed yesterday, and it's since been corrected. I had finally strained the processor a bit too far in the Game of Life, and the display would lock up and O_C required a restart. Scrolling through applets quickly was a reliable way to reproduce that crash. So if your build was too "old," you might have been caught up in that whole mess. Keep in mind that Hemisphere Suite is pre-beta and I'm still doing a lot of fooling around.


Right on, Thanks for getting back. Coincidenctly I did install this and pulled it from git hub yesterday early afternoon. Keep on keepin’ On man. Loving the ease of use.
chysn
Mashmore wrote:
Right on, Thanks for getting back. Coincidenctly I did install this and pulled it from git hub yesterday early afternoon.


Oh, yeah. The fix was pushed last night at about 10:30PM Eastern Time (U.S.A.), or 0230 GMT Wednesday morning. For a couple days prior to that, you'd definitely get crashes.
Mashmore
chysn wrote:
Mashmore wrote:
Right on, Thanks for getting back. Coincidenctly I did install this and pulled it from git hub yesterday early afternoon.


Oh, yeah. The fix was pushed last night at about 10:30PM Eastern Time (U.S.A.), or 0230 GMT Wednesday morning. For a couple days prior to that, you'd definitely get crashes.


Just to confirm. I loaded the new juice up. All good in the hood. applause
qbassa@gmail.com
Is there compiled (.hex) version of Suite available, or somebody can send it to me? Best!
chysn
qbassa@gmail.com wrote:
Is there compiled (.hex) version of Suite available, or somebody can send it to me? Best!


I'm just finishing a few things up. It will be available tomorrow. I'll post the URL in this topic.
chysn
qbassa@gmail.com wrote:
Is there compiled (.hex) version of Suite available, or somebody can send it to me? Best!


Officially, this is for release tomorrow, and I'll make posts on MuffWiggler, Facebook, and my blog in 24 hours.

But for you folks who have been paying attention this whole time, and helping me get it right, I can't do much more than offer my sincere thanks, and give you a chance to get it a day early.

https://github.com/Chysn/O_C-HemisphereSuite/releases
aragorn23
This is really impressive work! The suite has saved me buying at least three other modules hyper

If you're taking feature requests, I've always wished that the o_c sequencing stuff had clock shuffle on some of the apps (well, it would be nice if temps utile had that too). I'm guessing this wouldn't be too difficult to add, or that a dedicated clock shuffle (or even rotating clock divider) app may be possible?

Anyway, congrats again on creating so many useful extensions to o_c so quickly!
Shledge
Could just use a clock source that has shuffle
aragorn23
Indeed, but I don't have such a clock source, hence my hope that o_c could at some point fulfil this role hihi
chysn
aragorn23 wrote:
If you're taking feature requests, I've always wished that the o_c sequencing stuff had clock shuffle on some of the apps (well, it would be nice if temps utile had that too). I'm guessing this wouldn't be too difficult to add, or that a dedicated clock shuffle (or even rotating clock divider) app may be possible?


There will definitely be additional applets and full apps coming. With this firmware, Ornament and Crime is 55% full, so there's a lot of space now.

I'm reluctant to add a master clock, for various reasons. But a shuffling clock divider could be cool. I'll have to look into the functional details.
qbassa@gmail.com
Just checked Scope applet. Have some questions
1. CV1 draws scope, Cv2 only changes value on the bottom, but doesn't draw anything. Is that normal?
2. long pressing encoder goes in help screen but to return I need to hold and select app again. Is it normal?
3. Another mode with scope made of lines not points would be nice in some cases (square signals)
4. Now trigger sets bpm. Another mode with trigger resetting drawing position would be nice.
Anyway it is a great work!

Best,
Jakub Kostynowicz
SuperNiCd
Following this with interest - looks really cool!!

I think I understand but just for the sake of clarity...

Hemispheres is a (possibly complete?) lightweight version that leaves all the original O_c apps except MetaQ intact.

Hemispheres Suite is a full (in progress?) version that removes all O_c apps except References and adds more new applets and apps.

Is that right?
chysn
qbassa@gmail.com wrote:
Just checked Scope applet. Have some questions
1. CV1 draws scope, Cv2 only changes value on the bottom, but doesn't draw anything. Is that normal?


That's right.

Quote:
2. long pressing encoder goes in help screen but to return I need to hold and select app again. Is it normal?


Long-pressing the left encoder cycles through the help screens, and then back to the split applet screen. So, long-press left to see the first help screen, long-press left again to see the second help screen, and long-press left a third time to get back to the regular screen. If the same applet is selected for both hemispheres, you'll go back to the regular screen after the second long-press of the left encoder.

I considered making it so that an encoder movement or button push would bring you out of help mode. But I thought it was important for applets to continue to work as normal "in the blind," so that's why I'm using the long-presses instead.

Quote:
3. Another mode with scope made of lines not points would be nice in some cases (square signals)


This is a matter of adjusting the sample rate. If you get the right setting, it will look like lines instead of dots. Edit: I think I understand what you're saying. Maybe. I don't think people expect to see the vertical segments of square waves on a scope, because they're not real.

Quote:
4. Now trigger sets bpm. Another mode with trigger resetting drawing position would be nice.


I think that's a great idea. The second trigger could be used to auto-select the sample rate. I'll play around with this to see if it works right. It could be a real convenience.
chysn
SuperNiCd wrote:
Hemispheres is a (possibly complete?) lightweight version that leaves all the original O_c apps except MetaQ intact.

Hemispheres Suite is a full (in progress?) version that removes all O_c apps except References and adds more new applets and apps.

Is that right?


Yeah, that's exactly right. Hemisphere Suite is a full replacement.

Also, there's no room left to make any changes for Hemisphere, while Hemisphere Suite will be updated pretty frequently.
chysn
qbassa@gmail.com wrote:

4. Now trigger sets bpm. Another mode with trigger resetting drawing position would be nice.


This was a damn good idea. You'll be able to plug something like Maths's end-of-rise trigger into Trigger 2 and it'll set the scale and track Maths perfectly. Also works to "freeze" a grain waveform with Phonogene's EOS trigger. This will be available in next week's release. Thanks!
qbassa@gmail.com
Great to hear. Looking forward to play with it!

Best,
Jakub
Misk
Yeah, this is REALLY useful, and I love the way you've created everything as little applets, so there's no long-presses to get out to the menu to switch things up! at this point for me hemispheres is pretty indispensable for my uO_C, and it's allowed me to test the waters with a few logic toys without throwing down on modules!

guess it's time to build my other uO_C pcb!

I'm curious, I don't have a Temps Utile so I'm not as familiar with it, but it's mostly for clocking, gates and triggers right? is there much difference in the hardware between the O_C and the Temps?

Could one add, say, a probabilistic gate skipper/filter (like a skipmin) to the O_C? I'm not asking you—you're busy we're not worthy but just curious about the hardware differences between the two.

Massive thanks for sharing in your coding experiments!
chysn
Misk wrote:
Could one add, say, a probabilistic gate skipper/filter (like a skipmin) to the O_C? I'm not asking you—you're busy we're not worthy but just curious about the hardware differences between the two.


Hey, thanks for trying it out!

I don't know anything about Temps Utile (I mean... it LOOKS the same!), but I'd like to learn more about the probabilistic gate skipper. It sounds like the sort of thing I'd be interested in doing. Are there any existing modules that do this that I could steal their ideas? smile

I'm just finishing up the MIDI Out applet. I had no idea that playing a Korg M1 plug-in with my modular could be so magical!
DeltaTucanae
chysn wrote:
Misk wrote:
Could one add, say, a probabilistic gate skipper/filter (like a skipmin) to the O_C? I'm not asking you—you're busy we're not worthy but just curious about the hardware differences between the two.


Hey, thanks for trying it out!

I don't know anything about Temps Utile (I mean... it LOOKS the same!), but I'd like to learn more about the probabilistic gate skipper. It sounds like the sort of thing I'd be interested in doing. Are there any existing modules that do this that I could steal their ideas? smile

I'm just finishing up the MIDI Out applet. I had no idea that playing a Korg M1 plug-in with my modular could be so magical!


I would also love this.

Here's one
https://www.modulargrid.net/e/ladik-s-090-dual-probability-skipper

If that is possible I imagine a burst app would be also possible?
https://www.modulargrid.net/e/ladik-s-075-burst-generator
chysn
Probability Skipper is 15 minutes of work, and will be a great add, so consider it done. Probably Burst, too. Ladik is some fertile ground for ideas.
chysn
I've added the Dual Probability Skipper. Works pretty much exactly like the Ladik. It will go into next Friday's hex file, or it's available in the repo if you can compile it.
DeltaTucanae
chysn wrote:
I've added the Dual Probability Skipper. Works pretty much exactly like the Ladik. It will go into next Friday's hex file, or it's available in the repo if you can compile it.


That is great, thank you
Eurtrude
This is becoming more than amazing ! & no bugs to report so far.

I was wondering if it would be possible to include a Bernoulli Gate app ? Similar as Branches from Mutable ? That would be awesome smile

Thanks again for your great work !
chysn
Eurtrude wrote:
I was wondering if it would be possible to include a Bernoulli Gate app ? Similar as Branches from Mutable ? That would be awesome smile


Brancher was supposed to be this, but I got it all wrong. I might just update Brancher to work like it was supposed to in the first place, to be a Bernoulli gate. I think it's pretty pointless as it is now.

Quote:
Thanks again for your great work !


Thanks for trying it!
Eurtrude
chysn wrote:
Eurtrude wrote:
I was wondering if it would be possible to include a Bernoulli Gate app ? Similar as Branches from Mutable ? That would be awesome smile


Brancher was supposed to be this, but I got it all wrong. I might just update Brancher to work like it was supposed to in the first place, to be a Bernoulli gate. I think it's pretty pointless as it is now.



That's what I thought at the beginning, given the name of the app... but the current app is not pointless to my taste, it's like a probability base sequential switch and it can be a quick way to add a bit life in a patch.
chysn
I had to do an update because I found a serious problem with the restore-from-menu behavior. So there's a new version available:

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.1

This also adds the Clock Skipper (S-090 simulation), and changes Brancher to a Bernoulli Gate. (Don't worry, I plan to release the previous Brancher under a new name next week). This release also adds MIDI Out to go with MIDI In.

I'm still going to do a more comprehensive MIDI Interface full app.
Misk
chysn wrote:
I've added the Dual Probability Skipper. Works pretty much exactly like the Ladik. It will go into next Friday's hex file, or it's available in the repo if you can compile it.


My god, you're fuckin amazing!! Thanks so much! seeing all your enthusiasm with this has got me remembering that the O_C isn't just a module, but a whole platform of possibilities! It's got me thinking of all the other modules that use a teensy or some derivative...

Would it be too much to ask if you could update the varigate 8+ firmware after this? hihi
Misk
now that i'm thinking about it, I haven't built a Temps, but I know the DAC for the O_C has gotta be pretty nice, since it costs like $15 for one on mouser—and the Temps doesn't use the same DAC—I wonder if one could effectively add all the functions of the Temps into the O_C.

The only problem would be a lack of memory—but couldn't anyone who understood pcb design, put together an O_C with a microSD reader or something?

I don't know, I only have experience with basic electronics and building other peoples pcbs—but that could be cool—to have a way to store recorded CV, or sequencer data, etc.

a little off-topic and probably way more complicated than just slapping a microSD card onto the existing board tho.
chysn
Misk wrote:
seeing all your enthusiasm with this has got me remembering that the O_C isn't just a module, but a whole platform of possibilities!


Yeah, that's what I like about it. I've looked into pretty much everything that's open-source or programmable, and O_C really hits the sweet spot between flexibility and cost. It's really kind of ridiculous. Obviously, we always want more (I want LEDs and two more knobs and SD storage and better ADCs!), but if you know C++ and want to go from idea to patching in a few hours, there's nothing better

Quote:
The only problem would be a lack of memory—but couldn't anyone who understood pcb design, put together an O_C with a microSD reader or something?


I'm looking into storage options. It was important, for my purposes, to tackle MIDI first. Now that my O_C has a line to the outside world, I'm thinking that there's no reason that I can't transfer sets of data via system exclusive dumps. This would save me from having to do the MicroSD thing. But that's also on the table.
digitalohm
any thoughts on adding a dual gate delay app?
Mashmore
Over here dreaming about a ratcheting sequencer but to be honest I haven’t spent enough time with the divider/multiplier which may get me close.

So much to explore! Thanks again for all the work you have put in to this.
BaloErets
Loving the suite! O_c is one of the few modules that I've bought a 2nd of, and it's great to have the different firmware installed in each!

It's the Darkest Timeline part of the Hemisphere suite? I've been quite curious to try it out.
chysn
digitalohm wrote:
any thoughts on adding a dual gate delay app?


Does this delay the onset of a gate signal, or delay a clock signal? If it delays the onset of a gate, does it similarly delay the end of that gate, or does the gate end immediately when the incoming gate goes low?
chysn
BaloErets wrote:
Loving the suite! O_c is one of the few modules that I've bought a 2nd of, and it's great to have the different firmware installed in each!

It's the Darkest Timeline part of the Hemisphere suite? I've been quite curious to try it out.


Yeah, go out to the main menu and it's the second item, right under Hemisphere. Documentation is here: https://github.com/Chysn/O_C-HemisphereSuite/wiki
digitalohm
chysn wrote:

Does this delay the onset of a gate signal, or delay a clock signal? If it delays the onset of a gate, does it similarly delay the end of that gate, or does the gate end immediately when the incoming gate goes low?


At the fast rising edge of the gate, the time counter begins and the gate isn't produced at the outputs until it's reached.

my initial thoughts were trigger inputs 1 & 2 are the incoming gate signals to be delayed, CV 1 & 2 are CV control over the time to delay and the CV 1 & 2 are outputs are the final gate signal.

app parameters thus far (in my head) are just the gate delay time (1ms min? 5s max) and then any parameteres relevant to the incoming CV that would control the time.

I think delaying the whole gate would be a good use of this, if one is just looking to kind of get a bit of an off timing, not sure when/where the other scenario might be useful, could cause some weirdness if the incoming gate ends before the delay duration?
chysn
The way I'm doing it right now, it's basically a one-bit sampler with a two-second sampling time.
behndy
have a uo_C headed my way tomorrow with Hemmies Suite loaded on it, BEYOND excited. want to play with the menu and interface a few days first, but super plan to grab a second and replace my Branches.

curious - would the scope be able to be modded to work as an audio tuner? or is there a plan to have a tuning applet?

thanks for making such an amazing version available!
Shledge
I've noticed clock forwarding doesn't seem to work on the redesigned brancher applet - visually it seems to work, but it doesn't output gates unless I put a gate into the digital input.

Also an idea for brancher - give it the option to switch between outputting gates, flip-flopped gates and triggers. Pressing the encoder could be used to switch between them.
chysn
behndy wrote:
curious - would the scope be able to be modded to work as an audio tuner? or is there a plan to have a tuning applet?


A dedicated tuner is coming. That’s been surprisingly challenging, though, math-wise.
chysn
Shledge wrote:
I've noticed clock forwarding doesn't seem to work on the redesigned brancher applet - visually it seems to work, but it doesn't output gates unless I put a gate into the digital input.


Edit: Please disregard the stuff below. I'm testing the fix for this now, and it will be available as a hex file in the 7/27/2018 release, and in the repo in maybe 15 minutes.

Is Brancher on the right or on the left?

If it’s on the right, it won’t respond to forwarded clock, because it looks for changes in gate state, and not clock pulses. The simple way to phrase it is that Brancher is not a “clocked” applet.

Edit: Now I’m thinking about how to have it be both clocked and gated. Interesting.

However, if it’s on the left side, it should pass incoming clock signals to a clocked applet on the right.

Quote:
Also an idea for brancher - give it the option to switch between outputting gates, flip-flopped gates and triggers. Pressing the encoder could be used to switch between them.


I really like the current button function, which is to force the signal to the opposite branch on a one-time basis. It’s easy to set up those other things (e.g., swap the cables for flip-flop), but the ad-hoc override is a cool performance control that’s otherwise hard to do.
behndy
niiiiiiice! don't know if the o_C hardware allows this, but do you know if it will be able to pass through the audio?

thanks!
chysn
behndy wrote:
niiiiiiice! don't know if the o_C hardware allows this, but do you know if it will be able to pass through the audio?


Yes. Whether you’d want to use that signal for anything is another question!
behndy
lol. Unintentional Signal Degradation Awesomeness?

i. am. LISTENING.
chysn
chysn wrote:
If it’s on the right, it won’t respond to forwarded clock, because it looks for changes in gate state, and not clock pulses. The simple way to phrase it is that Brancher is not a “clocked” applet.


I just took a look at this, and I was mistaken; Brancher is a clocked applet, and it should respond to a clock signal coming into Digital 1 when it's in the right hemisphere and clock forwarding is enabled.

EDIT: I've reproduced the issue you're describing. There will be a fix of some kind in Friday's 1.2 release.
JonathanBedrava
This is a pretty nifty idea. I look forward to checking it out this weekend.
qbassa@gmail.com
The idea for burst applet is to use it with sequencers. Sometimes it is good to have side motif / melody, that only lasts several notes and then it is quiet until next trigger / gate. I see it as applet with two digital inputs and one digital output. The first digital input is base clock. The second is trigger / gate input that launches the output clock, but only for limited number of pulses. Additional parameters could be changing main clock ratio - clock divider / multiplier or maybe even more advanced clock manipulation (shuffle? euclidean rhythms?). I know it is possible to achieve this with using few modules (logical AND on clock and gate signal) but having such applet would be really neat solution.

Best,
Jakub
chysn
I did a basic Burst applet this morning. It's based on the Ladik S-075, meaning that it has a Number control (1-12) and a Spacing control (8ms - 500ms). Number and Spacing can also be set via CV.

Output A/C is the burst, and Output B/D is a gate that goes high while the burst sequence is in progress, like the S-075.

I'll most likely make it clockable before release.

I do want to do a second, pattern-based, burst generator.
dooj88
chysn wrote:
I do want to do a second, pattern-based, burst generator.


euclidean burst generator? might be neat, though i rarely do beat-oriented patches.
qbassa@gmail.com
Great to hear there will be burst generator! It would be nice to have both options clockable and with spacing control, sometimes going off the beat could be very inspiring (especially with cv control that you have already done). Any chance to test this version you already made?

Cheers!
chysn
dooj88 wrote:
chysn wrote:
I do want to do a second, pattern-based, burst generator.


euclidean burst generator? might be neat, though i rarely do beat-oriented patches.


I don't know about a Euclidean burst generator. Hemisphere Suite already has a dual Euclidean sequencer, though, called Annular Fusion.
chysn
qbassa@gmail.com wrote:
Any chance to test this version you already


If you know how to, or would like to learn to, compile the code in Arduino, it's available on GitHub, here: https://github.com/Chysn/O_C-HemisphereSuite

There are instructions for doing everything in my first few blog entries, here: http://butmostlycrime.blogspot.com/

If you don't want to go through all that, the hex file will be released on Friday (July 27, 2018).
qbassa@gmail.com
I definitely have to setup this Ide and try compiling it, as I used to work as java developer some time ago;). Back to the functionality, it would be great to have base spacing value and only modify it by Cv in the relative way, not absolute, so providing 0v would return to default value.

Best,
Jakub
chysn
qbassa@gmail.com wrote:
Back to the functionality, it would be great to have base spacing value and only modify it by Cv in the relative way, not absolute, so providing 0v would return to default value.


Yeah, I think you're right.
DeltaTucanae
chysn wrote:
I did a basic Burst applet this morning. It's based on the Ladik S-075, meaning that it has a Number control (1-12) and a Spacing control (8ms - 500ms). Number and Spacing can also be set via CV.

Output A/C is the burst, and Output B/D is a gate that goes high while the burst sequence is in progress, like the S-075.

I'll most likely make it clockable before release.

I do want to do a second, pattern-based, burst generator.


Thanks for this, can't wait to try it.
bloc
I'm goign to buy a o_c specialy for this.Really exited for the speed of development you have for small ideas. I only have a couple of app ideas wich would be awesome.

selected lfo:
pot determines lfo speed and button determines lfo shape. mabe something like waveshape included in a way (peaks lfo mode)

Tap tempo lfo:
Peaks tap tempo lfo with wave type by button and shape by knob(maybe a feature in selected lfo app)

random div/mult:
bassicly marbles yellow mode where the pot determines the bias know of marbles. i dont know what the button will do maybe freezing the div/mult to a static value (marbles manual https://mutable-instruments.net/modules/marbles/manual/)

midi clock in/out":
in is like a tap tempo algoritme wich forwards a midi clock.
out will output the current clock signal in 2 different pulse type (high speed and 1/4)

2 decays:
to decay envelopes in one app. From here on it gets a bit vague for apps smile


Melodie generator:
bassicly a arp generator where the signal can be a chord or a singal that is inputted. control over arp direction and rangde.

2 channel attenuator:
attenuating the 2 channels channels. button for switching channel and knob button for switching to offset mode where the knob adds a value. i know the signal will be aliased much but i still think it would be something nice to have


the apps are ordered on awesomeness
chysn
Good ideas, thanks! I haven't had much luck with MIDI clock with the Teensy USB MIDI. I might be able to get it to work some day, but it's documented poorly. I'll keep trying!

I'd definitely like to increase the LFO selection, that's a priority.
zerodivide
very curious about this but I use meta-q the most hmm
Shledge
There are quantiser applets which do a similar job.

Chysn - is it possible to make a simplified bytebeat applet? Mainly to playback/change speed of the bytebeats algos that come with o_C's firmware
chysn
Shledge wrote:
There are quantiser applets which do a similar job.

Chysn - is it possible to make a simplified bytebeat applet? Mainly to playback/change speed of the bytebeats algos that come with o_C's firmware


I’ve honestly never even once started Bytebeats up. I’ll take a look at it.
rayultine
you've never used the original firmware bytebeats!!!?! It's wild, dude!
Shledge
chysn wrote:
Shledge wrote:
There are quantiser applets which do a similar job.

Chysn - is it possible to make a simplified bytebeat applet? Mainly to playback/change speed of the bytebeats algos that come with o_C's firmware


I’ve honestly never even once started Bytebeats up. I’ll take a look at it.


Good way to get controlled-yet-random CV out. Nice concept to make music out of short lines of code too - I've edited some on my o_C Rockin' Banana!

chysn
rayultine wrote:
you've never used the original firmware bytebeats!!!?! It's wild, dude!


I've spent basically no time with the original firmware, other than to see how it works. I bought O_C specifically to write my own stuff, and started working on that on day one.

But it's been fairly easy to port a lot of O_C functionality, like the Braids quantizer, the Bjorklund tables for Euclidian rhythms, and the Lorentz stuff from Streams. So a port of Bytebeats isn't out of the question.
shellfritsch
i'm wondering if there is an easy way to have the option to use unipolar modulation sources where hemispheres is expecting a bipolar source?

i was thinking a top level preference setting.
chysn
shellfritsch wrote:
i'm wondering if there is an easy way to have the option to use unipolar modulation sources where hemispheres is expecting a bipolar source?

i was thinking a top level preference setting.


I think the easiest way is negative offset. Shades is a ridiculously good companion for Ornament and Crime.
chysn
Hemisphere Suite 1.2 is now available

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.2

I'm pretty excited about some of the updates here, and I think the basic UI has pretty much gelled. The help system is now activated by double-clicking the selection buttons; clock forwarding has been moved to the left encoder, so that selection buttons are used only for applet management; and there's now a system to filter applets by category, which is slightly menu-divy, but it's there for convenience.

A few applets have been added, mostly consisting of requests at this thread. Carpeggio is a blast. Somebody suggested a Cartesian sequencer a few pages back. I started working on that, and it morphed into a Cartesian arpeggiator than can ALSO be used as a Cartesian sequencer.
Mashmore
Fucking love the double tap for the help menu applause
srogers
Such a good idea, thank you for all the work you put into this!

Maybe I missed something earlier in the thread but I'm only getting 13 applets when I install Hemisphere, it seems that the gated VCA isn't showing up?
chysn
srogers wrote:
Maybe I missed something earlier in the thread but I'm only getting 13 applets when I install Hemisphere, it seems that the gated VCA isn't showing up?


The product you have is the older one, which includes the O_C apps (except for Meta-Q).

The newer one is Hemisphere Suite, which removes the Ornament and Crime apps, and has (as of today) 30 applets. You can get the hex file for that here:

https://github.com/Chysn/O_C-HemisphereSuite/releases
srogers
chysn wrote:
srogers wrote:
Maybe I missed something earlier in the thread but I'm only getting 13 applets when I install Hemisphere, it seems that the gated VCA isn't showing up?


The product you have is the older one, which includes the O_C apps (except for Meta-Q).

The newer one is Hemisphere Suite, which removes the Ornament and Crime apps, and has (as of today) 30 applets. You can get the hex file for that here:

https://github.com/Chysn/O_C-HemisphereSuite/releases


Yeah that much I knew, I wanted Hemisphere in lieu of the Suite in order to keep the original apps intact. Reading over the Hemisphere release notes/revs I see now what has been added/removed so far!
chysn
srogers wrote:
I wanted Hemisphere in lieu of the Suite in order to keep the original apps intact. Reading over the Hemisphere release notes/revs I see now what has been added/removed so far!


Gotcha. Officially, the end of that line is Beta 5, which is available here:

https://github.com/Chysn/O_C-Hemisphere/releases/tag/hemisphere.beta5

It's stable and fairly bug-free, but I'm not going to do anything else with it. My original plan was to have a system that would work within the O_C firmware. But it turned out that there wasn't any room left to do what I wanted to do, so I've thrown the original plan out the window. My new paradigm is "replacement."
Shledge
I'm honestly tempted to sell some of my modules because of these applets - even some of my distings. They work so well and are not laden under loads of menus/questionable UX decisions.
BaloErets
1.2 is a great release. I've been having great fun with Carpeggio and the gate delay. Might be time for a 3rd O_c!

A suggestion for either a modification, or an alternate version of Carpeggio would be using the 2 digital clocks for sequence advancement. As in;

gate/trig into digital 1 moves the sequence 1 step on the x axis
gate/trig into digital 2 moves the sequence 1 step on the y axis.
1 CV out provides the sequence CV and the 2nd with a gate output when either axis moves. This is great for making evolving, yet familiar sequences!
Shledge
Having fun with the CV recorder. Manually send a gate, use an offset module and you can make it into a 64 step sequencer. Use a sequential switch and it's a 128 step sequencer.
Eurtrude
This. is. pure. genious.

I didn't have time to thank you for the new Brancher, so I do. This is awesome smile

I'm loving more & more my O_c !
chysn
Shledge wrote:
Having fun with the CV recorder. Manually send a gate, use an offset module and you can make it into a 64 step sequencer.


That's why I was saying that Shades (or any offset/polarizer/attenuverter module) is a ridiculously good companion to O_C. It basically provides the knobs that O_C doesn't have! Even better: Tetrapad in Sliders Mode.

Eurtrude wrote:
I didn't have time to thank you for the new Brancher, so I do.


You're quite welcome.
Shledge
chysn wrote:

That's why I was saying that Shades (or any offset/polarizer/attenuverter module) is a ridiculously good companion to O_C. It basically provides the knobs that O_C doesn't have! Even better: Tetrapad in Sliders Mode.


Shades is so good that I have two of them. One of the few mixers out there that is fully buffered for pitch CV too.
itsmeaflo
Great work! Is there also a 2 channel ASR applet?
chysn
itsmeaflo wrote:
Great work! Is there also a 2 channel ASR applet?


No, there's not. Do you think there's much value to a two-output ASR?
shellfritsch
You mentioned a full midi applet a while back. Is there a chance we will be seeing a 4 channel cv/gate -> midi function?
Thanks again for all your amazing work!
Shledge
I don't think that's possible, considering there are only 4 outputs. Two CV/gates at most.

EDIT: Nevermind, you meant CV to MIDI.
chysn
shellfritsch wrote:
You mentioned a full midi applet a while back. Is there a chance we will be seeing a 4 channel cv/gate -> midi function?
Thanks again for all your amazing work!


Funny you mention it! I'm working on the full MIDI Interface app this week. Unlike my other stuff, MIDI Interface is going to be an app in the "classic" O_C style, with the native menus and user experience.

Each of the Gate/CV pairs will be assignable to different controllers and on different channels. So if you want to use all four Digital/CV pairs for note information, you'll be able to do that. As Shledge pointed out, you can only do two note/gate channels of MIDI-to-CV, but you'll be able to make those assignments as well.

The MIDI Interface will have two independent Setups that can be saved. So you can create an all-note Setup, and also have a Setup with pitch bend, or whatever.

Over the weekend, I was working on system exclusive dumps. So for every app I do, it will send a sysex dump when you return to the main menu. And it will listen for sysex dumps as the app is running, so you can set the state of any of my apps (including Hemisphere) via MIDI.

The O_C hardware makes for a great MIDI hub. I've tried CV-to-MIDI with Q-Nexus, MicroBrute, SQ-1, and some other stuff. There was always some problem, like noise, or configuration issues. Now I've finally got a modular MIDI interface that I like. I can't promise anyone else will like it, but it ticks most of the major boxes.

Have any of you tried the MIDI In and Out in Hemisphere? Since you're hooking your O_C up to a computer, anyway, it's worth trying those out. The Legato mode for MIDI Out is very playable.
shellfritsch
yesssss


this is great news!

one of my all time favorite sequencers, kria, only exists as a euro module and i've often wished i could use it to control software. i'll finally be able to!
behndy
got mines in, love them. first time i've felt attached to and comfy with an o_C.

just checking, because i can be DUMB about things like this. there is the Hemispheres Suite then a version that keeps (almost all of) the original o_C apps? or does Suite also keep the other apps at root?

like, if i hold right knob down to back out of Hemispheres, the screen where i would hold right knob again to save, there are all the Copier_Maschine etc that can be selected, then Hemisphere at the end.
chysn
behndy wrote:
there is the Hemispheres Suite then a version that keeps (almost all of) the original o_C apps? or does Suite also keep the other apps at root?

like, if i hold right knob down to back out of Hemispheres, the screen where i would hold right knob again to save, there are all the Copier_Maschine etc that can be selected, then Hemisphere at the end.


Hemisphere Suite takes all the native apps out, except for References, and I'm probably going to port over References so it, too, can be removed.

You can tell which version you have in a few ways, but the easiest is that the bootup splash screen says "Hemisphere Suite" instead of "Ornaments and Crimes" (sic??) at the top. If it says this, you'll have Hemisphere with 30 applets, The Darkest Timeline sequencer, and some placeholders for stuff I'm working on.

In what is now a month of retrospect, it was a mistake to make two versions. So I might start pretending that the other one doesn't exist. smile
behndy
lol. deny deny DENY.

yeah. that does complicate things a bit. i mean, again, i'm an idiot about some things, so it doesn't have to be TOO complicated to get my brain bits all wubbly.

cooooooool. i'll try to load up the Suite tonight.
behndy
also, i'm sort of a git about loading firmwares, is there a walk through on loading up Hemisphere Suite?

edit - found this - http://ornament-and-cri.me/firmware/#method_a

hurpadurp. looks simple.

do i have to calibrate again?
behndy
hurm. so, got the hex file from here -

https://github.com/Chysn/O_C-HemisphereSuite/releases

used the Teensy.exe Windows app i've used to update my FLXS1, loads the hex file fine. when i hit Program it says PROGRAMMING, progress bar ends. then when i hit REBOOT the uo_C reboots, and it goes to a screen that says......

Hemisphere Suite
-------------------------

[L] => Calibration
[R] => Select app

v.1.2 (OC1.3.4)


and a progress bar is going down. if i let the bar reach end, it freezes where i can't select anything. if i reboot and hold down R knob until the progress bar hits zero, then it goes into the app selection screen where i have.....

Hemisphere
Darkest Timeline
MIDI Interface
Pong
Neural Network
References

it lets me scroll around, and i can select and enter any of those apps, but if i click on Hemisphere it freezes and none of the buttons do anything.

tried the HemisphereSuite1_2.hex and the HemisphereSuite1_2_FLIP180.hex files. one of my o_C's is skiff orientation, so thanks so much for already having a Flip file ret to go!

what am i doing wrong?

thanks!
Shledge
You'll need to make sure the aurdino IDE, the compile settings and teensyduino match what is mentioned in the wiki.
behndy
gotchaaaaaaa.

i mean. i understand what you're SAYING, but that's a bit beyond my codey skills right now. will research tomorrow.

thanks!
behndy
if i'm using the Hex from here - https://github.com/Chysn/O_C-HemisphereSuite/releases

then the Ardudino IDE and Compile Settings would be correct now? i'm not grabbing a code file and compiling it myself, just using Teensy.exe on my PC to program the Hex file from the Hemisphere site.

edit - never mind this - - - > i'm not sure what the Teensyduino is?

looked it up.
behndy
also, i didn't build these uo_C's myself, they came with the Hemisphere (non-suite) version running correctly, so the Teensyduino would be installed on the Teensy? the non-suite Hemisphere needs that to run properly no?
Zymos
Shledge wrote:
You'll need to make sure the aurdino IDE, the compile settings and teensyduino match what is mentioned in the wiki.


To load an already compiled hex file??
behndy
yeah, i'm using the official release Hex's from here -

https://github.com/Chysn/O_C-HemisphereSuite/releases

for what it's worth, i just loaded the 1.1 file (WHICH I SHOULD HAVE THOUGHT TO DO TO TEST THE FIRST TIME DURP), and it worked perfectly. shows the Calibration option menu, when the progress bar hits zero it jumps to Hemispheres Suite. lets me back out and see Darkest Timeline, References etc.

1.1 seems to load fine, 1.2 hangs after loading.

i tried installing 1.1 and then programming in 1.2 in case there was supposed to be a transition between the two or something, 1.2 still freezes.

moop?
chysn
You might need to reset the O_C's EEPROM. Hold down the Up and Down buttons while starting O_C and keep them held until the O_C is done booting. You'll then be prompted to press the right encoder to erase the EEPROM.

I suspected that this might need to be done in some cases, because Hemisphere Suite 1.2 uses a different format for saving settings data.
behndy
awesome! will try that when i get home. will i need to do a re-calibration? lol. never have, so that is a bit DAUNTING.

thanks Chysn! this is such a BEAUTIFUL module to use now.
sushiluv
I love the Hemisphere Suite, some amazing stuff in there, thinking about getting a second o+c, thanks!

Do you pan on imlementing midi out for all sequence related applets like the turing machine and so on as well? would be great!

cheers!
chysn
behndy wrote:
awesome! will try that when i get home. will i need to do a re-calibration? lol. never have, so that is a bit DAUNTING.


I just tried this out, and calibration settings seem to survive erasing the EEPROM.

sushiluv wrote:
Do you pan on imlementing midi out for all sequence related applets like the turing machine and so on as well? would be great!


I have considered it, and haven't ruled it out, but there are a few things that give me pause. One is, there aren't always enough controls or screen space available for setting the MIDI out channel.

Also, things like sequencers and Turing Machines are really common in DAWs already, so I think it would be of limited utility. By their nature, Hemisphere sequencers will never be full-powered things.

Further, you can combine Turing Machine with MIDI Out. In fact, in my YouTube demo of the MIDI Out applet, I do exactly this (https://www.youtube.com/edit?o=U&video_id=cVnJ3RqdbJU).

That said, expect to see MIDI become a bigger part of this system in the near future.
behndy
chysn wrote:
You might need to reset the O_C's EEPROM. Hold down the Up and Down buttons while starting O_C and keep them held until the O_C is done booting. You'll then be prompted to press the right encoder to erase the EEPROM.

I suspected that this might need to be done in some cases, because Hemisphere Suite 1.2 uses a different format for saving settings data.


YEP. little bugger is GOLDEN now.

now to load it onto the other uo_C.

thanks!
Timmy
chysn wrote:
behndy wrote:
awesome! will try that when i get home. will i need to do a re-calibration? lol. never have, so that is a bit DAUNTING.


I just tried this out, and calibration settings seem to survive erasing the EEPROM.


Um, no.

The calibration data are stored in the EEPROM. If you erase it, your calibration data are gone for good, unless you noted down the values with pencil and paper (which you can indeed do, takes just a few minutes).

Now, the O&C firmware stores the calibration data in a different area of EEPROM from the apps setting data, and you can erase all the apps settings without erasing the calibration data. No idea how Hemisphere behaves in this respect, but if it erases the entire EEPROM, then the calibration data is definitely gone.
REVIVER
Timmy wrote:
Um, no.

The calibration data are stored in the EEPROM. If you erase it, your calibration data are gone for good, unless you noted down the values with pencil and paper (which you can indeed do, takes just a few minutes).


d'oh!

Whelp....

EDIT: Curious then...if the calibration data has been erased, should all of the calibration settings default back to a "zero" state?
Timmy
REVIVER wrote:
Timmy wrote:
Um, no.

The calibration data are stored in the EEPROM. If you erase it, your calibration data are gone for good, unless you noted down the values with pencil and paper (which you can indeed do, takes just a few minutes).


d'oh!

Whelp....


Yeah, I think there’s a general principle when writing and releasing alternative firmware of "primum non nocere" to borrow the medical dictum.

REVIVER wrote:

EDIT: Curious then...if the calibration data has been erased, should all of the calibration settings default back to a "zero" state?


No, the default calibration settings are zero, they are instead reasonable starting points for the required values if the module has been built with the specified precision components. They will still need tweaking for every module, though.
REVIVER
Timmy wrote:
No, the default calibration settings are zero, they are instead reasonable starting points for the required values if the module has been built with the specified precision components. They will still need tweaking for every module, though.


Got it. Thanks.

seriously, i just don't get it
pld
Timmy wrote:
chysn wrote:
behndy wrote:
awesome! will try that when i get home. will i need to do a re-calibration? lol. never have, so that is a bit DAUNTING.


I just tried this out, and calibration settings seem to survive erasing the EEPROM.


Um, no.

The calibration data are stored in the EEPROM. If you erase it, your calibration data are gone for good, unless you noted down the values with pencil and paper (which you can indeed do, takes just a few minutes).

There's maybe two cases to distinguish:
If you erase the EEPROM though some external tool (if that's even supported in the Teensy loader) it would lose the calibration.
Using the o_C reset function at startup _should_ only erase the apps' settings and global data (sequences and scales), but leave calibration data intact (erasing starts at EEPROM_GLOBALSETTINGS_START). The only way to nuke calibration should be to start calibration, use defaults, then save. Should smile And even if somehow the calibration data is lost, it should fall back on those (reasonably sane) default values.

As an aside, it's possible to make app data incompatible with previous releases and avoid some issues that way.
chysn
pld wrote:
Using the o_C reset function at startup _should_ only erase the apps' settings and global data (sequences and scales), but leave calibration data intact


Yes, this is what I was talking about when I was advising behndy. I tried it several times because I know it takes like a half hour to calibrate an O_C with a meter, and I didn't want to steer anyone wrong. If you follow the reset procedure onboard the module, it definitely does not affect calibration settings.

Thanks for the clarification!

Timmy wrote:
No idea how Hemisphere behaves in this respect, but if it erases the entire EEPROM, then the calibration data is definitely gone.


Hemisphere does nothing in this respect. It's just an app that runs on the normal O_C 1.3.4 framework. I'm just talking about the [R] RESET SETTINGS AND ERASE EEPROM procedure that you guys provided.
chysn
chysn wrote:
you can only do two note/gate channels of MIDI-to-CV, but you'll be able to make those assignments as well.


It turns out, you will be able to do four notes of MIDI-to-CV with MIDI Interface.

The way it works is, each output (A, B, C, D) can be assigned to a MIDI event independently, and on any MIDI channel. So you can assign all four outputs to Note and control pitch on up to four oscillators. If you set it this way, you just need your VCA control to come from somewhere else. Or you can assign three outputs to Note and a single Gate... or the traditional two Notes/two Gates.

I'll probably post an intro video this weekend. This is still on target for August 10 release.
Maximars
This is awesome! Thank you very much!
An idea: A simple AD envelope with End of Cycle (EOC) trigger out = Happy Krell Patch Rockin' Banana!
CV vor A and D

Thank you!
chysn
Maximars wrote:
An idea: A simple AD envelope with End of Cycle (EOC) trigger out = Happy Krell Patch Rockin' Banana!
CV vor A and D


Good idea. It's really only a few lines of code away from the existing Slew applet, so it should be pretty easy.
selfdestroyer
Installed 1.2 on a spare O&C I had and I wanted to say thank you for releasing this! This has been a blast to use and its very intuitive. I look forward to watching this project grow.

Thanks again!
Timmy
chysn wrote:


Timmy wrote:
No idea how Hemisphere behaves in this respect, but if it erases the entire EEPROM, then the calibration data is definitely gone.


Hemisphere does nothing in this respect. It's just an app that runs on the normal O_C 1.3.4 framework. I'm just talking about the [R] RESET SETTINGS AND ERASE EEPROM procedure that you guys provided.


Ok, no worries then. When you wrote “hold down the R encoder to erase the EEPROM” it set alarm bells ringing...
mazzyboy
Hello!

Im not sure if you are taking request for apps, but here goes...

Would it be possible to include a clock shuffler app? Something in the style of the ladik S-188?

(I know that you can kind of do this with the gate delay app, but you still need to use external modules. I'm looking for something more self-contained)
chysn
mazzyboy wrote:
Hello!

Im not sure if you are taking request for apps, but here goes...

Would it be possible to include a clock shuffler app? Something in the style of the ladik S-188?

(I know that you can kind of do this with the gate delay app, but you still need to use external modules. I'm looking for something more self-contained)


Yeah, I'm totally interested in hearing what folks want. And I love working from Ladik's stuff as templates for functionality because so often they map really well to Hemisphere controls.

There won't be too many new applets for the 1.3 release because I'm spending all my time on Captain MIDI, which is a full app. But I was able to finish a CV mixer and the AD envelope.
mazzyboy
Thats great news! I look forward to seeing how things develop!
qbassa@gmail.com
Hi again,

Didn't have time to try new features from suite 1.2 until now. I have a question on Bursts. What clock format is used for input and output? When connecting output 1 to external sequencer (Xaoc Moskwa or Antumbra ROT8) sequencer doesn't progress the set number of steps. I see burst is generated on the screen but sequencer seems not to catch the output tics (or catch only 1 or two from 10). Another question concerns input clock. When tried regular gate signals with slow tempo on digital 1 the spacing was set fairly fast like 20ms. I tried tap tempo clock from 0-coast around 1 second and the spacing was set to 100ms. Maybe there is a bug in the internal time calculator or I am doing something wrong.

Best,
Jakub
chysn
Hey, Jakub,

Burst sends a 2ms trigger, which might not be long enough. I'll put it at the normal (I think, 5ms) trigger and see if it works. I've actually got a Moskwa around that I can test with, I just need to put it in my case.

Update: Yeah, 2ms just isn't long enough. When I just leave it at the default, it works a lot better. This will be fixed in the 1.3 release (this coming Friday). If you compile yourself, the fix is available now.

I'll review the calculation for the display, to make sure that's right. Note that the displayed Spacing value depends on the number of bursts, so if you've got a 60BPM clock and 4 bursts, it should read (around) 250ms, and then if you change the bursts to 8, it would go to 125ms. The Spacing display, in other words, should be the amount of time (in ms) from the start of one trigger output to the start of the next.

Thanks for helping me make this better!
qbassa@gmail.com
Thanks for fast checking and explaining me spacing value. So maybe third parameter with divide / multiply values could be introduced here. /8 /4 /2 1 *2 *4 *8?


Best!
Z_Hyper
I tried MIDI Out.
I was able to control NL 3 with Din MIDI using oplab as USB Host.
I am impressed! !

https://instagram.com/p/BmGRp_fDYe2/
adh82
I'd really love to see a cv controlled precision adder like the beasts chalkboard.
And I'd really love to see some slow evolving random apps with options to go really slow, something like the sloth random module or the low rents app on o&c.
chysn
Z_Hyper wrote:
I tried MIDI Out.
I was able to control NL 3 with Din MIDI using oplab as USB Host.
I am impressed! !

https://instagram.com/p/BmGRp_fDYe2/


Nice! If you’re impressed by MIDI Out, wait until you try Captain MIDI.
chysn
adh82 wrote:
I'd really love to see a cv controlled precision adder like the beasts chalkboard.
And I'd really love to see some slow evolving random apps with options to go really slow, something like the sloth random module or the low rents app on o&c.


Precision anything isn’t really an option with O_C ADCs. But I’ll look into how Sloth works. I’d like to do something like that.

Have you tried LowerRenz?
qbassa@gmail.com
Once again about Bursts. A clock divider for achieving longer bursts than fraction of the main clock would be really handy. Also odd divisions 3, 5, 6 would be musically sensible.

Best
chysn
qbassa@gmail.com wrote:
Once again about Bursts. A clock divider for achieving longer bursts than fraction of the main clock would be really handy. Also odd divisions 3, 5, 6 would be musically sensible.

Best


Oh, yeah, this is done already for the 8/10 release.
Maximars
Hi chysn

Thank you for adding the AD Envelope!

Another idea: Slope Dedector App

Option to select one of 4 states on both outputs:

CV is rising, Gate
CV is falling, Gate
CV is steady, Gate
CV changes direction, Trigger

Maybe passing CV1 to the second seide of Hemisphere

Somehow like:

Ladik, J-110 Derivator
or

Elby Desings, Slope Detector

I do not know if the sensitivity controller is necessary with a digital module.

Eny chance for this app?

Tank You!
chysn
Maximars wrote:
Hi chysn
Another idea: Slope Dedector App


Yeah, I like that one! I'm pretty sure a sensitivity control will be required, since there's some jitter from one reading to the next.

I still want to do the shuffle applet, but it's been challenging to learn exactly how such a thing works. Any ideas?
behndy
played a thing last night and whilst nerding out about gear the 4 or 5 people that asked about Hemispheres and jiggled around with mines were blown away.

because.

AWESOME.
Maximars
For a shuffled clock every odd time pulse has to be delayed a bit.
Steevio explains it pretty well in this video.
https://www.youtube.com/watch?v=x6hJa2lRRgM @ 1:35

Every odd pulse is a delayed pulse from a straight one. I do not know how the Ladik module works exactly.
If you drive sequencer with a shuffled clock you have to deal with reset signals somehow, therefore a dedicated module is to prefer. anyway, everybody needs "pamelas new workout" love

Hope this helps somehow ...
chysn
Maximars wrote:
For a shuffled clock every odd time pulse has to be delayed a bit.


So the first pulse is delayed, the second one is on the clock, the third one is delayed...
Maximars
No, vice versa

The first is on the Clock and the Offbeat is delayed. Roughly said. That creates this swing feeling. Sometimes it almost sounds like a triplets. It is somehow a wonky clock meh

1....+..2....+...3....+...4....+
|.....|...|.....|...|.....|...|.....|...

May search some vids for Swing and Shuffle.

Hope my english is not to bad..
chysn
Maximars wrote:
No, vice versa

The first is on the Clock and the Offbeat is delayed. Roughly said. That creates this swing feeling. Sometimes it almost sounds like a triplets. It is somehow a wonky clock meh

1....+..2....+...3....+...4....+
|.....|...|.....|...|.....|...|.....|...

May search some vids for Swing and Shuffle.

Hope my english is not to bad..


Thanks! I just made it so that you can set a different delay for both even and odd beats. The delay is expressed as a percentage of the clock time from one beat to the next.
Maximars
I’m excited, looking forward to the next release! nanners
aragorn23
The shuffle applet is probably going to be the clincher for me building a second o_c SlayerBadger!
Maximars
Luckily I already have two o_c and I still have four displays on my workbench SlayerBadger!
Maximars
Because I'm impatient, I tried to compile the latest source code. I can compile the code with the usb midi enabled. I can upload the hex file to the Teensy but the display remains dark hmmm.....
Shledge
You'll need to make sure it's compiled and uploaded with the proper settings and versions of arduino IDE/teensyduino. Usually it directly uploads to the teensy from the arduino IDE - no need to make a hex file.
aragorn23
Random suggestion that may well be out of scope, but would it be at all possible to port Grids across as an app or create something that uses the same collection of patterns spaced on a 2d grid in order of fill density, with the ability to navigate through them?
chysn
Maximars wrote:
Because I'm impatient, I tried to compile the latest source code. I can compile the code with the usb midi enabled. I can upload the hex file to the Teensy but the display remains dark hmmm.....


Usually the display staying dark means that your Teensy has been erased but not programmed. The solution is "try it again." Ensure that the Teensy Loader shows the "Programming" progress bar, and then says "Reboot OK."

aragorn23 wrote:
Random suggestion that may well be out of scope, but would it be at all possible to port Grids across as an app or create something that uses the same collection of patterns spaced on a 2d grid in order of fill density, with the ability to navigate through them?


I don't see any technical barriers to a Grids-like app, but you'd have to sacrifice some of the outputs. I'd probably use three outputs for the drums, and make the fourth output assignable to one fill or one accent.
piersr
Maximars wrote:
Because I'm impatient, I tried to compile the latest source code. I can compile the code with the usb midi enabled. I can upload the hex file to the Teensy but the display remains dark hmmm.....


I had the same problem when using the latest Arduino and Teensyduino libs, check the O_c build docs to make sure you're using the reccomended versions of each.
chysn
Here's a preview of one of the features in tomorrow's Hemisphere Suite 1.3 release. The sysex dump is demonstrated with The Darkest Timeline, but the same workflow applies to all current and future Hemisphere Suite apps. Except Pong. Pong doesn't have a care in the world.

https://youtu.be/6R99A0Yvqlk

The motivation behind this is to provide a way for me to make bigger systems that use more data. With the classic O_C firmware, you can save one thing, and you can lose it in lots of different ways. With sysex support, you can spend a long time on something and bring it back in the studio or during a performance. Maybe it's not the Modular Way, but I think some will appreciate it.
chysn
https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.3A
selfdestroyer
chysn wrote:
https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.3A


Thanks! Can't wait to get home from work and give it a go. Was having a blast with 1.2 last night.
4eyez
Does Hemispheres offer support for microtonal tunings in the quantizers? It seems like the standard O&C does, but I'm not seeing if that flows through to Hemisphere too.
chysn
4eyez wrote:
Does Hemispheres offer support for microtonal tunings in the quantizers? It seems like the standard O&C does, but I'm not seeing if that flows through to Hemisphere too.


Hemisphere currently has two quantizer applets available.

Dual Quantizer gets its scales from the same source that Quantermain uses, and user scales can be hacked in exactly the same way. Likewise, you can compile Hemisphere Suite with the Buchla support to use the built-in non-octaval scales, although I have not tried this.

The other quantizer, Scale Duet, focuses on creating "masks" of standard pitches, kind of like µScale, and thus does not support microtonal scales.

I'd be interested in hearing more about the needs of microtonal musicians for a potential future applet focusing specifically in this area.
Dogma
OP this is amazing! I sold my O_C immediately as i didnt like the menu setup nor the choice of apps - this is perfect....thank you!!
4eyez
chysn wrote:
Dual Quantizer gets its scales from the same source that Quantermain uses, and user scales can be hacked in exactly the same way.


Very cool, thanks for the info!
chysn
4eyez wrote:
chysn wrote:
Dual Quantizer gets its scales from the same source that Quantermain uses, and user scales can be hacked in exactly the same way.


Very cool, thanks for the info!


You know, I wasn't totally decided on my big project for 1.4, and you've inspired me. I now know that I'm doing a standalone scale editor app. I like the idea of importing an existing scale, and being able to edit it. Or being able to edit a note one-handed, without holding down a button while turning a knob. Or loading up scales via MIDI dump.

Thanks!
noisanceritual
wow i just noticed this, really excited to install it, does it lose calibration?
Mashmore
noisanceritual wrote:
wow i just noticed this, really excited to install it, does it lose calibration?


Nope. you should be good to go! Rockin' Banana!
AnalogAssailant
Quote:
I'd be interested in hearing more about the needs of microtonal musicians for a potential future applet focusing specifically in this area.



Yes! Not sure what you have in mind, but for me an easy way to enter in microtonal values (and save) would be awesome rather than having to edit and compile first and reflash.
Abyssinianloop
This is why I built two o_Cs. There is already so much use for two with the primary firmware. But I hoped robust alternate firmwares like this would come along and make these things even more powerful. Kudos!
blackjam
every time i leave this forum for a while I come back to great new things!
4eyez
chysn wrote:
You know, I wasn't totally decided on my big project for 1.4, and you've inspired me. I now know that I'm doing a standalone scale editor app.


And you in turn have inspired me to get an uO_C! I have an 8hp hole I want to fill in my small case with some kind of multi-function module and was debating between two Distings vs. an O&C and I think this tips the scales. Exciting stuff!
chysn
4eyez wrote:
chysn wrote:
You know, I wasn't totally decided on my big project for 1.4, and you've inspired me. I now know that I'm doing a standalone scale editor app.


And you in turn have inspired me to get an uO_C! I have an 8hp hole I want to fill in my small case with some kind of multi-function module and was debating between two Distings vs. an O&C and I think this tips the scales. Exciting stuff!


Cool! The 1.4 release is scheduled for August 24 and that will have the on-board microtonal scale editor. That will be the last bi-weekly update before I slow down to a monthly release for the remainder of 2018.
4eyez
Wow, that's quick, even cranking them out on a monthly basis is impressive! Thanks in advance for your work!
Attackman
Spotted this browsing through forum topics. Awesome. I love the look of Hemisphere - & the pace of development is amazing.

Given the overall complexity and rich menu's of o_C I'd like to give Hemisphere a whirl. It looks more immediate, less recourse to the manual and the dual set up might prove to be more useful overall.

I will really miss Quantermaine with its Turing Machines coupled to Quantizers. I put some effort into getting that app under my belt. But I'll forego it for a while and see how I get on with Hemisphere.

The level of generosity among developers and hackers is really something. Thanks. Can't wait to get stuck in to this new suite of apps!
chysn
Attackman wrote:
I will really miss Quantermaine with its Turing Machines coupled to Quantizers. I put some effort into getting that app under my belt. But I'll forego it for a while and see how I get on with Hemisphere.


Whether you keep it or not, I appreciate you trying it out. In general, similar functions are less comprehensive but (hopefully) more immediate.

Hemisphere's Turing Machine has one quantized output and one unquantized output. Since the quantized output uses 5 bits of the register, and the unquantized output uses 8 bits, there's not a strong relationship between the two outputs as they shift; it's like having two different Turing Machines. And, of course, you can run two at once. For the next update, the Turing Machine will allow scale selection, to take advantage of this:

For those of you who compile from source, the Scale Editor is available. I think there are some advantages of a standalone editor over one that's embedded in another app. Specifically:

* You can use any of the Braids scales or User scales as a starting point to edit your own scales
* There's a level of Undo, if you need to get a note value back where it was
* Like everything else, Scale Editor supports transfer via MIDI system exclusive, one scale at a time
* There are two types of monitoring: full scale monitoring (which quantizes CV input) and selected note monitoring (which just continuously sends whichever note you're editing at the moment, with adjustable octave).
* Monitoring is in real time for both of these, with no need to re-trigger the note

Full documentation is here: https://github.com/Chysn/O_C-HemisphereSuite/wiki/Scale-Editor
Attackman
chysn wrote:
Attackman wrote:
I will really miss Quantermaine with its Turing Machines coupled to Quantizers. I put some effort into getting that app under my belt. But I'll forego it for a while and see how I get on with Hemisphere.


Whether you keep it or not, I appreciate you trying it out. In general, similar functions are less comprehensive but (hopefully) more immediate.


I can't wait to try it out.

Hoping to install this evening or tomorrow morning. Quite excited! if it goes well I imagine a 2nd o_C entering my rack. Tho whether that'll be for the o_C I'm losing or for a 2nd Hemisphere may be a tough call. What you have achieved with Hemisphere is a solution to the drawback of o_C ie. lack of 'playability' in realtime. That for me was probably why I settled on Quantermaine - apart from some initial scale selection I could pretty much let it run and use triggers and cv to play it. Not that i don't love the o_C, you find ways to work with the limitations of things but it was always a niggle that it was difficult to get hands on in realtime. Especially when I had to crack out the reading glasses! Lol.
4eyez
chysn wrote:
For those of you who compile from source, the Scale Editor is available.


Hot damn, that was fast. we're not worthy

Now if I could only get my hands on the actual module that fast....
Attackman
chysn wrote:
Attackman wrote:
I will really miss Quantermaine with its Turing Machines coupled to Quantizers. I put some effort into getting that app under my belt. But I'll forego it for a while and see how I get on with Hemisphere.


For those of you who compile from source, the Scale Editor is available.


Yeah this looks ace. If I've understood it, scales can be prepared on a computer and dumped via sysex into Hemisphere? Note monitoring is a clever touch. This is such an amazing project!
chysn
Attackman wrote:
If I've understood it, scales can be prepared on a computer and dumped via sysex into Hemisphere?


The intent is storage; you can create scales on the module, then dump them to the computer to save them. So instead of just having the four user scales that the original firmware allows, you can create and save as many as you want.

However, my SysEx format is really simple, and it wouldn't be too hard for some ambitious developer to write a computer-based scale editor. I'm pretty sure that you'll find it easy to edit scales with the app.
rayultine
chysn: in the main o_C thread, I shared a file with most of the Braids tunings in MTS format. These can be easily produced via Scala.

Here's a workflow:
http://xen-arts.net/microtuning-virtual-instruments-part-5/
chysn
rayultine wrote:
chysn: in the main o_C thread, I shared a file with most of the Braids tunings in MTS format. These can be easily produced via Scala.

Here's a workflow:
http://xen-arts.net/microtuning-virtual-instruments-part-5/


Nicely done!

I'm tempted to keep going with the Scale Editor and add MTS support. Unfortunately, the 1.3.5 Teensy USB MIDI Library can only receive up to 60 bytes (yeah, you read that right) from system exclusive files. I get around this in my code by splitting data into multiple dumps and adding a page number to sort them out during reception. But I can't think of a way to successfully receive an MTS dump.

But, I can certainly create an MTS dump. You know, for all the people who want to edit scales on a module and then use them in a DAW smile
Attackman
Wow Chysn. I thought this looked great but FLIPPIN ECK.

I installed it 90 minutes ago & I've never had so much fun with an o_C module. It's a complete game changer. I love it! This is a bit gushing but I'm really impressed - it's opened up new possibilities for me and it's kind of an effortless experience.

I've got this thing going now with Scale Duet and Trending, receiving from Batumi and Maths and playing with an STO. Just ACE.

Can't wait to see how this project develops. A massive thanks for developing and sharing this. I know what I'm doing for the rest of the week now!
behndy
also crazy happy with all that Hemmies opens up. beautifuuuulllll.

a neat addition to the gate sequencers is being able to have 2 patterns you can switch between. if that's a doable thingie.
Attackman
Looks like you dropped the S&H which appeared in an earlier iteration. Will you reintroduce it? I can always use one.

In the realm of utility, would a basic chromatic tuner be possible? It's not exciting but useful when needed.

I was late to bed last night! Couldn't stop playing with Hemisphere.
4eyez
Attackman wrote:
Looks like you dropped the S&H which appeared in an earlier iteration. Will you reintroduce it? I can always use one.


After reading the manual obsessively in anticipation of the arrival of my uO_C this week, I believe S&H has been moved to one of the operations of the Calculate applet.

Attackman wrote:
In the realm of utility, would a basic chromatic tuner be possible? It's not exciting but useful when needed.


That's a great idea. And since we're talking about microtonal stuff, if it could also be used to tune microtonally in addition to chromatic, that would be next level. Though I'm trying to think of how that would work, would it display cents in addition to the chromatic letter notes? Or would you select one of the scale files and match to that, like a guitar tuner but instead of displaying for each string you would be displaying for each step in the scale. hmmm.....

I actually have a custom made lap steel guitar being made for me and I was thinking of how I could tune the individual strings microtonally, something like this would make that easy. Heavy.
chysn
4eyez wrote:
I believe S&H has been moved to one of the operations of the Calculate applet.


That’s right, consolidated to Calculate along with random.

Quote:
Attackman wrote:
In the realm of utility, would a basic chromatic tuner be possible? It's not exciting but useful when needed.


That's a great idea. And since we're talking about microtonal stuff, if it could also be used to tune microtonally in addition to chromatic, that would be next level. Though I'm trying to think of how that would work, would it display cents in addition to the chromatic letter notes?


Yeah, it would basically be a port of the tuner from References. It’s been on my list for a while now, and it’s probably time to finish it up.
chysn
behndy wrote:
a neat addition to the gate sequencers is being able to have 2 patterns you can switch between. if that's a doable thingie.


You mean a trigger sequencer with CV-selectable patterns?
Attackman
4eyez wrote:
Attackman wrote:
Looks like you dropped the S&H which appeared in an earlier iteration. Will you reintroduce it? I can always use one.


I believe S&H has been moved to one of the operations of the Calculate applet.


Thanks. I missed that. It is there. Appreciate you pointing it out.

4eyez wrote:
Attackman wrote:
In the realm of utility, would a basic chromatic tuner be possible? It's not exciting but useful when needed.


if it could also be used to tune microtonally in addition to chromatic, that would be next level.


Yeah! That didn't even occur to me but how awesome.

4eyez wrote:

I actually have a custom made lap steel guitar being made for me and I was thinking of how I could tune the individual strings microtonally, something like this would make that easy. Heavy.


That's an idea. I play mandolin. Never thought of throwing that into strange tunings but I can see how great that'd be. I should be more adventurous! I have recorded it into Morphagene and had some lovely results. I say lovely - my aesthetic veers sharply towards abstract noise and glitchy clicks or perturbing ambiance.
sushiluv
Bad news, pong seems to be broken in 1.3.4!

On a more serious note, it seems like Darkest Timeline doesn´t output midi.
Am i right or am i doing something wrong?

Thanks for the great work, love what you did with o+c!
chysn
sushiluv wrote:
Bad news, pong seems to be broken in 1.3.4!


I just checked it out and it's working okay. What kind of problem are you seeing?

Quote:
On a more serious note, it seems like Darkest Timeline doesn´t output midi.


It does not output MIDI notes yet. It should perform a system exclusive dump when you suspend it (long-press the right encoder to go back to the main menu). This is just for saving sequences that you've created.

Note: I do plan on updating The Darkest Timeline very soon. There are still two unused inputs and the left button press is undefined. I'll probably use that to open up an output configuration screen to set a MIDI channel and scale.
in_sherman
Just here to express my appreciation for your efforts in enhancing an already useful and pleasurable tool. Thrilled to try this out.
chysn
in_sherman wrote:
Just here to express my appreciation for your efforts in enhancing an already useful and pleasurable tool. Thrilled to try this out.


Thanks, I'm looking forward to hearing your thoughts!

Meanwhile, my new panel arrived from Front Panel Express!

Zymos
What, you ditched the Magpie? Or is it a new build?
chysn
Zymos wrote:
What, you ditched the Magpie? Or is it a new build?


Same O_C, I just thought it would be fun to design a panel.
chysn
OK, I finished the Tuner after work today. It should be pretty useful with microtonal scales, in that it tells you the number of cents away from in-tune you are, and you can set the A4 reference between 400-500Hz.

I don't know how (or whether) it will work with acoustic instruments.

Since Tuner is essentially a ported subset of References, and uses the same Arduino frequency-measuring library, it's hard-coded to accept its input on Digital 4. I think the capture input is a compile-time setting, and I don't think there's a way to change it without interfering with References. In other words, the Tuner will only work when it's in the right hemisphere. It will gently remind you of this.
chysn
behndy wrote:
a neat addition to the gate sequencers is being able to have 2 patterns you can switch between. if that's a doable thingie.


In the next release, for both TrigSeq and TrigSeq16, if the CV1 input is high, the A and B outputs will be swapped. For TrigSeq, this will effectively behave like a CV-controlled pattern switcher.
rayultine
ehh chysn,
I'm suddenly getting a locked o_C if I switch from a vanilla app to Hemispheres. Using the recommended arduino/teensy versions and upgraded to beta 5. Still happening. :(
chysn
rayultine wrote:
ehh chysn,
I'm suddenly getting a locked o_C if I switch from a vanilla app to Hemispheres. Using the recommended arduino/teensy versions and upgraded to beta 5. Still happening. :(


Yeah, the most recent commit of the old repo does crash, but I consider it a dead project and don't plan any future commits. I would recommend installation via the final hex file for Beta 5, which is here:

https://github.com/Chysn/O_C-Hemisphere/releases

Basically, I'm tossing out the non-Suite version of Hemisphere. I took a mulligan.
mxmxmx
chysn wrote:

Since Tuner is essentially a ported subset of References, and uses the same Arduino frequency-measuring library, it's hard-coded to accept its input on Digital 4. I think the capture input is a compile-time setting, and I don't think there's a way to change it without interfering with References. In other words, the Tuner will only work when it's in the right hemisphere. It will gently remind you of this.


yeah, input capture only works with TR input #4 (= PTA13), but not #1 - #3; those aren't connected to any of the FTM channels. it's a hardware thing (and using input capture was an afterthought).
behndy
chysn wrote:
behndy wrote:
a neat addition to the gate sequencers is being able to have 2 patterns you can switch between. if that's a doable thingie.


In the next release, for both TrigSeq and TrigSeq16, if the CV1 input is high, the A and B outputs will be swapped. For TrigSeq, this will effectively behave like a CV-controlled pattern switcher.


oooooo! very elegant suh.
chysn
mxmxmx wrote:
chysn wrote:

Since Tuner is essentially a ported subset of References, and uses the same Arduino frequency-measuring library, it's hard-coded to accept its input on Digital 4. I think the capture input is a compile-time setting, and I don't think there's a way to change it without interfering with References. In other words, the Tuner will only work when it's in the right hemisphere. It will gently remind you of this.


yeah, input capture only works with TR input #4 (= PTA13), but not #1 - #3; those aren't connected to any of the FTM channels. it's a hardware thing (and using input capture was an afterthought).


Thanks for the clarification!

I suspected it was a hardware issue when I opened the OC_FrequencyCapture file and it looked like the sort of thing that I better leave alone.

Using the FTM was a nice afterthought, though. I tried doing a frequency counter with the ADC inputs and that was just a disaster. A million isn't evenly-divisible by (440 * 60).
Maximars
One little idea for the Logic App:

Set/Reset Latch Flip Flop


Or maybe a little Flip-Flop App like the Ladik B-011?
http://ladik.ladik.eu/?page_id=1910
With the following types:

RS-Flip-Flop
T-Flip-Flop (I know, it's a frequency divider)
D-Flip-Flop

What do you think?
chysn
Maximars wrote:
One little idea for the Logic App:

Set/Reset Latch Flip Flop


Or maybe a little Flip-Flop App like the Ladik B-011?
http://ladik.ladik.eu/?page_id=1910
With the following types:

RS-Flip-Flop
T-Flip-Flop (I know, it's a frequency divider)
D-Flip-Flop

What do you think?


I'm trying to wrap my brain around it. The T is, like you said, a frequency divider. The D seems a lot like a binary sample and hold. And an SR latch is basically a two-step sequencer with one step set to 5V and another set to 0V. So it seems like all this functionality is already there, unless some important concept is whizzing over my head, which happens.

You can also get some interesting "latching" behaviors out of things like Switch and Threshold Logic Neuron.

I will add voltage offset controls to the Switch applet so it will work without any inputs. This will literally turn it into a 2-step sequencer, and should fill the role of the SR latch.
Maximars
RS-Flip Flop pseudo-code


on risingedge of trigger input 1

if mybit = false then
mybit = true
set output1 high
end if

end on

on risingedge of trigger input 2

if mybit = true then
mybit = false
set output1 low
end if

end on

rs flip flop wold be very nice :-)
chysn
I see, so one input always turns it on, and one input always turns it off. Try this in Threshold Logic Neuron ("TL Neuron"):
    * Set dendrite 1 to 1
    * Set dendrite 2 to -1
    * Set dendrite 3 to 1
    * Set threshold to 0
    * Patch Output B into CV 1

Now Digital 1 sets and Digital 2 resets, and your output comes from Output A.
mazzyboy
Wow - I suggested a shuffled app, and boom... here it is a few days later!!!

Thanks so much for this. Really dig what you are doing with these new apps
rayultine
chysn wrote:
rayultine wrote:
ehh chysn,
I'm suddenly getting a locked o_C if I switch from a vanilla app to Hemispheres. Using the recommended arduino/teensy versions and upgraded to beta 5. Still happening. :(


Yeah, the most recent commit of the old repo does crash, but I consider it a dead project and don't plan any future commits. I would recommend installation via the final hex file for Beta 5, which is here:

https://github.com/Chysn/O_C-Hemisphere/releases

Basically, I'm tossing out the non-Suite version of Hemisphere. I took a mulligan.


Hrmph... Beta 5 was locking up too. Ok then I took the plunge and put the Suite on my o_C. Haven't taken it for a spin though.

feature request: any chance the gated VCA outputs could have the option of being inverted?
chysn
rayultine wrote:
Hrmph... Beta 5 was locking up too.


Huh. If you installed from the hexfile, there shouldn't be much that can go wrong. I'd suggest resetting the EEPROM by holding down the Up and Down buttons while restarting your system, hold them down until the O_C boots up, and then click the right encoder. This will clear out the data from your O_C*, and make sure that nothing is confusing Hemisphere.

Quote:
feature request: any chance the gated VCA outputs could have the option of being inverted?


What do you mean by "inverted" here? You mean using a gate to turn off the output? Or voltage inversion of the output (e.g., from positive to negative)? If you mean the former, I think what I'll do is make it so that Digital 2 going high turns off the B/D output. In almost all cases, it'll behave exactly the same, but it'll just be a bit more flexible.

If you meant "voltage inversion," I usually resist that in the applets because the O_C's I/O is asymmetrical, and I think it'll confuse people. To invert O_C's outputs, it's way better to use Shades or Kinks or whatever.

__________
* This operation will not destroy your O_C's calibration
chysn
I did a couple more applets today, because I was off work sick. In addition to handling a couple requests here, I added:

* Dual Schmitt Trigger (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Schmitt-Trigger)
* This strange binary counter (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Binary-Counter)

I'm not totally certain whether the binary counter will have any applications, but I've never seen anything like it. The Schmitt Trigger is good for squaring off ragged input, and other things.
Maximars
Thank you for your efforts! This module will save me a bucket full of utility modules thumbs up
Attackman
chysn wrote:
OK, I finished the Tuner after work today. It should be pretty useful with microtonal scales, in that it tells you the number of cents away from in-tune you are, and you can set the A4 reference between 400-500Hz.


I'm looking forward to the update release next week. I'm definitely appreciating what you've done with Hemisphere. It's excellent, so many useful little apps and largely either intuitive to use or quick to find a use for. One of the great things you've done with Hemisphere is the Help menu - that is SOOOO brilliant. It allows for an uninterrupted flow - a quick check of a menu and I know what I'm doing or how to best get on with it. That cannot be underestimated and was probably my biggest gripe when switching apps in o_C. Needing to refer to a manual.

As you seem open to suggestions ... I'd love a dual low pass gate emulation of an Optomix or failing that a regular audio vca.

BTW that Carpeggio applet is little short of genius. Such a neat idea. Ran it all over the O Coast last night. Ace
chysn
Attackman wrote:
As you seem open to suggestions ... I'd love a dual low pass gate emulation of an Optomix or failing that a regular audio vca.


I'd really like a platform on which to do audio processing stuff and oscillators, but Ornament and Crime isn't it. If you want to see how well it handles audio, run an audio signal through Gated VCA. That's literally about as good as it's capable of getting. The voltage range is asymmetrical (-3V ~ 6V), the interrupt can only access the I/O at 17kHz, and it has a grand total of 64K available for volatile data system-wide. So it's just not cut out for that kind of thing.

Salt is a bit out of my league, price-wise, and lacks a screen, and it's pretty much unobtanium anyway. So I'll wait for the next bump in processing power and get into DSP "some day."

Quote:
BTW that Carpeggio applet is little short of genius. Such a neat idea. Ran it all over the O Coast last night. Ace


I'm glad that Carpeggio is getting some love! It's one applet that, if I had the means, I'd really like to see as a 20HP full-sized physical module with great big friendly buttons. If somebody wants to build it, they can even have the name "Carpeggio," which is like the best name in the world.
rayultine
chysn wrote:

If you meant "voltage inversion," I usually resist that in the applets because the O_C's I/O is asymmetrical, and I think it'll confuse people. To invert O_C's outputs, it's way better to use Shades or Kinks or whatever.


It could be a useful feature for those who don't have Shades or Kinks razz
Attackman
chysn wrote:

I'd really like a platform on which to do audio processing stuff and oscillators, but Ornament and Crime isn't it.


It's not my area, I only have a rudimentary grasp of these things but I wonder if this is any good? I don't do programming but I find these sorts of things appealing.

https://1010music.com/product/euroshield1
https://www.thonk.co.uk/shop/1010-music-euroshield-1-bundle/

Alternatively, Axoloti
http://www.axoloti.com/product/axoloti-core/

but it's heading out of eurorack. I think some people have tried to mod it but essentially it's not eurorack. Very nice tho.
chysn wrote:

I'm glad that Carpeggio is getting some love! It's one applet that, if I had the means, I'd really like to see as a 20HP full-sized physical module with great big friendly buttons. If somebody wants to build it, they can even have the name "Carpeggio," which is like the best name in the world.


Funny you should say that. As I was using it I was kind of laughing thinking how ridiculous that this sits on such a tiny screen when it feels like it should be the size of a Rene. I like a good 18/20hp size module. It's good to get in and use but once rack space is tight it's a harder decision - well, that and the cost. Strangely with Carpeggio, it works perfectly in a tiny form factor because the encoder can hop over values and quickly change notes etc. So it's easy to play with. It ensures you use your ears as much as the screen.
chysn
Attackman wrote:

https://1010music.com/product/euroshield1
https://www.thonk.co.uk/shop/1010-music-euroshield-1-bundle/


I've seen those before, but I haven't spent much time learning about them. Could be interesting.
blinosynth
Hey it could be possibile to add an envelope follower?
chysn
blinosynth wrote:
Hey it could be possibile to add an envelope follower?


It’s been on my list, but I’ve been putting it off as it promises to be tricky. Maybe I’ll work on figuring it out this weekend.
digitalohm
extending the AD EG or making another AD EG mode that is just two AD EG instead of the reverse mode might be cool. I keep thinking it would be handy to have 2xAD EG alongside the Dual Quantizer since it would move towards more voices.

The dual AD EG would just be tr1 and tr2 for triggering the envelope, CV control over either attack or decay (user definable) and CV out for the envelope.

also cheers on that gate delay grin
chysn
digitalohm wrote:
extending the AD EG or making another AD EG mode that is just two AD EG instead of the reverse mode might be cool. I keep thinking it would be handy to have 2xAD EG alongside the Dual Quantizer since it would move towards more voices.

The dual AD EG would just be tr1 and tr2 for triggering the envelope, CV control over either attack or decay (user definable) and CV out for the envelope.

also cheers on that gate delay grin


It's a good idea. The ADSR actually works the way you're suggesting, if you want to have two envelopes.

For the AD, the second output is already the end-of-cycle, which I think mixes things up pretty nicely. Since the second output was spoken for, the reverse trigger was just sort of a quirky bonus, which may or may not prove useful. You can patch EOC into Reverse to create a cycling function that starts in one direction, but cycles in the opposite direction!
chysn
blinosynth wrote:
Hey it could be possibile to add an envelope follower?


All right, here's the deal. I don't have an envelope follower. I've never had an envelope follower. What I've got here is probably more like a caricature of an envelope follower. So don't feel bad about telling me if it's awful.
starthief
Do you have a slew? Function, Maths? You can patch up a usable envelope follower that way.

A fast-ish rise, and a fall just slow enough to smooth out the audio-rate wiggles.

My preferred settings on Function are Rise at about 10-11 o'clock, fall at about 3-4 o'clock, and Lin/Log at about 2 o'clock. This matches the envelope follower in Bastl Dynamo pretty well.

How to test it: feed your follower Rings or an LPG pluck or a drum sound or something, watch it on a scope and listen to it. Ideally you should only hear a click or thump at the attack, not the sound of the source audio during the decay. If you can hear the source, the fall needs to be slower.

A slower attack is also very useful -- just increase the rise time so the thump is barely audible (about 9 o'clock on Function).
chysn
starthief wrote:
A slower attack is also very useful -- just increase the rise time so the thump is barely audible (about 9 o'clock on Function).


Thanks! What I cooked up was sort of a fast slew over a 10ms maximum. I might be fine with just the slew part. I'll play around with it. The big limitation is that the inputs aren't normalized for audio, so the dynamic range is poor. With LoFi Tape, that was kind of a feature; but with an envelope follower, dynamic range is kind of important and I'm really feeling its limitation.
Kroaton
Could Captain Midi be modified to include a MPE high-resolution midi mode?

That'd make it the first open-source Eurorack module to support controllers with MPE midi like the Roli / Linnstrument / Haken Continuum etc.

Here is the MPE 1.0 specification:

https://drive.google.com/file/d/1B5S0umtu8V4sxP5GZgD_APJZv4je90Gw/view
erstlaub
Great work, thanks.

I seem to sometimes get a weird issue with game of life running where it just seems to null out and it doesn't matter what I feed into it, nothing moves. Sorry if that's too vague, I've had it happen a couple of times but I couldn't yet figure out how I broke it. A power cycle is the only thing that gets it moving again (or maybe I've managed to cv it into a 'game over' state?)?

The only thing I really miss from the original is Sequins as it was always handy to have a more flexible sequencer in rack.

Cheers.

EDIT: Nevermind, just noticed that there's a new version and GoL has been retired.
chysn
Kroaton wrote:
Could Captain Midi be modified to include a MPE high-resolution midi mode?

That'd make it the first open-source Eurorack module to support controllers with MPE midi like the Roli / Linnstrument / Haken Continuum etc.


That's a good question, and it has several answers.

The too-long-didn't-read answer is "yes." I can add support for CC#74 easily, and other controllers that are important to MPE devices. MPE doesn't add any new types of MIDI messages, so I don't need to worry about support from Teensyduino. Furthermore, I have a pretty good idea of how zones would be defined and used. And finally, some aspects of MPE are already supported, based on the flexibility of assigning controllers and notes to channels.

There's a more nuanced answer which involves things like

(1) Assuming that there's not much value in implementing the entire specification, what use cases are people most interested in?
(2) I don't, myself, have an instrument that supports MPE and I don't see getting one in the foreseeable future, so testing would be a pain on my end.
(3) I'd rather support MPE than MIDI clock, so maybe it's something to shoot for.

You've got me thinking about it, certainly.

erstlaub wrote:
DIT: Nevermind, just noticed that there's a new version and GoL has been retired.


There are some states that won't reproduce, or will collapse immediately. But, yeah, it's on the way out, anyway, for good reason. It consumes a lot of memory and processing power for something that stretches the definition of "musical" just a bit too much. Mostly I wanted to see if the Teensy could handle it, which it can, for the most part (maybe not at audio rates...)

The coolest thing that could be done with it is plugging the LowerRenz X and Y outputs into GoL's X and Y inputs and feeding it a clock. Then it basically looks like the Low Rents screensaver. smile

(This would seem to suggest that Pong is on the chopping block next. But Pong is relatively featherweight, and I have to give people a chance to beat my score of 24 when played autonomously by a Eurorack patch)
chysn
Kroaton wrote:
Could Captain Midi be modified to include a MPE high-resolution midi mode?


I'm going to start by supporting the Y-Axis controller (CC#74) in version 1.4. That's the super-low-hanging fruit of MPE support. I don't think I'd support multiple zones, because the O_C simply doesn't have enough outputs to justify that work. But the next step would be support of MPE-style polyphony, which is basically like Captain MIDI's current polyphonic assignment, but listening to a single zone instead of to a single channel.

From a technical standpoint, it's not that difficult a problem. But I'm having some reservations about the practicality. The idea of MPE is to allow a performer to apply expression to individual notes by assigning those notes to different MIDI channels. But O_C has a total of 4 CV outputs. Even if you decline to assign gates, how much polyphonic expression can you really achieve? I'd like to know what a performer's aims might be before going too far into it.
Timmy
Kroaton wrote:
Could Captain Midi be modified to include a MPE high-resolution midi mode?

That'd make it the first open-source Eurorack module to support controllers with MPE midi like the Roli / Linnstrument / Haken Continuum etc.

Here is the MPE 1.0 specification:

https://drive.google.com/file/d/1B5S0umtu8V4sxP5GZgD_APJZv4je90Gw/view


Expert Sleepers FH-2 supports MPE: https://matttechmodular.co.uk/product/fh-2/
Kroaton
chysn wrote:
Kroaton wrote:
Could Captain Midi be modified to include a MPE high-resolution midi mode?


I'm going to start by supporting the Y-Axis controller (CC#74) in version 1.4. That's the super-low-hanging fruit of MPE support. I don't think I'd support multiple zones, because the O_C simply doesn't have enough outputs to justify that work. But the next step would be support of MPE-style polyphony, which is basically like Captain MIDI's current polyphonic assignment, but listening to a single zone instead of to a single channel.

From a technical standpoint, it's not that difficult a problem. But I'm having some reservations about the practicality. The idea of MPE is to allow a performer to apply expression to individual notes by assigning those notes to different MIDI channels. But O_C has a total of 4 CV outputs. Even if you decline to assign gates, how much polyphonic expression can you really achieve? I'd like to know what a performer's aims might be before going too far into it.


Considering the lack of outputs needed for proper polyphony, support for Y-Axis is probably as good as it can get on the O_c without some sort of expander module (at that point the FH-2 might be better suited, as Timmy pointed out).

With Y and Z-Axis support though it'd be a great companion to the cheap Roli Seaboard Block instrument to control mono-voiced modules such as Elements/Rings etc. (X - 1v/oct pitch, gate, Y and continuous pressure - Z)
Supervillain
THANKS FOR THAT FIRMWARE !!!

to improve some of the app i got a few things in mind.

In quantizer app it would be usefull to have a root note selection parameter.

In 5 step sequencer app it would be nice to have one of the CV input to act as a voltage adder to transpose the whole sequence.

last but not least, Would it be possible to make some kind of sidechain/ducking app ? maybe to use along with the VCA app
starthief
Supervillain wrote:
last but not least, Would it be possible to make some kind of sidechain/ducking app ? maybe to use along with the VCA app


That's one thing an envelope follower is great for, though you'd want to invert and offset the output.
chysn
Supervillain wrote:
In quantizer app it would be usefull to have a root note selection parameter.


Yeah, I think that's a good idea. There's enough room on the screen, I just need to re-arrange it a bit. Dual Quantizer was one of the first applets, and my visual conventions have evolved since then, to where I think a minor re-thinking is in order.

Quote:
In 5 step sequencer app it would be nice to have one of the CV input to act as a voltage adder to transpose the whole sequence.


I love this idea so much that I already did it last week! It'll be in this Friday's release.

Quote:
last but not least, Would it be possible to make some kind of sidechain/ducking app ? maybe to use along with the VCA app


As Starthief mentioned, the envelope follower is coming. I'll look at adding settings that'll allow it to be used for ducking. Or maybe I'll make it so that there's one follower channel and one ducking channel.
4eyez
Got my uO_C on Friday and immediately installed Hemispheres. It's really nicely done, thanks for your work!

Took a quick stab at trying to tune my guitar with the tuner, unsurprisingly it wasn't really working, it was a long shot at best. No biggy, tuned my oscillators and that's the bigger thing.

I also tried out the scale editor. It's really slick, but I'm not sure if the increments work. Looking through the O&C predefined scale list, most of them use whole numbers and with the way the values increase you can never get an exact whole number. Maybe it would make most sense to click through for whole numbers and then a second click to decimal place. And I don't think you really need to go more than one decimal place. But, tbh, I'm not a deep microtonal user, just starting to dabble. Just some thoughts.

With Palimpsest is there a way to have CV control for either Compose or Decompose? I'm not sure which is better, maybe there's no difference from a certain perspective. Right now, without CV for either, at some point the accents fill up or empty out until you adjust one or the other. Does that make sense?

Is it possible to have a syncable LFO? Where triggers would determine the overall length of the LFO? Something like the tap mode on Peaks? Peaks can learn trig patterns of a certain length, that would be nice, but even just setting length from the last two triggers is sufficient.

Supervillain wrote:

In quantizer app it would be usefull to have a root note selection parameter.


Don't you simply do that by tuning the oscillator. A quantizer just limits how far apart each step is in the scale. So, 0 in the scale could be a C or an F depending on how you tune the oscillator, no? That's how I'm used to it working.
Maximars
A little idea for an app: Raw precise voltages


The voltage is selectable in semi tone steps of 0.0833 volt, in the positive as well as in the negative range.

It has a gate input. If the gate is high then the app outputs the selected voltage. If the gate is low it output 0 volt. Maybe zero volt has to be fine tune able.

I think this is use full in connection with en external precision adder for transpositions. It would also be possible to plug this direct in to the Exp FM input of an VCO.


the background behind it.
I like to mix some lfo's and then put the cv through a quantizer. Sometimes this gehts out of control, then you get very high or very low pitches. These could be caught for instance with the Schmittriger App and then transpose it up or down an octave.
chysn
4eyez wrote:
I also tried out the scale editor. It's really slick, but I'm not sure if the increments work. Looking through the O&C predefined scale list, most of them use whole numbers and with the way the values increase you can never get an exact whole number.


Oh, yeah, those increments are right. You're tuning in increments of 1/1.28 cents, so the whole numbers will land on the semitone (increments of 100 cents), diatonic quarter-tone (50 cents) and diatonic eighth-tones (25 cents/75 cents). An equal-tempered 12-note scale goes 0, 100, 200, 300, etc., because each note is 100 cents higher than the last one.

Quote:
With Palimpsest is there a way to have CV control for either Compose or Decompose? I'm not sure which is better, maybe there's no difference from a certain perspective. Right now, without CV for either, at some point the accents fill up or empty out until you adjust one or the other. Does that make sense?


Yeah, I like that. Palimpsest has its origins in a application I developed for Peaks, which had no CV control, so I didn't really consider that kind of modulation. I might make it so that positive modulation increases Compose and negative modulation increases Decompose. I've got some time to play around with it.


Quote:
Is it possible to have a syncable LFO? Where triggers would determine the overall length of the LFO? Something like the tap mode on Peaks? Peaks can learn trig patterns of a certain length, that would be nice, but even just setting length from the last two triggers is sufficient.


Yeah, that's on the drawing board. I want at least one more LFO, and I need sync to compensate for the loss of ModBox.

Quote:
Don't you simply do that by tuning the oscillator. A quantizer just limits how far apart each step is in the scale. So, 0 in the scale could be a C or an F depending on how you tune the oscillator, no? That's how I'm used to it working.


Tuning the oscillator is more like a transposition than a root change. For many scales, you can change the root without changing the transposition at all. For example, let's say you have a major (Ionian) scale with a root of C, and your quantized note is a G, and then you change the root to D; it will sound like nothing is happening at all, because Ionian with a root of D also has a G in it (just as 4 instead of 5). Whereas, with transposition (or oscillator tuning change), you're maintaining a diatonic relationship, so you'd move from G to A.
chysn
Maximars wrote:
A little idea for an app: Raw precise voltages


That's kind of fun. It's like a two-note quantizer! Hemisphere Suite is missing an offset generator, and this could be a neat alternative that can also be something else. I wonder if it should be two independent channels?
insoul8
So, just to be clear, the only original o_C app that had to be deleted to make room for this was MetaQ? I won't lose any other o_C functionality? If so, this looks friggin' amazing and I can't wait to give it a try.
chysn
insoul8 wrote:
So, just to be clear, the only original o_C app that had to be deleted to make room for this was MetaQ? I won't lose any other o_C functionality? If so, this looks friggin' amazing and I can't wait to give it a try.


That was the original plan, but I scrapped that idea. This is now total replacement. The only original O_C app that will remain is References.
synkrotron
chysn wrote:


That was the original plan, but I scrapped that idea. This is now total replacement. The only original O_C app that will remain is References.


Ah... I see. So there is no quadrature LFO now, or am I missing something?

Oh, hi, by the way. I'm new to o_C and I have been keeping on eye on Hemispheres smile
insoul8
chysn wrote:
insoul8 wrote:
So, just to be clear, the only original o_C app that had to be deleted to make room for this was MetaQ? I won't lose any other o_C functionality? If so, this looks friggin' amazing and I can't wait to give it a try.


That was the original plan, but I scrapped that idea. This is now total replacement. The only original O_C app that will remain is References.


Ahh, got it. Certainly still worth trying but I'm not sure i can give up Copier Maschine and Quantermain permanently. Those are definitely two of the apps i use the most.
chysn
synkrotron wrote:
chysn wrote:


That was the original plan, but I scrapped that idea. This is now total replacement. The only original O_C app that will remain is References.


Ah... I see. So there is no quadrature LFO now, or am I missing something?


That's right. There wasn't enough memory left in the O_C to make something that could live alongside the current firmware, and later I wanted to extensively support MIDI, so I gave up on trying to preserve the original stuff. You can think of Hemisphere Suite as a totally different module that looks like an O_C and shares some of its cultural sensibilities.

But I'm already facing the same problem that the O_C team faced, which is that I'm running out of space. Hemisphere Suite uses 75% of the program memory, so the main phase of development will be pretty much over by the end of the year.
synkrotron
chysn wrote:

You can think of Hemisphere Suite as a totally different module that looks like an O_C and shares some of its cultural sensibilities.


Okay, thanks. You're doing a great job here but I've only just got o_C and was using the quad LFO last night and I think I need that more than some of the Hemisphere stuff.

I'll still be keeping an eye on developments here, of course... A voltage offset thing would be great for BSP users...


cheers

and thanks again,

andy
uebl
So, does hemisphere now come with a "gate delay" app? I see it's mentioned in this thread, but I can't find anything in the release notes and the .ino files on GitHub. Sorry if I just overlooked it.

Oh, and would installing hemisphere require re-calibrating my uO_C?

Edit: I see it shouldn't require it.
chysn
synkrotron wrote:
chysn wrote:

You can think of Hemisphere Suite as a totally different module that looks like an O_C and shares some of its cultural sensibilities.


Okay, thanks. You're doing a great job here but I've only just got o_C and was using the quad LFO last night and I think I need that more than some of the Hemisphere stuff.


That's cool, thanks for checking it out!
chysn
uebl wrote:
So, does hemisphere now come with a "gate delay" app? I see it's mentioned in this thread, but I can't find anything in the release notes and the .ino files on GitHub. Sorry if I just overlooked it.

Oh, and would installing hemisphere require re-calibrating my uO_C?

Edit: I see it shouldn't require it.


Yeah, Gate Delay is in Hemisphere Suite from v1.2, and the documentation is here: https://github.com/Chysn/O_C-HemisphereSuite/wiki/Gate-Delay

You don't need to re-calibrate.
Maximars
Ideas for the offset generators (all semitones quantised):
- Encoder selectable Oktave offsets -2, -1, 0, +1, +2 Volt
- Encoder selectable semitones offsets
- by gate switching between two selectable voltages (0.08333 Volt quantized)
- if cv input goes over a selectable threshold (with hysteresis), it outputs a selectable voltage.
- if cv goes below a selectable threshold (with hysteresis), it outputs a selectable voltage.
-4 selectable voltages selected with two gate inputs

For all of this an analog unity mixer or a precision adder is helpfully.
May it’s called a stepped offset generator nanners
Edit:
Yes two channels if possible thumbs up
digitalohm
chysn wrote:

But I'm already facing the same problem that the O_C team faced, which is that I'm running out of space. Hemisphere Suite uses 75% of the program memory, so the main phase of development will be pretty much over by the end of the year.


one idea might be to separate the apps into different branches so that certain functionalities, ie hemisphere - gate utilities, hemisphere - cv utilities - hemisphere - functions etc

or maybe a streamlined way to pick and choose x number of apps, compile them and flash!
srogers
digitalohm wrote:


or maybe a streamlined way to pick and choose x number of apps, compile them and flash!


This would be amazing. If there were a reasonably easy way to be able to choose a number of apps from both the original and Hemisphere Suite firmwares I'd be pretty thrilled - which I guess was the idea of the original Hemisphere firmware but it'd be cool to choose the apps.
chysn
https://github.com/Chysn/O_C-HemisphereSuite/wiki/Voltage
debolish
insoul8 wrote:
Ahh, got it. Certainly still worth trying but I'm not sure i can give up Copier Maschine and Quantermain permanently. Those are definitely two of the apps i use the most.


Same here. Just scooped a second o_c off the BST to enjoy the best of both.
uebl
chysn wrote:
uebl wrote:
So, does hemisphere now come with a "gate delay" app? I see it's mentioned in this thread, but I can't find anything in the release notes and the .ino files on GitHub. Sorry if I just overlooked it.

Oh, and would installing hemisphere require re-calibrating my uO_C?

Edit: I see it shouldn't require it.


Yeah, Gate Delay is in Hemisphere Suite from v1.2, and the documentation is here: https://github.com/Chysn/O_C-HemisphereSuite/wiki/Gate-Delay

You don't need to re-calibrate.


I missed the jump to "Hemisphere Suite" and was still looking at the beta 1-5 releases d'oh!

I'll give it a try today, thank you so much, chysn thumbs up thumbs up we're not worthy we're not worthy
uebl
I just installed Hemisphere Suite v1.3 and tried out a few applets incl. gate delay. Thank you so much for this!

One thing I noticed it that it seems the encoders resolution to set the gate delay time is 10ms, not 1ms as written in the documentation (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Gate-Delay):
Quote:
Encoder Turn: Set delay time in 1ms increments, from 0ms to 2000ms


(I'd love to have the finer resolution!)
Supervillain
chysn wrote:
Supervillain wrote:
In quantizer app it would be usefull to have a root note selection parameter.


Yeah, I think that's a good idea. There's enough room on the screen, I just need to re-arrange it a bit. Dual Quantizer was one of the first applets, and my visual conventions have evolved since then, to where I think a minor re-thinking is in order.

Quote:
In 5 step sequencer app it would be nice to have one of the CV input to act as a voltage adder to transpose the whole sequence.


I love this idea so much that I already did it last week! It'll be in this Friday's release.

Quote:
last but not least, Would it be possible to make some kind of sidechain/ducking app ? maybe to use along with the VCA app


As Starthief mentioned, the envelope follower is coming. I'll look at adding settings that'll allow it to be used for ducking. Or maybe I'll make it so that there's one follower channel and one ducking channel.


You're the man ! nanners

One small thing, In scope mode my CV display (Bottom of the screen) is very wiggly. Is it normal ?
chysn
uebl wrote:
One thing I noticed it that it seems the encoders resolution to set the gate delay time is 10ms, not 1ms as written in the documentation


Yep, good catch! I'll correct it to 1ms shortly.

Update: I might have to make it 2 or 3ms, or it takes forever to get from one place to another, even with acceleration. But it'll definitely be way finer than 10ms.

Supervillain wrote:
One small thing, In scope mode my CV display (Bottom of the screen) is very wiggly. Is it normal ?


It is normal. There seems to be a "noise floor" in the ADC, and I'm not using the "smoothed" function. In the upcoming update, that display will be in volts, and it's slightly less wiggly because I go to only two decimal places. But still a little wiggly.
rayultine
chysn,
I seem to recall earlier discussion of the ASDR snappiness. Would it be possible to make the attack a little more fine-grained? It seems like 0-setting is very clicky and requires a lot of fussing with the VCA. The next value is quite slow.
uebl
chysn wrote:
uebl wrote:
One thing I noticed it that it seems the encoders resolution to set the gate delay time is 10ms, not 1ms as written in the documentation


Yep, good catch! I'll correct it to 1ms shortly.

Update: I might have to make it 2 or 3ms, or it takes forever to get from one place to another, even with acceleration. But it'll definitely be way finer than 10ms.


Does o_C support something like a "pressed turn" of an encoder? This way, it could be 10ms per encoder step, and 1ms per "pressed"-encoder step. Otherwise something like 2ms or 3ms is of course perfectly fine.

Or do something like 1ms steps from 0 to 100ms, then 2-5ms from 100 to 250 and 10ms from thereon..
chysn
uebl wrote:
Does o_C support something like a "pressed turn" of an encoder?


It's supported, yes. However, in the original O_C firmware, you'll only see this done for the left encoder, because long-press-and-release of the right encoder always takes you back to the main menu. Since each Hemisphere applet uses only one encoder, and it can be either of the encoders, that particular control combination isn't available to me.

Quote:
Or do something like 1ms steps from 0 to 100ms, then 2-5ms from 100 to 250 and 10ms from thereon..


That's sort of how the Scope fine-tuning works. Even if it maxed out at 5ms, it would be easy to get around.
chysn
rayultine wrote:
chysn,
I seem to recall earlier discussion of the ASDR snappiness. Would it be possible to make the attack a little more fine-grained? It seems like 0-setting is very clicky and requires a lot of fussing with the VCA. The next value is quite slow.


The increments are already 0-255, which is pretty fine-grained. But I'll double-check the math that's going on between 0 and everything else.
rayultine
I like the 0 value click, but I think the 1 value is realllllyyyy different. If that makes sense.
chysn
rayultine wrote:
I like the 0 value click, but I think the 1 value is realllllyyyy different. If that makes sense.


It does. It comes out to about 15ms right now, which is probably too long.
Supervillain
Chysn for president !

Your raectiveness is absolutely mad ! SlayerBadger!

You restored my faith in O_c, so thanks a lot
chysn
Hemisphere Suite 1.4 is available at

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.4
rayultine
unreal!!! Thanks chysn, also thank you for the 180flipped hex, which I might try out as I rearrange my case.
Mashmore
SlayerBadger! Many Thanks SlayerBadger!
chysn
Okay... here's the Scala-to-Hemisphere Suite converter:

http://www.beigemaze.com/scala/

You paste in (or upload) your Scala document, and download the System Exclusive file, which you then transfer to your O_C while Scale Editor is running.
Boofin Moonrocks
Would hemisphere be able to run on an ornament and crime modified to have a voltage range of 0-9v?
chysn
Here's a spoiler for upcoming development, scheduled for release near the end of September.

When I started planning the Turing Machine applet, I had no idea how a Turing Machine worked. I thought it was just a random sequence. So I asked some questions on the Facebook Eurorack thread, and Tom Whitwell was kind enough to explain everything in great detail, enabling me to reproduce the behavior pretty faithfully.

Since then, I've wanted to "go big" with the Turing Machine concept, but I wasn't sure what form that would take. With the sixteen-bit register, the Turing Machine provides probably the best memory consumption-per-step ratio in melody generation. It's pretty incredible what you can get, and no wonder they're so popular in so many forms.

My new project, which I started today, is called Enigma. It's a full-app Turing Machine workstation, with which you'll be able to:

* Maintain a library of 32 Turing Machines in memory (four banks of eight), and work on them one at a time. When a good melody (or rhythm) emerges, you'll be able to mark it as a favorite and lock it in place.

* You'll be able to create songs of up to four tracks. Each track will be a chain of Turing Machines, selected in any order from your library. You'll be able to specify how many times each Turing Machine in the chain repeats. Each Turing Machine will evolve as it plays, based on its set probability. But whenever it is re-introduced, it will return to the form it has in the library.

The Song Mode is going to be an homage to the Alesis MMT-8's Song Mode, but with Turing Machines instead of Parts.

* You'll be able to assign tracks to CV outs (A/B/C/D), and you can specify the type of output for each track (quantized pitch, unquantized modulation, trigger (a pulse for each "on" bit), or gate (a gate will span consecutive "on" bits)).

* You'll also be able to assign MIDI out to tracks for pitch, modulation, velocity

* You'll be able to save your TM library and songs via sysex dump, as always

I'm pretty geeked about it. I wish I had it right now.
Zymos
Clever name Mr. Green
chysn
Zymos wrote:
Clever name Mr. Green


Thanks! Seemed a nice tribute to Mr. Turing.

Other candidates were: TuringMasaurus-Rex, Auntie Em
behndy
very cool. can't wait to see it!
chysn
Boofin Moonrocks wrote:
Would hemisphere be able to run on an ornament and crime modified to have a voltage range of 0-9v?


I would think so. But most modulation ranges would wind up being from 3V to 8V, so it would always lie about things. How does the normal O_C software behave with such a modification?

Since all of the Hemisphere I/O goes through a single set of methods, the software could be compiled with offsets in just a few places to compensate for the difference.
Godphaser
Is there a version of hemisphere I could load that has the original apps (minus Meta-Q) and the bow tie app of yours?

Thanks!
chess1
Dang the Enigma applet sounds sweet.

If I was a bad ass I would design a dual o-c mega plate in 20hp with a huge screen and extra stuff.

But im not so ill sit back and wait for my O-C to arrive in the mail.
Dogma
So i think this could now be a totally different module altogether - i mean the orignial o_c isnt even there at all

It would be a damn shame to stop this development because of space restrictions - its 2018 smile
'Would a simple upgrade of the teensy be sufficient?
pld
Hehe, various incarnations of poly-Turings have been on my todo-list for a while, but running on a dedicated thing... and I didn't have a cool name.

Dogma wrote:
So i think this could now be a totally different module altogether - i mean the orignial o_c isnt even there at all

It would be a damn shame to stop this development because of space restrictions - its 2018 smile

Yeahbut the platform is from 2014-ish and the longevity (at least as far as I can tell) somewhat unplanned...

Quote:
Would a simple upgrade of the teensy be sufficient?

Yes. No. Maybe. It depends smile
There are a few potential upgrade paths, mostly it comes down to finding or motivating someone to do it.
It's vaguely possible to MacGyver the 3.6 on there, but it'd still need some work to fixup the drivers. While it also brings a faster processor, it's still hampered by all the other restrictions of the original design (shared SPI between DAC/screen, few physical controls) so to me that doesn't gain much, and it's a bit unclear how to support nice things like the FPU, but not totally cut off backwards compatibility -- at least without creating a logistics nightmare. Plus there are more interesting approaches, like replacing the Teensy with a STM32F4 daughterboard and ditching the Teensy/Arduino crutches. YMMV, especially for the definition of "interesting" razz

One might also make something to stack under the existing Teensy with pogo pins and unlock additional ports (serial MIDI? I2C storage? Additional analog ins?).

For that matter, if sysex is working for people, it should also be possible to code up a way to dynamically load apps... although that plan has a high Baldrick factor.
chysn
pld wrote:
For that matter, if sysex is working for people, it should also be possible to code up a way to dynamically load apps... although that plan has a high Baldrick factor.


You might have a different mechanism in mind; but after I mastered sysex, I imagined a little applet scripting language that would be transpiled into sysex. I realized that I'd be giving up lots of volatile memory for something that, in the end, uses the same cable to get where it needs to go. I don't know if there's a way to store things in program memory during runtime?

The real game-changer would be "programmable over wi-fi."

What does "Baldrick factor" mean?

Dogma wrote:
It would be a damn shame to stop this development because of space restrictions - its 2018


On the other hand, I go have to get back to composing at some point.

I imagine O_C as sort of an "evergreen" module. As long as the parts are available, we work with what it does, and don't worry about what it doesn't do.

Quote:
Is there a version of hemisphere I could load that has the original apps (minus Meta-Q) and the bow tie app of yours?


No, I'm afraid not. There's been enough of a schism that applets for the current Hemisphere framework won't even work with the old one. But, trust me, it's worth making the switch to have a sequential Boss Bow Tie! Yeah, my Tetrapad is an eight-step sequencer now!
Supervillain
Hey
Tested some app on the new firmware and I have some ideas for improvement.

On the enveloppe follower it would be nice to have a release control and maybe the ability to switch each to normal or inverted. If you want to sidechain multiple tracks for exemple.

On 5 step sequencer it seems the cv input isn't tracking 1v/oct like a simple adder would.

Didn't got time to test other apps enough be will try soon!
pld
chysn wrote:
You might have a different mechanism in mind; but after I mastered sysex, I imagined a little applet scripting language that would be transpiled into sysex. I realized that I'd be giving up lots of volatile memory for something that, in the end, uses the same cable to get where it needs to go. I don't know if there's a way to store things in program memory during runtime?

The real game-changer would be "programmable over wi-fi."

Scripting is one way (and one of my on-going projects) but in this case I meant sending compiled code. Or split the difference and use a built-in compiler, a la micropython which can generate assembler internally IIRC.
And yes, it's possible to write to flash from a running program -- I just tend to avoid it on Teensy because (AFAIK) it looks like there's no easy way (like jtag/swd) to unbrick it if things go horribly wrong. They're cheap, but not that cheap grin

Quote:
What does "Baldrick factor" mean?

In that it's a particularly cunning plan with some horrible flaws. See: Blackadder.
chysn
Supervillain wrote:
On the enveloppe follower it would be nice to have a release control and maybe the ability to switch each to normal or inverted. If you want to sidechain multiple tracks for exemple.

On 5 step sequencer it seems the cv input isn't tracking 1v/oct like a simple adder would.


Good suggestion on the envelope follower. The ducking feature was sort of a last-minute tack-on, whereas providing a setting on each channel would be my usual practice. I'll certainly make that update.

As for the transpose... yeah. It wasn't intended to work like an adder, it's just a vague bi-polar up/down value. I can see how your suggestion would be a lot more useful (and expected), so I'm putting that on the list, too. I'll also do that with The Darkest Timeline's transpose, which has the same flaw.

Thanks!

pld wrote:
And yes, it's possible to write to flash from a running program -- I just tend to avoid it on Teensy because (AFAIK) it looks like there's no easy way (like jtag/swd) to unbrick it if things go horribly wrong.... it's a particularly cunning plan with some horrible flaws.


Yikes, yeah. Also, I think I'm too dumb to pull it off. So screw that!

Here's the video of the scale editor, demonstrating Scala file transfer: https://www.youtube.com/watch?v=UKX79rkSdIQ

I mass-converted about 3000 Scala files in the Scala Archive to the HS sysex format, but I'm waiting for permission from Manuel Op de Coul or the Huygens-Fokker Foundation to distribute them. I could find no license information for the archive.
chysn
Note: I updated the hex file to 1.4A because I found a buffer overwrite error resulting from forgetting a length range check on the Scale Editor. A crash could only be caused by a poorly-formatted sysex file, but I thought it was important to release a patch.

Also, the transposition now works as Supervillain suggested.
erstlaub
I think it's been mentioned further up but I would love to see some sort of system for idiots like me that aren't good with github/code to be able to pick and choose what goes into a custom hex package to DL and install.

While I totally appreciate that they are incredibly useful to some, the midi stuff, the boss bow tie and more complex turing machines are totally surplus to me and I'd be far happier having a few original O+C bits and the pick of Hemisphere (and have my cake and eat it obviously).

Maybe it'd make for an interesting project for someone, from an analytical point of view, it should be fairly easy to list the sizes of various components and then via a tick box of something, be able to tell what can be squeezed into the limited space available?

Amazing work, I think I'll probably stick on 1.3 for the time being as I'm really enjoying playing with Game Of Life still.
chysn
erstlaub wrote:
Maybe it'd make for an interesting project for someone, from an analytical point of view, it should be fairly easy to list the sizes of various components and then via a tick box of something, be able to tell what can be squeezed into the limited space available?


I'd be happy to lend moral support and a sympathetic ear, and maybe even some technical guidance, to anyone willing to undertake such a project.

Now that I'm doing my own thing, I've felt at liberty to sort of screw around with the main O_C code here and there, so there might be slight--not insurmountable--issues with simply dropping applets in.

Quote:
While I totally appreciate that they are incredibly useful to some, the midi stuff, the boss bow tie and more complex turing machines are totally surplus to me


Honestly, I don't know if MIDI is useful to anyone at all. For me, it totally changed how I do everything and Hemisphere Suite has become "a MIDI interface that also does some other things." I didn't know this about myself until now, but apparently what I really missed after going modular was saving stuff.

Also, I think everyone should have a Boss Bow Tie. It's like the best module ever made.
starthief
I use MIDI occasionally via CV.OCD. Definitely curious to play with it in Hemisphere once I pick up an O_C again.
rayultine
chysn wrote:

Honestly, I don't know if MIDI is useful to anyone at all. For me, it totally changed how I do everything and Hemisphere Suite has become "a MIDI interface that also does some other things." I didn't know this about myself until now, but apparently what I really missed after going modular was saving stuff.


MIDI<->CV in an o_C would be infinitely useful to me ... if I had the time/skills/energy to carve out of the back of my case to get easy access to the module's USB port. Then I also start thinking about whether or not I have a case that I am going to own forever and do I want to start customizing it.

Oh wait... I might actually have a large enough gap between two modules to keep a USB cable in there permanently? hmmm..... ok gonna do some potentially game changing experiments this week. hah
pld
chysn wrote:
I'd be happy to lend moral support and a sympathetic ear, and maybe even some technical guidance, to anyone willing to undertake such a project.

Ditto.

Given the shared code between apps (and/or whole program optimizations, if enabled) I'm not sure if the resulting code size for a combination is predictable. So it'd seem to either need brute force (compile & see), or preprocessing brute force (compile all combinations and remember results). And the settings storage requirements might have to be parsed out. Or just best-guess if the error margins allow...

...and by all that I of course mean to say it'd be a super fun and amazingly rewarding project! w00t
REVIVER
Really enjoying the updated Turing applet.

Any possibility of squeezing in a pre-quantize attenuator a la the Disting Quantized Shift Register? Maybe just even on the A/C output?

I see that there's a tiny, tiny bit of screen space to the right of the scale...
chysn
rayultine wrote:
MIDI<->CV in an o_C would be infinitely useful to me ... if I had the time/skills/energy to carve out of the back of my case to get easy access to the module's USB port. Then I also start thinking about whether or not I have a case that I am going to own forever and do I want to start customizing it


I had the same reservations. I built my own case, so there wasn't really any kind of risk, but it's still always an uneasy feeling to whip out the drill. A set of spade bits makes it easy. And, I'd highly recommend this Switchcraft thing:

https://www.amazon.com/gp/product/B00DJ4Q6FW

There's no soldering here. The back of this is just a regular USB jack, and you just use a regular USB cable.

REVIVER wrote:
Any possibility of squeezing in a pre-quantize attenuator a la the Disting Quantized Shift Register? Maybe just even on the A/C output?


I'll keep that in mind. I'm totally rethinking the Turing applet in light of the Enigma app. The reason is, I'm taking the global setting memory that O_C reserved for user chords, and claiming it for user Turing Machines instead *. So even though Enigma is a full app, I'll be able to access its Turing Machines from a Hemisphere Applet. The idea here is to convert Turing Machines from these disposable ephemeral things into durable little objects that let you say, "Oh, I like that, let's keep it for later." So there will be a mini-Enigma applet either in addition to, or instead of, the current Turing. Probably in addition to.

* Turns out there's room for 40 of them without upsetting other apple carts.
REVIVER
chysn wrote:
I'll keep that in mind. I'm totally rethinking the Turing applet in light of the Enigma app.


Yeah, I hesitated about making that suggestion after reading your concept of the Enigma app which I like very, very much.

Great work all around.

screaming goo yo
Dogma
Alright this is a long shot but i think may be doable - is there provision for select bus saving, managing presets? Then we would be able to save Mungo presets! And all the others too but Mungo!
chysn
Dogma wrote:
Alright this is a long shot but i think may be doable - is there provision for select bus saving, managing presets? Then we would be able to save Mungo presets! And all the others too but Mungo!


Probably not right now, because I don't understand a word of that. But I'm interested in learning. What are Mungo presets?
pld
chysn wrote:
Dogma wrote:
Alright this is a long shot but i think may be doable - is there provision for select bus saving, managing presets? Then we would be able to save Mungo presets! And all the others too but Mungo!


Probably not right now, because I don't understand a word of that. But I'm interested in learning. What are Mungo presets?

There's a kind-of-standard for modules to use the underused gate/cv pins on busboards as a data bus to do preset save/recall: See Storage Strip, Varigate?, Mungo, and I believe Disting? One device plays master to send commands/clock and there can be multiple slaves.

That would be a case for some board below the Teensy, a) to get access to some serial TX/RX pins and b) be able to add the necessary opamps to drive the signals. The functionality could then be "fairly easy" to implement because the preset protocol is really just MIDI commands, and the actual save/recall is handled by the slave devices. Personally, if I had any skin in the game (i.e. modules that supported it) I'd probably even make something that would translate TT I2C to preset bus, but really I never looked to closely because it'd still be an almost clone of someone else's module.
sushiluv
thanks for midi out in darkest timeline, i love it. one thing i noticed, the alternate output has no probability, the sequence plays just through, is this by design or maybe a bug?

Bow tie, i don´t get this app, could this be used without the bow tie module in the case? what would i need to get a use out of it.

Could we eventually have an applet for switching octaves and transpose by semitones? having two inputs and outputs it would be great to switch octaves or transpose two oscillators independently.

thanks! smile
chysn
pld wrote:
There's a kind-of-standard for modules to use the underused gate/cv pins on busboards as a data bus to do preset save/recall


OK, thanks. I don't have any of those modules, so I'd be hard-pressed to test such a thing even if the hardware were available.

sushiluv wrote:
thanks for midi out in darkest timeline, i love it. one thing i noticed, the alternate output has no probability, the sequence plays just through, is this by design or maybe a bug?


It is a bug, thanks for finding it! I'll put out a corrected hex file later today.

Quote:
Bow tie, i don´t get this app, could this be used without the bow tie module in the case? what would i need to get a use out of it.


ALM's Bow Bow Tie module is a CV-controlled switch in 6HP. With the offset set to 0, and polarity fully positive, the CV required to select I/O 1-8 are .65V apart. With the applet, whenever a clock comes into Clock 1, the output increases by .65V to a voltage smack-dab in the middle of the Boss Bow Tie's range for the next step, until I/O 8, at which point it goes back to 0V. The purpose is to enhance Boss Bow Tie's utility by turning it into a sequential switch, like an eight-I/O Doepfer A-150.

Can you use it without Boss Bow Tie? Sure. It's really just a fixed-voltage 8-step sequencer that increases by about 2/3 of an octave per step. It could be used as a sort of counter in conjunction with a comparator. At first glance, this application seems like nothing more than a clock divider, but except a clock divider doesn't tell you how close it is to its next step.

Now that I think about it, I should have just called it "Counter," and people would find ways to use it smile

Quote:
Could we eventually have an applet for switching octaves and transpose by semitones? having two inputs and outputs it would be great to switch octaves or transpose two oscillators independently.


It would have to be quantized voltage, or it would be a little off; so basically a third quantizer, maybe one with a shared scale and root but separate transpose settings per channel. That would be nice, I think.
bemushroomed
Great work with this smile

I've just installed Hemisphere Suite 1.4A. However long-pressing left encoder (or right encoder, I tried both) does not bring up any help screen for me?
chysn
bemushroomed wrote:
Great work with this smile

I've just installed Hemisphere Suite 1.4A. However long-pressing left encoder (or right encoder, I tried both) does not bring up any help screen for me?


Thanks for trying it!

Please see here for general instructions: https://github.com/Chysn/O_C-HemisphereSuite/wiki/_Hemisphere-Basic-In structions

Specifically, here: https://www.youtube.com/watch?v=3I734jXkrvA&feature=youtu.be
bemushroomed
chysn wrote:

Thanks for trying it!

Please see here for general instructions: https://github.com/Chysn/O_C-HemisphereSuite/wiki/_Hemisphere-Basic-In structions

Specifically, here: https://www.youtube.com/watch?v=3I734jXkrvA&feature=youtu.be


Awesome, thanks.

Really love Neuron.. goes so well with my NoiseEngineering Manis Iteritas we're not worthy
chysn
Manuel Op de Coul granted permission for me to distribute the Scala Scale Archive as sysex files, so here are about 3000 scales that you can try!

http://beigemaze.com/downloads/HS_Scala_Archive.zip

This includes all of the scales in the Archive that are compatible with the Ornament and Crime scale structure (that is, between 4 and 16 notes, within a one-octave range).
Maximars
@ sushiluv
You can use the Voltage App for this purpose, but you need a external precision adder or some VCOs have two pitch cv inputs.

@ chysn

I think the Voltage App wold be more usefull if the grid of the encoder wold be exact one semitone per step and for fine tune maybe Push and Turn the encoder.
chysn
Maximars wrote:
I think the Voltage App wold be more usefull if the grid of the encoder wold be exact one semitone per step and for fine tune maybe Push and Turn the encoder.


Yeah, that makes sense. I imagined some use cases where people would want to dial in more precise voltages for specific parameters, but I guess that's what VCAs are for.
Maximars
I noticed something else. If I feed a Pattern into the Clock Skipper App, so an irregular pattern and turn up the probability 100% i get not the same patter as the one i feed in. Is this app just for regular clocks? should that not be like the Brancher without switching the outputs? seriously, i just don't get it
I end up using the Brancher and losing one channel sad banana
chysn
Maximars wrote:
I noticed something else. If I feed a Pattern into the Clock Skipper App, so an irregular pattern and turn up the probability 100% i get not the same patter as the one i feed in. Is this app just for regular clocks?


No, it should work with any clock signal, regular or not.

Just to clarify, you're saying that the probability is at 100%, and it still skips clocks? How often would you say this happens?

Also, just make sure that you don't have any negative CV going in that might be reducing the probability.
Maximars
No negativ CV applied.

Example, I feed in somting like

Taaa Ti Ti Taaa Ti Ti Taaa ...

And get

Taaa Ti Taaa Ti Taaa Ti Taaa Ti...

No Skipping, just a new Pattern. Very strange ...

Is the Gate or Trigger lenght reflectet to the output?
Supervillain
Noticed something strange in the dual quantizer app : when I feed a already quantized voltage, from a keyboard sequencer for exemple, and set quantization to semitone. I get a different sequence. Is it normal?
chysn
Maximars wrote:
No negativ CV applied.

Example, I feed in somting like

Taaa Ti Ti Taaa Ti Ti Taaa ...

And get

Taaa Ti Taaa Ti Taaa Ti Taaa Ti...

No Skipping, just a new Pattern. Very strange ...

Is the Gate or Trigger lenght reflectet to the output?


I'll check it out with some different patterns tonight. Note that Clock Skipper only responds to, and sends, triggers. The length of its trigger outputs will always be around 4ms, regardless of what comes in.
chysn
Supervillain wrote:
Noticed something strange in the dual quantizer app : when I feed a already quantized voltage, from a keyboard sequencer for exemple, and set quantization to semitone. I get a different sequence. Is it normal?


I wouldn't think it should be way off. Do you happen to have a meter handy?
mxmxmx
pld wrote:

Dogma wrote:
Would a simple upgrade of the teensy be sufficient?

Yes. No. Maybe. It depends :)


let's see ... if this happens, it should be physically compatible at least (even pin-compatible).
chysn
chysn wrote:
Maximars wrote:
No negativ CV applied.

Example, I feed in somting like

Taaa Ti Ti Taaa Ti Ti Taaa ...

And get

Taaa Ti Taaa Ti Taaa Ti Taaa Ti...

No Skipping, just a new Pattern. Very strange ...

Is the Gate or Trigger lenght reflectet to the output?


I'll check it out with some different patterns tonight. Note that Clock Skipper only responds to, and sends, triggers. The length of its trigger outputs will always be around 4ms, regardless of what comes in.


I've been playing around with it for about ten minutes at 100%, and it's working as expected. I'm sending sequenced patterns from a Varigate 4+ into a DSM03 Trigger In. If I set pulse width to its highest value, it does change the rhythm, but that's because there's no release between consecutive pulses at 100%.

I'd say, check your pulse width. It's a clock skipper; the pulses need to be triggers, meaning under maybe 10ms. If you're still having some problems, please tell me a little about your clock source.
chysn
Supervillain wrote:
Noticed something strange in the dual quantizer app : when I feed a already quantized voltage, from a keyboard sequencer for exemple, and set quantization to semitone. I get a different sequence. Is it normal?


Well, I'm striking out on observing phenomena today! I put quantized voltage from my Tetrapad into the Dual Quantizer and passed its output to Scope, and it read the correct voltage over about 5 octaves. So if this is still an issue, I'd like to learn more about your voltage source.

I remember Douglas Hofstadter writing "twice enlightened is unenlightened." I don't think the same is true of twice quantized.
Supervillain
I tested it back again too and I was wrong. In fact, the issue is when I feed a C minor melody in the quantizer set on aeolian scale (which is minor, right?)
chysn
Supervillain wrote:
I tested it back again too and I was wrong. In fact, the issue is when I feed a C minor melody in the quantizer set on aeolian scale (which is minor, right?)


OK, good to hear!

There's a 1.4B maintenance release at https://github.com/Chysn/O_C-HemisphereSuite/releases

This fixes the Parallel Universe probability calculation for The Darkest Timeline, sets the Voltage increments at one semitone, addresses some MIDI issues, and a range problem with CV Recorder.
Zymos
This whole thing is amazing. We have pages and pages of "can it do this?" ..."OK, added that."

Unlimited props to Chysn It's peanut butter jelly time! nanners Rockin' Banana!
rayultine
Zymos wrote:
This whole thing is amazing. We have pages and pages of "can it do this?" ..."OK, added that."

Unlimited props to Chysn It's peanut butter jelly time! nanners Rockin' Banana!


Chysn,
Can you program o_C to do all of the biochemical modeling automation that I've been working on for the last 8 years?

twisted
chess1
I am getting an OC this week to enjoy.

I have a feature request (forgive me if this already exists and I missed it)

Would it be possible to have sysex template for the Yamaha Tx8z that allowed the user to select cv destination of params OR even a radomize location feature that could be sequenced somehow. Similar in spirit to the steropping controllers but with cv input!


Chysn, you are the man! Miley Cyrus
myecholalia
Quick question: I already have a uO_c running Hemisphere in my little eurorack system. But I'm also running a slightly larger Buchla system for which I'm thinking about getting the Northern Light Modular Dual CV Polymorpher
Model 2OC 1.5 (essentially 2 x O_c behind a Buchla panel with additional attenuators and obviously using Buchla voltage standards.

I don't think that Hemisphere would work out of the box, or would it? If there are any changes to be done, would it be possible to put those on the roadmap? Cheers!
DeltaTucanae
myecholalia wrote:
Quick question: I already have a uO_c running Hemisphere in my little eurorack system. But I'm also running a slightly larger Buchla system for which I'm thinking about getting the Northern Light Modular Dual CV Polymorpher
Model 2OC 1.5 (essentially 2 x O_c behind a Buchla panel with additional attenuators and obviously using Buchla voltage standards.

I don't think that Hemisphere would work out of the box, or would it? If there are any changes to be done, would it be possible to put those on the roadmap? Cheers!


https://www.instagram.com/p/BlXGGrknD7J/?taken-by=kuro303909
mxmxmx
myecholalia wrote:

I don't think that Hemisphere would work out of the box, or would it? If there are any changes to be done, would it be possible to put those on the roadmap? Cheers!


the Buchla-specifics happen in hardware (gain (1.2V/oct) / offset), so it should _mostly_ work out of the box.

much of the rest will be cosmetics, ie display related, except the value for "kOctaveZero", which is a/the index into the DAC code for 0V (different for euro / Buchla). i'm not sure how hemisphere handles the min values, but IIRC i spent some time making sure the various OC apps don't go below 0V when compiled for 2OC. (the 2OC hardware can go below 0V (somewhat), just so to make sure there's no offset, but apparently some $uchla modules aren't properly input protected, so that might be an issue, or not)

anyways, the essential bits are still there, it looks like (see OC_options.h: #define BUCHLA_4U )
chysn
DeltaTucanae wrote:
myecholalia wrote:
Quick question: I already have a uO_c running Hemisphere in my little eurorack system. But I'm also running a slightly larger Buchla system for which I'm thinking about getting the Northern Light Modular Dual CV Polymorpher
Model 2OC 1.5 (essentially 2 x O_c behind a Buchla panel with additional attenuators and obviously using Buchla voltage standards.

I don't think that Hemisphere would work out of the box, or would it? If there are any changes to be done, would it be possible to put those on the roadmap? Cheers!


https://www.instagram.com/p/BlXGGrknD7J/?taken-by=kuro303909


Holy shit, man. I can die happy now.

I was going to say: I don't have any way to test it, but, as mxmxmx said, the more low-level systems are where the Buchla support happens, and I haven't modified those. Obviously, there might be cases where I inadvertently disrupted something, with no way to catch it, so proceed with the usual caution.

Note that wrote my own quantizer for MIDI notes, which probably won't work. I'd be willing to take a shot at updating it, though, if you're planning to use MIDI.
chysn
chess1 wrote:
I have a feature request (forgive me if this already exists and I missed it)

Would it be possible to have sysex template for the Yamaha Tx8z that allowed the user to select cv destination of params OR even a radomize location feature that could be sequenced somehow. Similar in spirit to the steropping controllers but with cv input!


That's a fascinating one. I've had a grand total of three TX81zs in my life, two at once back in the 90s. Honestly, I think it's a bit narrow as a general app. But you know, if you buy me a TX81z to test it on, I'll do it!

Geez... I should also mention that the same offer goes for the Buchla folks! nanners
mxmxmx
chysn wrote:

I was going to say: I don't have any way to test it, but, as mxmxmx said, the more low-level systems are where the Buchla support happens, and I haven't modified those. Obviously, there might be cases where I inadvertently disrupted something, with no way to catch it, so proceed with the usual caution.


the "support" is mostly about menu adjustments, so that's handled on the app-level. also most of the 0V stuff. Buchla was more than an afterthought (nor ours).

i can't remember what the 2OC full range is (maybe something like -0.5V to 12.5V ...? ); and I can't tell how dangerous this is. fwiw, the potentially critical bits is when bypassing the calibration stuff when writing to the DAC, in which case you can end up with values below -3V (euro) resp. below 0V (2OC/Buchla) (= below OC::calibration_data.dac.calibrated_octaves[CHANNEL_NUM][OC::DAC::kOct aveZero])). typically that would be the case for LFOs, envelopes, etc.

it looks as if you use OC::DAC::set_pitch pretty much through-out though (rather than OC::DAC::set), ie

void Out(int ch, int value, int octave = 0) {
OC::DAC::set_pitch((DAC_CHANNEL)ch, value, octave);
outputs[ch] = value + (octave * (12 << 7));
}

this will actually make sure you're not going < 0, but in some cases will introduce some unwanted quantization, if i see it right (= not using the entire 16bit). i *think*. i've only had a quick glance at the code during lunch...
chysn
Yeah, I think your assessment is right. I'm using only pitch_set(), which observes kOctaveZero. So I think the bottom line is: Is it safe? Yes. Will it always work right? Dunno.

Thanks for the input!
Maximars
@chysn
Clock Skipper works. My Trigger Signal was bad. had two triggers very narrow to each other from a sequential switch. I'm sorry for that :-(

But I have a wish. Is it possible to have an option that the output reflects the length of the input gate like Brancher does? It would be so much more useful hyper
I constantly modulate my gate length with the a-142. nanners

Found a repeatable Bug which freezes the Module (1.4.B)

-Left side:
LowerRenz

-Right side:
LoFi Tape (for example)

-Save State

-Power down

-Power Up

-Select App Selection on the right side

-Turn right encoder CCW <- Important!
when You get to the LowerRenz app it freezes! Dead Banana
chysn
Maximars wrote:
But I have a wish. Is it possible to have an option that the output reflects the length of the input gate like Brancher does? It would be so much more useful hyper


It might not need to be an "option." It'll probably continue to work as a clock skipper if I make it a gate skipper. I'll check it out.

Quote:
Found a repeatable Bug which freezes the Module (1.4.B)


Interesting. LowerRenz has a Highlander complex. There can be only one. I'll look into that.
chysn
Maximars wrote:

when You get to the LowerRenz app it freezes! Dead Banana


I think I was just pushing the processor way too hard, asking it to process two independent Streams Lorenz generators at once. So I added something called a "Singleton" object that manages a single Lorenz generator between both hemispheres. The Lorenz generator is really a dual Lorenz generator, anyway, so it made perfect sense to just share it. I've been playing around with it for a while, with abusive modulation, and it seems to be behaving okay.

I've committed the fixes to GitHub, but I haven't created a hexfile yet. I'll probably do that later in the week.
myecholalia
chysn wrote:
Yeah, I think your assessment is right. I'm using only pitch_set(), which observes kOctaveZero. So I think the bottom line is: Is it safe? Yes. Will it always work right? Dunno.

Thanks for the input!


Had been in touch with Northern Lights to briefly discuss Hemisphere as they are putting out a new version of the Buchla o_C. If you like (and I know it's probably very low on your list of priorities), I could link you up with them for a small exchange just to be sure. Happy to then be the guinea pig to test if everything works properly once the new module is out and I've had a chance to put Hemisphere on it.
esquilofreniko
Amazing app (specially in terms of UI/playability)! Has made my O_C much more fun than it used to be smile
Sugestion: would be amazing to have a range parameter (attenuator) on the turing machine applet smile
Thanks for the great work!
Maximars
The strange thing is, that the second LowerRenz just freezes when selected counterclockwise after a power-up. Clockwise - no problem.
If you select it first clockwise and then counterclockwise also no problem.

It seems to be an initialization problem after rebooting.
chysn
myecholalia wrote:
If you like (and I know it's probably very low on your list of priorities), I could link you up with them for a small exchange just to be sure. Happy to then be the guinea pig to test if everything works properly once the new module is out and I've had a chance to put Hemisphere on it.


Sure, sounds good.

esquilofreniko wrote:
Amazing app (specially in terms of UI/playability)! Has made my O_C much more fun than it used to be smile


Thanks for trying it!

Maximars wrote:
The strange thing is, that the second LowerRenz just freezes when selected counterclockwise after a power-up. Clockwise - no problem. If you select it first clockwise and then counterclockwise also no problem.


I couldn't reproduce it in exactly that way. It was mostly via selecting LowerRenz in both hemispheres and playing around with them for between two seconds and thirty seconds. But I haven't been able to crash it since moving over to a single Lorenz instance.
pld
chysn wrote:
I think I was just pushing the processor way too hard,

Part of the debug menu should show the time being used by the core interrupt (min/avg/max/%). Except of course you can't see it if the current app is already permanently using > 100% smile
rayultine
can any of the Hemispheres applets do CV rectification?
chysn
pld wrote:
chysn wrote:
I think I was just pushing the processor way too hard,

Part of the debug menu should show the time being used by the core interrupt (min/avg/max/%). Except of course you can't see it if the current app is already permanently using > 100% smile


Woah! That's going to be super-useful. Thanks for the protip!

In the old (current) version, when two independent Lorenz generators were running, the percentage was in the 90s and then it crashed. With the singleton instance, it stays in the 50s whether you're running one LowerRenz or two.

rayultine wrote:
can any of the Hemispheres applets do CV rectification?


Some of them rectify CV internally for some reason or other, but none of them do it as a function. O_C's I/O is asymmetrical, so it would be potentially confusing. That's an area where Kinks is best.
rayultine
I think I asked you a similar question like a week ago. Sigh. My case is almost full.
SavageMessiah
This turned my uO_C from the thing with the crappy UI that I reluctantly reached for when I needed a quantizer into the thing I reach for all the time for any number of utilities. It's basically obsoleted a couple of other modules now. Thanks and job well done! thumbs up
chess1
Question about the firmware upgrade process

I got th teensy loader running on my windows 7 laptop.

Downloweded and put the folder of the firmware on may desktop.

When I go to open the hex file with the Teensy loader all I get is empty folders...it is not seeing the hex files.

Even if it did see the hex files does the process involve uploading each applet's hex one by one ?

Appreciate the help.
chysn
chess1 wrote:
Question about the firmware upgrade process

I got th teensy loader running on my windows 7 laptop.

Downloweded and put the folder of the firmware on may desktop.

When I go to open the hex file with the Teensy loader all I get is empty folders...it is not seeing the hex files.

Even if it did see the hex files does the process involve uploading each applet's hex one by one ?

Appreciate the help.


Hey!

There's just one hex file. Go to this page:

https://github.com/Chysn/O_C-HemisphereSuite/releases

And click on the file named HemisphereSuite1_4.hex and save that on your computer somewhere. That's the file that you open with Teensy Loader.

Meanwhile, on another topic, I see you have a chess computer as your avatar. Do you have a strong interest in chess computers? I've got a pretty crazy one...
chess1
thanks bud, I am finally all sorted. I dont know why I struggle so much to get the right pages in github.

As for chess computers, it is a love/hate. I am rated 2365 uscf and probably would never have reached this level without learning from them. But then again the competition....

We should continue the talk in the Chess thread!
chess1
Wait...hold...up

Are you sitting on a integrating music chess playing AI interface that will be dropped in a later release ??


Dead Banana Dead Banana Dead Banana Rockin' Banana!
chysn
chess1 wrote:
Wait...hold...up

Are you sitting on a integrating music chess playing AI interface that will be dropped in a later release ??


It should come as no surprise that it's crossed my mind. When I was in college, I developed a method to convert chess games into musical notation for a class assignment, using a programming language for Amiga called AREXX. And I sometimes think of ways to select from lists of moves via CV. And a 64x64 pixel half-screen is perfect to render a chessboard with little 8x8 pixel pieces. Who knows whether I'll actually get around to doing it, but it's something I think about regularly.
chess1
How about a mini usb to vga (or usb to computer screen??) stream. It would be sick to use some of these wallpapers as a light show for Aesthetics.

Or a blown up screen of this rumored chess playing app! (my high score in pong is 6 w00t )
chess1
PGN (portable game notation) to pitch converter as an easter egg!

meanwhile I keep lusting for this dongle and dreaming of improving my pong score...


Shledge
I don't think that would be feasible without a considerable rewrite of the core o_C firmware, or some sort of adapter to the screen pins.
chysn
chess1 wrote:
meanwhile I keep lusting for this dongle and dreaming of improving my pong score...


Well, I can help you with one of those.

As with most Eurorack problems, it's Maths to the rescue. Patch O_C output C to Maths Ch. 2 input, O_C output D to Maths Ch. 3 input. Maths Sum output to O_C CV input 1. Set Maths Ch. 2 knob at about the 9:00 position, and the Ch. 3 knob at about the 3:00 position.

Output C is the ball's Y position, and Output D is the paddle's Y position. CV Input 1 moves the paddle. So you're basically setting up a comparator to output positive or negative voltage depending on where the paddle needs to be.

You can improve the score by adjusting the knobs slightly. You can rack up scores into the 20s pretty easily (highest I've seen is 27), but at some point the level is too fast for a Maths-based approach to keep up.

A more advanced strategy would be listening to Output A and B pings to anticipate the future position of the ball, eliminating the need to simply follow it, including a slow synced LFO to adapt to the speed/distance changes. I could do that with a second O_C, but it's more of a thought experiment....
rayultine
chysn:
I dunno if you have ever used Reaktor, but the Pong app might get super weird and cool if you added some of the features of this Reaktor ensemble:

https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/3163/

By changing paddle width/height, paddle spacing, box size, and ball size, you can make some very strange, cool, and hilarious sounds/patterns.
chysn
rayultine wrote:
chysn:
I dunno if you have ever used Reaktor, but the Pong app might get super weird and cool if you added some of the features of this Reaktor ensemble:

https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/3163/

By changing paddle width/height, paddle spacing, box size, and ball size, you can make some very strange, cool, and hilarious sounds/patterns.


There's nothing new under the sun, I guess!
Attackman
Installed the latest firmware release the other night. Just loving Hemisphere more and more as I explore it. The Hemisphere o_C is now the module that's in every patch. And so excited to see the Enigma implementation of the Turing as that will easily replace Quantermaine (which I still miss) and seal the deal on Hemisphere as a keeper.

I'm wondering if The Darkest Timeline or Captain Midi could be used to send midi to a Blofeld (or any other midi device) via a midi DIN?

Presumably I'd need an adapter or DIY interface to get from the usb interface to a DIN - any suggestions on what that might be and things I might have to consider?

Of course if I get that working I'll be requesting polyphony across simultaneous channel outs.

I don't have a midi sequencer and for the most part everything in my system is CV. Downside is that I can only play the Blofeld via a keyboard and I don't have much interest in keys. I bought Arturia's BSP but the usb snapped off and after numerous fumbled attempts I failed to fix it. (So frustrating. Built like a tank and then this stupid little frail socket barely strong enough to be there. Uh?)

It'd be ACE to employ Hemisphere on boxes outside of the modular!
chysn
Attackman wrote:
I'm wondering if The Darkest Timeline or Captain Midi could be used to send midi to a Blofeld (or any other midi device) via a midi DIN?


The Teensy can't act as a MIDI host, so you'd need something to function in that capacity. I haven't tried this, but I'd love to know whether or not it works:

https://www.amazon.com/USB-MIDI-Host-Module-controllers/dp/B0752M869X

It's the sort of thing that should "theoretically" work, so if you give it a shot, please report back!

Quote:
Of course if I get that working I'll be requesting polyphony across simultaneous channel outs.


You don't have to request it, it's already done. Captain MIDI handles up to four notes of polyphony quite well. If you use all four O_C outs for notes, you need to get your gate elsewhere, of course.

Quote:
I don't have a midi sequencer and for the most part everything in my system is CV. Downside is that I can only play the Blofeld via a keyboard and I don't have much interest in keys.


I really like Captain MIDI's MIDI out capabilities. Using my Tetrapad to play Wavestation plug-in patches via Ableton is a blast. Don't overlook the Legato output type; it makes playing via Eurorack controllers feel very natural.
trumman
just received one of these this week, can confirm that it works well with captain midi for midi out

http://www.hobbytronics.co.uk/usb-midi-converter
chysn
trumman wrote:
just received one of these this week, can confirm that it works well with captain midi for midi out

http://www.hobbytronics.co.uk/usb-midi-converter


Thanks for the confirmation!
chysn
I've been thinking about this... I don't use screensavers for anything. Basically, the only difference between a screensaver and a regular screen is that the header isn't present, and you can't interact with buttons or encoders. To protect the screen, inasmuch as it needs any protection at all, I blank it after 30 minutes.

So, in 1.5, I'm going to eject the concept of the screensaver in Hemisphere Suite completely. All my apps will stay active until the 30-minute screen-blank.

Good news, everyone! This frees the long-pressed Up button. I might universally assign it to MIDI Data Dump, which is something all my apps do. That would be a convenience I'd welcome. But the long-pressed Right button already does that, as a side-effect of taking you to the Main Menu. It might be nice to have a totally separate control for a MIDI dump, one that doesn't interrupt your workflow as much.

It could also be valuable as another general control. I will ponder it carefully. Any ideas?
ipassenger
Nothing to add here other than installed hemisphere suite for first time today and it seems awesome! Thanks smile
chysn
ipassenger wrote:
Nothing to add here other than installed hemisphere suite for first time today and it seems awesome! Thanks smile


I appreciate you giving it a go!

me wrote:
This frees the long-pressed Up button... I will ponder it carefully.


My pondering has led me to the boring conclusion that the long-press Up needs to blank the screen. The 1.5 changes will be:

* The screensaver option in calibration will be changed from "Screensaver Timeout (s)" to "Screen Blank (minutes)"
* After that many minutes of inactivity, the screen goes completely dark. The app keeps working, etc.
* Long-pressing the Up button darkens screen immediately

Yeah, not so exciting of a change, but the screensaver behavior annoyed me even after a minute, and I think it'll be nice to be able to darken your O_C if you're planning to step away for a few.
thebrotherspus
I'd love to try out hemisphere. sadly, I get the hex file too large warning. I had a look at the other instructions, but they looked too daunting for me. waah
edit: never mind. I tried a different cable and it worked out fine.
looking forward to trying this out tonight : )
chysn
thebrotherspus wrote:
I'd love to try out hemisphere. sadly, I get the hex file too large warning. I had a look at the other instructions, but they looked too daunting for me. waah


Where are you getting the hex file?
thebrotherspus
chysn wrote:
thebrotherspus wrote:
I'd love to try out hemisphere. sadly, I get the hex file too large warning. I had a look at the other instructions, but they looked too daunting for me. waah


Where are you getting the hex file?


I downloaded 1_4c.hex from github.com/Chysn/O_C-HemisphereSuite/releases
I tried again with a better cable and it seemed to work now.
I guess now it's time to figure out how hemisphere works.... I'm assuming I'll probably be back later with many more questions hihi
thanks for all your hard work!
thebrotherspus
I'm really enjoying this so far!
I have my first question. I'm using two of the shift register apps with the piston Honda mk3 and it's really fun since the piston is a dual oscillator so each shift register is sequencing a different oscillator with different sequence lengths and probabilities. the problem I'm having is that the piston Honda only takes positive voltages in its v/oct inputs and it seems like the outputs of the shift registers are bipolar since roughly half of the notes in my sequences end up being the same note, i.e. when ever the o_c outputs a negative voltage.

anyway, I'm just curious if there is a way to switch the shift register app (or any of the pitch related apps for that matter) to output only positive voltages for those instances when we're working with oscillators that don't accept negative pitch cv.

thanks so much for your time!!
chess1
Ok I will take a crack at a feature request for long press up button: Randomization of value or reset of values depending on the app. I do not know if there is enough memory left to do this kinda of thing but just a thought.
chysn
thebrotherspus wrote:
I'm really enjoying this so far!


Glad to hear it!

Quote:
anyway, I'm just curious if there is a way to switch the shift register app (or any of the pitch related apps for that matter) to output only positive voltages for those instances when we're working with oscillators that don't accept negative pitch cv.


With respect to pitch, I've been working on standardizing pitches to the following throughout the system:

A4=Tuning reference (default of 440Hz) and MIDI note #69
C4=Middle C, MIDI note #60
C4=0V

This standard maximizes the effective range of MIDI note numbers (that is, #120, or C9, can be reached at +5V).

The Turing Machine (or, as it's now called, ShiftReg) has a melody output with a 5-bit resolution (or, a 2.5-octave range), which is shifted up 48 semitones at the quantizer. So it can go about 1V negative. Same story with Sequence5.

I could shift these to 60 semitones; however, it won't change the more general fact that you'll see negative voltages throughout the system. It's sort of baked in, due--in part--to the Braids quantizer that's used, in part to the deepening integration of MIDI, in part to the fact that O_C is inherently bi-polar. It would be really hard to stamp these out.

This means, your best bet might be putting a precision offset generator in front of your V/oct inputs. Erica's Pico MScale seems like a good companion for 0-10V oscillators, since it can take on or off exactly 5V. Shades is also pretty high-precision, if you want something that's more generally useful.
mazzyboy
chysn wrote:
Here's a spoiler for upcoming development, scheduled for release near the end of September.

When I started planning the Turing Machine applet, I had no idea how a Turing Machine worked. I thought it was just a random sequence. So I asked some questions on the Facebook Eurorack thread, and Tom Whitwell was kind enough to explain everything in great detail, enabling me to reproduce the behavior pretty faithfully.

Since then, I've wanted to "go big" with the Turing Machine concept, but I wasn't sure what form that would take. With the sixteen-bit register, the Turing Machine provides probably the best memory consumption-per-step ratio in melody generation. It's pretty incredible what you can get, and no wonder they're so popular in so many forms.

My new project, which I started today, is called Enigma. It's a full-app Turing Machine workstation, with which you'll be able to:

* Maintain a library of 32 Turing Machines in memory (four banks of eight), and work on them one at a time. When a good melody (or rhythm) emerges, you'll be able to mark it as a favorite and lock it in place.

* You'll be able to create songs of up to four tracks. Each track will be a chain of Turing Machines, selected in any order from your library. You'll be able to specify how many times each Turing Machine in the chain repeats. Each Turing Machine will evolve as it plays, based on its set probability. But whenever it is re-introduced, it will return to the form it has in the library.

The Song Mode is going to be an homage to the Alesis MMT-8's Song Mode, but with Turing Machines instead of Parts.

* You'll be able to assign tracks to CV outs (A/B/C/D), and you can specify the type of output for each track (quantized pitch, unquantized modulation, trigger (a pulse for each "on" bit), or gate (a gate will span consecutive "on" bits)).

* You'll also be able to assign MIDI out to tracks for pitch, modulation, velocity

* You'll be able to save your TM library and songs via sysex dump, as always

I'm pretty geeked about it. I wish I had it right now.


This sounds absolutely fantastic! I often find that there is a lack of repeatable randomness in eurorack - it's usually one or the other. The ability to save 'seeds', mutate them and - most importantly - be able to return to them after a power cycle is very exciting!

One suggestion/request: I often find that randomly generated melodies need to be shifted a step or two to the right or left in order to sit right beside the rest of a patch. I usually use an oddly placed reset signal to take care of this. Would it be possible to include a shift pattern right/left command, so that everything can be saved 'in sync' and work from the same rest signal?

PS - Love your work!
jonkers303
1000 thanks chysn! Hemisphere is beyond awesome!

While my vanilla O&C was amazing its complexity made me not use it very much, i was even on the verge of selling it.
But with the simpleness of hemispheres applets it now gets a ton of use and gives me lots of inspiration.

Hemisphere actually just made me order two micro O&C to replace my full size one since they seem perfect for hemisphere and I quickly use up the two slots.

The help screens are genius and just enough info once you learned a specific applet.

Played around with The Darkest Timeline as well and it seems really cool, need to experiment more with that one.

I think i found a small "bug" in latest HemisphereSuite1_4C :
In Annular Fusion Euclidean Drummer the outputs for left applet is A and B and for right applet its C and D.
No really a problem but just wanted to let you know.

Keep on keeping on!
jonkers303
Eeer never mind my so called "bug" , it behaves as documented

Its me thats confused apparently
thebrotherspus
chysn wrote:
...I could shift these to 60 semitones; however, it won't change the more general fact that you'll see negative voltages throughout the system. It's sort of baked in, due--in part--to the Braids quantizer that's used, in part to the deepening integration of MIDI, in part to the fact that O_C is inherently bi-polar. It would be really hard to stamp these out.

This means, your best bet might be putting a precision offset generator in front of your V/oct inputs. Erica's Pico MScale seems like a good companion for 0-10V oscillators, since it can take on or off exactly 5V. Shades is also pretty high-precision, if you want something that's more generally useful.


dang... that's too bad.
I understand the hesitation of messing around with stuff when it already works well under most circumstances.
It's too bad for me, as my system is super small- I only have one precision adder (contained in Disting mk4) and no room for anymore. I was hoping it would be possible as an option in a "main menu" type of thing to switch only the pitch related cv to positive voltages. Other bipolar voltages in hemisphere are fine since everything else in my system (including piston honda's other cv inputs, excluding v/oct) can accept bipolar.
I understand, however, if this is simply not feasible or just something you're not interested in.
Thanks for your time.
chysn
mazzyboy wrote:
One suggestion/request: I often find that randomly generated melodies need to be shifted a step or two to the right or left in order to sit right beside the rest of a patch. I usually use an oddly placed reset signal to take care of this. Would it be possible to include a shift pattern right/left command, so that everything can be saved 'in sync' and work from the same rest signal?


This is already part of the Library mode. Each stored Turing Machine has only two pieces of information: the register and the length. But the Library mode lets you use an encoder to rotate a register right and left to determine where it starts when it's inserted into a song.

thebrotherspus wrote:
dang... that's too bad.
I understand the hesitation of messing around with stuff when it already works well under most circumstances.
It's too bad for me, as my system is super small- I only have one precision adder (contained in Disting mk4) and no room for anymore. I was hoping it would be possible as an option in a "main menu" type of thing to switch only the pitch related cv to positive voltages. Other bipolar voltages in hemisphere are fine since everything else in my system (including piston honda's other cv inputs, excluding v/oct) can accept bipolar.
I understand, however, if this is simply not feasible or just something you're not interested in.


It's something I'll continue to think about, and it's something I'll keep in mind as I design apps and applets. I think the problematic areas are realistically very limited:

* For most applications of the quantizer, you choose what goes in, so that shouldn't be a problem.
* If you're using MIDI, you can transpose at the output device to keep things above middle C.
* Carpeggio starts low, but it has a global transpose, which should get you into the right range. If you try it, and it doesn't transpose high enough, let me know (but make sure you're using 1.4C, because there are some Carpeggio tuning changes in that version).

So if I understand correctly, the only problems are ShiftReg and Sequence5. And I don't think anybody would object if I bumped up the offset to 60 semitones instead of 48.
thebrotherspus
chysn wrote:

So if I understand correctly, the only problems are ShiftReg and Sequence5. And I don't think anybody would object if I bumped up the offset to 60 semitones instead of 48.


I haven't used too many apps yet, but I assume that you're probably correct- it's likely just those two apps that cause my issues with negative voltages. As you say, there seem to be work arounds for the other apps.
I would love it if this could be implemented in shiftreg and sequence5. Especially considering the upcoming developments with your indepth turing machine. This would make it super useful for me. Thanks so much for at least considering this.
chysn
thebrotherspus wrote:
chysn wrote:

So if I understand correctly, the only problems are ShiftReg and Sequence5. And I don't think anybody would object if I bumped up the offset to 60 semitones instead of 48.


I haven't used too many apps yet, but I assume that you're probably correct- it's likely just those two apps that cause my issues with negative voltages. As you say, there seem to be work arounds for the other apps.
I would love it if this could be implemented in shiftreg and sequence5. Thanks for considering this.


Yeah, those changes will go out in the Sept. 28 release.
thebrotherspus
chysn wrote:
thebrotherspus wrote:
chysn wrote:

So if I understand correctly, the only problems are ShiftReg and Sequence5. And I don't think anybody would object if I bumped up the offset to 60 semitones instead of 48.


I haven't used too many apps yet, but I assume that you're probably correct- it's likely just those two apps that cause my issues with negative voltages. As you say, there seem to be work arounds for the other apps.
I would love it if this could be implemented in shiftreg and sequence5. Thanks for considering this.


Yeah, those changes will go out in the Sept. 28 release.


That's fantastic!
So would this apply to your upcoming in depth Turing Machine app too?
rayultine
chysn wrote:
Shades is also pretty high-precision, if you want something that's more generally useful.


lol, it seems like the only real dead end with Hemispheres users is that now we all need a Shades. chysn, are you getting kick-backs from Olivier? razz
chysn
rayultine wrote:
chysn wrote:
Shades is also pretty high-precision, if you want something that's more generally useful.


lol, it seems like the only real dead end with Hemispheres users is that now we all need a Shades. chysn, are you getting kick-backs from Olivier? razz


That would be AWESOME!

But seriously, Shades should just be built into the O_C hardware. Sort of like this:

http://northernlightmodular.com/preorder/

It's eight hundred euros, but it's got an attenuator on every single in and out.
chess1
unless there is already a way to do this...could long top button press save the applet selection to be the same upon the next reboot ?
chess1
ah nevermind. I just learned how to save state! hmmm.....
chysn
thebrotherspus wrote:
chysn wrote:
thebrotherspus wrote:
chysn wrote:

So if I understand correctly, the only problems are ShiftReg and Sequence5. And I don't think anybody would object if I bumped up the offset to 60 semitones instead of 48.


I haven't used too many apps yet, but I assume that you're probably correct- it's likely just those two apps that cause my issues with negative voltages. As you say, there seem to be work arounds for the other apps.
I would love it if this could be implemented in shiftreg and sequence5. Thanks for considering this.


Yeah, those changes will go out in the Sept. 28 release.


That's fantastic!
So would this apply to your upcoming in depth Turing Machine app too?


Enigma has pitch selections that range from 3 to 7 bits. 3- to 6-bit pitches will be offset by 60 semitones, so they'll all start around 0V (depending on the lookup generated by the quantizer). The 7-bit pitch will go negative, because it has 127 notes. Each song step has its own transpose value, as well.
yghartsyrt
just wondering, if there's still room for a rungler applet...
chysn
yghartsyrt wrote:
just wondering, if there's still room for a rungler applet...


There's plenty of room for applets. But except I don't know what a rungler is. What would this do?
synkrotron
chysn wrote:
yghartsyrt wrote:
just wondering, if there's still room for a rungler applet...


There's plenty of room for applets. But except I don't know what a rungler is. What would this do?


https://www.muffwiggler.com/forum/viewtopic.php?t=92049&sid=f35a37a5aa 1b409de5319628333ba286
yghartsyrt
synkrotron wrote:
chysn wrote:
yghartsyrt wrote:
just wondering, if there's still room for a rungler applet...


There's plenty of room for applets. But except I don't know what a rungler is. What would this do?


https://www.muffwiggler.com/forum/viewtopic.php?t=92049&sid=f35a37a5aa 1b409de5319628333ba286


This ^^

Basically it is a non linear feedback shift register:
http://electro-music.com/forum/viewtopic.php?t=38081
chysn
Has anybody ever written a software rungler?
starthief
chysn wrote:
Has anybody ever written a software rungler?


Funnily enough, I've been playing with a 4-bit shift register design in Monome Teletype over the last couple of days. What I have now is not very far from a rungler (I'd have to add the second "oscillator" to sample via script delays if I didn't want to use an external source.)

The problem there is, Teletype is meant more for sequencer time scales so my rungler would have a theoretical max clock rate of 500Hz. (Meaning a max of 250Hz audio when the output values happen to alternate between high and low...)

I'm guessing O_C could run fast enough to be interesting but not match the CMOS speeds of a "real" rungler.
yghartsyrt
chysn wrote:
Has anybody ever written a software rungler?


https://cycling74.com/forums/sharing-is-rungling-stepped-havoc-in-gen/

and there is a rungler in michael Hetricks Reaktor Blocks.
https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/9093/

evene more background:
http://electro-music.com/forum/viewtopic.php?t=38081


thanks for taking interest
chysn
starthief wrote:
The problem there is, Teletype is meant more for sequencer time scales so my rungler would have a theoretical max clock rate of 500Hz. (Meaning a max of 250Hz audio when the output values happen to alternate between high and low...)

I'm guessing O_C could run fast enough to be interesting but not match the CMOS speeds of a "real" rungler.


Yeah, the O_C is meant more for modulation time scales. I'm pretty sure it wouldn't have the dynamic range to do this convincingly.

I bet it could be done pretty well with the 1010 Euroshield.
yghartsyrt
chysn wrote:
starthief wrote:
The problem there is, Teletype is meant more for sequencer time scales so my rungler would have a theoretical max clock rate of 500Hz. (Meaning a max of 250Hz audio when the output values happen to alternate between high and low...)

I'm guessing O_C could run fast enough to be interesting but not match the CMOS speeds of a "real" rungler.


Yeah, the O_C is meant more for modulation time scales. I'm pretty sure it wouldn't have the dynamic range to do this convincingly.

I bet it could be done pretty well with the 1010 Euroshield.


it doesn't need to go into oscillator range. a rungler at lfo speed is still a great random source – something similar to a complex s&h
hinterlands303
Question - does the clocked mode of the dual quantizer work the same way as the O_C qauntermain app (or the doepfer a-156)? Meaning that every time it receives a gate or trigger it samples another voltage? Also, does this mode have an unquantized setting? I ofter use the quantermain app as a straight up sample and hold for non-pitch related duties and the unquantized option is nice for this.
chysn
hinterlands303 wrote:
Question - does the clocked mode of the dual quantizer work the same way as the O_C qauntermain app (or the doepfer a-156)? Meaning that every time it receives a gate or trigger it samples another voltage? Also, does this mode have an unquantized setting? I ofter use the quantermain app as a straight up sample and hold for non-pitch related duties and the unquantized option is nice for this.


Each half of Dual Quantizer starts off in “continuous mode,” meaning that it quantizes as CV comes in. If you provide a clock at a digital in, it switches to clocked mode, and a clock icon will appear. It remains in clocked mode until you change the scale.

If you set the scale to Off, it will be unquantized.

There’s also an actual sample and hold in Calculate. Calculate also has a clocked random setting if you want to sample and hold from a noise source.
chysn
yghartsyrt wrote:
chysn wrote:
starthief wrote:
The problem there is, Teletype is meant more for sequencer time scales so my rungler would have a theoretical max clock rate of 500Hz. (Meaning a max of 250Hz audio when the output values happen to alternate between high and low...)

I'm guessing O_C could run fast enough to be interesting but not match the CMOS speeds of a "real" rungler.


Yeah, the O_C is meant more for modulation time scales. I'm pretty sure it wouldn't have the dynamic range to do this convincingly.

I bet it could be done pretty well with the 1010 Euroshield.


it doesn't need to go into oscillator range. a rungler at lfo speed is still a great random source – something similar to a complex s&h


So you have a clock source coming in. When this clock goes high, it reads the ADC for a second signal. Based on the read, it pushes a high or low bit to an 8-bit shift register. Bit 7 falls off the end. You take the low three bits and “do something.” What do you do? Feed it back to a frequency parameter? What frequency does this parameter control? Where does the output signal actually come from?

If the signal is based on the last three ADC reads, why does it matter how big the shift register is? You never see a bit again after three reads.
starthief
yghartsyrt wrote:
it doesn't need to go into oscillator range. a rungler at lfo speed is still a great random source – something similar to a complex s&h


I think at higher clock rates there is something pretty cool about a rungler, Benjolin etc. but at lower speeds I don't think it has any advantage over a Turing Machine style LSFR.


What I wrote in Teletype was basically half an Omiindustriies DDSR, except slow smile In effect it's a length 4 Turing Machine where the chance of flipping a bit is either 0% or 100% according to an external input.

To make that imitate a rungler I clocked it internally, dropped the CV output resolution to 3 bits, made the clock speed dependent on the CV output and patched an external oscillator to the data input.

I preferred the more open version before I did that though smile


With all this in mind... what are the chances of adding more input to the Shift Register app? CV1/3 could be assigned to probability. Digital 1/3 could force 0% probability when high to freeze the pattern, and digital 2/4 could force 100% probability when high (overridden by the freeze)
venus_3036
EDIT: saw you'd already had this chat about runglers!

Someone tried, its a fun little m4l device infact !

http://www.maxforlive.com/library/device/4670/simple-rungler
chysn
starthief wrote:
what are the chances of adding more input to the Shift Register app? CV1/3 could be assigned to probability. Digital 1/3 could force 0% probability when high to freeze the pattern, and digital 2/4 could force 100% probability when high (overridden by the freeze)


There's going to be another Turing Machine applet based on Enigma that will have additional CV. However, I don't like probability modulation for a Turing Machine at all. It doesn't add much value. At some point, you tend to have something that sounds random, and it doesn't sound appreciably different whether you modulate the probability to 15% or 65%.

So the two CV inputs for the Enigma-based TM will be (1) Transposition, which is a good, high-impact modulation, and (2) Organize.

I'm guessing that Organize is something never before seen with Turing Machines. With Enigma, you'll be able to curate a library of 40 Turing Machines and save them to memory or via sysex. So Organize in the applet will let you choose, via CV, from among a subset (or all, your choice) of your Turing Machine library.
sushiluv
chysn wrote:
starthief wrote:
what are the chances of adding more input to the Shift Register app? CV1/3 could be assigned to probability. Digital 1/3 could force 0% probability when high to freeze the pattern, and digital 2/4 could force 100% probability when high (overridden by the freeze)


There's going to be another Turing Machine applet based on Enigma that will have additional CV. However, I don't like probability modulation for a Turing Machine at all. It doesn't add much value. At some point, you tend to have something that sounds random, and it doesn't sound appreciably different whether you modulate the probability to 15% or 65%.

So the two CV inputs for the Enigma-based TM will be (1) Transposition, which is a good, high-impact modulation, and (2) Organize.

I'm guessing that Organize is something never before seen with Turing Machines. With Enigma, you'll be able to curate a library of 40 Turing Machines and save them to memory or via sysex. So Organize in the applet will let you choose, via CV, from among a subset (or all, your choice) of your Turing Machine library.


oh, can´t wait for it!
starthief
Enigma sounds very cool.

I realize my mistake now: Probability on a Turing Machine is the chance of randomly choosing a new value. What I was doing was using a (Euclidean) pattern sequence XORd against the feedback bit with no randomness.

My typical use cases for Turing Machine, or Deja Vu on Marbles, are:

-- low but constant chance of change (no CV needed)

-- no chance of change for the most part, but interrupt it with a high chance of change to create a new pattern (Enigma sounds like it will handle that case brilliantly)
hinterlands303
Cool - I think my confusion was around the term "clocked" which I thought could mean a steady clock vs. sampling every time a gate or trigger goes high.
Does the digital input of the sample and hold respond to audio rate signals like O_C does? I sometimes use quantermain as for sample rate reduction.

I think I'll try to load Hemisphere on my uO_C this weekend. I'm pretty much always in quantermain mode but I rarely use more than two channels of quantization so being able to load another applet is really appealing (along with the simplified interface). Great work!

chysn wrote:
hinterlands303 wrote:
Question - does the clocked mode of the dual quantizer work the same way as the O_C qauntermain app (or the doepfer a-156)? Meaning that every time it receives a gate or trigger it samples another voltage? Also, does this mode have an unquantized setting? I ofter use the quantermain app as a straight up sample and hold for non-pitch related duties and the unquantized option is nice for this.


Each half of Dual Quantizer starts off in “continuous mode,” meaning that it quantizes as CV comes in. If you provide a clock at a digital in, it switches to clocked mode, and a clock icon will appear. It remains in clocked mode until you change the scale.

If you set the scale to Off, it will be unquantized.

There’s also an actual sample and hold in Calculate. Calculate also has a clocked random setting if you want to sample and hold from a noise source.
chysn
hinterlands303 wrote:
Cool - I think my confusion was around the term "clocked" which I thought could mean a steady clock vs. sampling every time a gate or trigger goes high.


I see. The concepts are pretty much interchangeable in terms of what sort of signal the receiving device sees. I wish we used a word like "pulse" in Eurorack, like Buchla does.

Quote:
Does the digital input of the sample and hold respond to audio rate signals like O_C does? I sometimes use quantermain as for sample rate reduction.


Yes, it can sample a million times a minute, or 16.7kHz.
hinterlands303
chysn wrote:
hinterlands303 wrote:
Cool - I think my confusion was around the term "clocked" which I thought could mean a steady clock vs. sampling every time a gate or trigger goes high.


I see. The concepts are pretty much interchangeable in terms of what sort of signal the receiving device sees. I wish we used a word like "pulse" in Eurorack, like Buchla does.


Mostly interchangeable but I've got a couple of digital modules that have "clocked modes" where they are expecting a steady clock otherwise they freak out so I just wanted to be sure.

Looking forward to trying it out!
starthief
Just got a uO_C (built by ZZ Ardoz, who I'd recommend) and almost the first thing I did was install Hemisphere Suite.

I confused myself at first by accidentally having clock forwarding enabled -- it's what I get for diving in without paying close enough attention to the wiki -- but now I'm cruising around it just fine.

Very cool stuff here. I wonder if I'm going to have to clear another 8HP for a second uO_C Mr. Green
selfdestroyer
starthief wrote:
Very cool stuff here. I wonder if I'm going to have to clear another 8HP for a second uO_C Mr. Green


Exactly what I did.. well worth having them both around.
chysn
starthief wrote:
Just got a uO_C (built by ZZ Ardoz, who I'd recommend)


Thanks for the recommendation of a builder. I think it's inevitable that I'll have to switch over to a µO_C sooner or later in order to make room for The Bends. I need an additional 6HP to support The Bends, and that's the difference between O_C and µO_C.
jonkers303
Only thing i miss from original fw is copier machine.

Can you replikate the functionality of a 2 channel copier machine with any of the applets?
chysn
jonkers303 wrote:
Only thing i miss from original fw is copier machine.

Can you replikate the functionality of a 2 channel copier machine with any of the applets?


Possibly! I think somebody mentioned it before, but didn't respond to my questions about it. Can you give me a description of what Copier Maschine does, and an ideal of how a 2-channel version would work?
Shledge
Copier Maschine is basically 4 S&Hs which act in sequence, moving the held voltage from one to the other (essentially making it an ASR), and if quantised, allows for interesting chords.

A 2 channel version would essentially be the same thing, only making 2 notes rather than 4, although I would imagine it could be easily patched with another instance to make 4 channels (or more if you have more o_Cs).
chysn
Shledge wrote:
Copier Maschine is basically 4 S&Hs which act in sequence, moving the held voltage from one to the other (essentially making it an ASR), and if quantised, allows for interesting chords.

A 2 channel version would essentially be the same thing, only making 2 notes rather than 4, although I would imagine it could be easily patched with another instance to make 4 channels (or more if you have more o_Cs).


Okay, thanks! That sounds like a cool thing. It does make sense for both hemispheres to share the same set of data. So if one instance is running, A/C is the most recent sample, and B/D is the previous one. If two instances are running, it would go A, B, C, D, like Copier Maschine.

How does the "freeze" work? I think it just disables the input so that the same 4-step pattern is rotated. Is that right?
pld
chysn wrote:
How does the "freeze" work? I think it just disables the input so that the same 4-step pattern is rotated. Is that right?

Yes, freeze stops adding values to the internal ring buffer. That contains more than 4 values though, it's up to the last 256 values. Another way of describing it might be that it's a 4-tap delay, and you can control where the taps tap into the buffer and their spacing (index + delay parameters, IIRC).
But it's probably possible to construct arbitrary descriptions by mixing and mangling the phrases "S&H", "delay", "tap" and "shift register" smile
chysn
pld wrote:
chysn wrote:
How does the "freeze" work? I think it just disables the input so that the same 4-step pattern is rotated. Is that right?

Yes, freeze stops adding values to the internal ring buffer. That contains more than 4 values though, it's up to the last 256 values. Another way of describing it might be that it's a 4-tap delay, and you can control where the taps tap into the buffer and their spacing (index + delay parameters, IIRC).
But it's probably possible to construct arbitrary descriptions by mixing and mangling the phrases "S&H", "delay", "tap" and "shift register" smile


Ah, thanks. I suspect it would be best to load up the firmware and actually try it out for a while.
jonkers303
Great thanks everyone that helped explain copier machine.

Chysn if your thinking of implementing this as a new applet thats awesome!

Then i really think it should follow the tradition of hemisphere and keep it really simple with focus on the core functionality, but maybe that goes without saying.

Just let me know if i can help in any way!
chysn
jonkers303 wrote:
Chysn if your thinking of implementing this as a new applet thats awesome!


I think it is certain.
hinterlands303
chysn wrote:
starthief wrote:
Just got a uO_C (built by ZZ Ardoz, who I'd recommend)


Thanks for the recommendation of a builder. I think it's inevitable that I'll have to switch over to a µO_C sooner or later in order to make room for The Bends. I need an additional 6HP to support The Bends, and that's the difference between O_C and µO_C.


Another recommendation for ZZ Ardoz. He built my uO_C and has been really helpful with some questions I had about updating firmware etc.

Also - I just got Hemisphere up and running and I'm really enjoying it. The simplicity of the applets is just what I needed to get out of my O_C rut.
Godphaser
I add my recommendation for ZZ Ardoz too. He built an uO_C for me and one for my friend, both beautiful modules, great service.
chysn
How do you get hold of ZZ Ardoz? Contact him here on MuffWiggler, or does have have an online store on Reverb or something?
Zymos
PM him here.
jonkers303
chysn wrote:
jonkers303 wrote:
Chysn if your thinking of implementing this as a new applet thats awesome!


I think it is certain.


Niice, thanks!
Shledge
I've noticed there is noticeable slight latency when you put a gate into the quantiser modes, especially if you tie it to a VCO being controlled by a VCA or filter - you'll hear a small part of the previous note as tit updates.

If you let it autodetect, there's no latency.
chysn
Shledge wrote:
I've noticed there is noticeable slight latency when you put a gate into the quantiser modes, especially if you tie it to a VCO being controlled by a VCA or filter - you'll hear a small part of the previous note as tit updates.

If you let it autodetect, there's no latency.


The latency is intentional. The digital read is faster than the ADC read, so the latency is there to assure that the right value is used. However, I'm pretty sure I went way overboard with it, in terms of time. I'll take another look at that and have a fix in this month's release.
mxmxmx
chysn wrote:
Shledge wrote:
I've noticed there is noticeable slight latency when you put a gate into the quantiser modes, especially if you tie it to a VCO being controlled by a VCA or filter - you'll hear a small part of the previous note as tit updates.

If you let it autodetect, there's no latency.


The latency is intentional. The digital read is faster than the ADC read, so the latency is there to assure that the right value is used. However, I'm pretty sure I went way overboard with it, in terms of time. I'll take another look at that and have a fix in this month's release.


fwiw, this branch handles the ADC stuff somewhat differently now, speeding things quite up a bit (see "Commits on Aug 30, 2018"); it could be made to run even faster; generally speaking though the problem begins upstream, hence the configurable latency in o_C.
chysn
Thanks! I’ll take a look at that, too. I think the primary cause of the problem Shledge reported stems from my misreading of the O_C code, and I used the maximum latency value.
yghartsyrt
chysn wrote:
yghartsyrt wrote:
chysn wrote:
starthief wrote:
The problem there is, Teletype is meant more for sequencer time scales so my rungler would have a theoretical max clock rate of 500Hz. (Meaning a max of 250Hz audio when the output values happen to alternate between high and low...)

I'm guessing O_C could run fast enough to be interesting but not match the CMOS speeds of a "real" rungler.


Yeah, the O_C is meant more for modulation time scales. I'm pretty sure it wouldn't have the dynamic range to do this convincingly.

I bet it could be done pretty well with the 1010 Euroshield.


it doesn't need to go into oscillator range. a rungler at lfo speed is still a great random source – something similar to a complex s&h


So you have a clock source coming in. When this clock goes high, it reads the ADC for a second signal. Based on the read, it pushes a high or low bit to an 8-bit shift register. Bit 7 falls off the end. You take the low three bits and “do something.” What do you do? Feed it back to a frequency parameter? What frequency does this parameter control? Where does the output signal actually come from?

If the signal is based on the last three ADC reads, why does it matter how big the shift register is? You never see a bit again after three reads.


The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.
chysn
yghartsyrt wrote:
The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.


What effect does the rungler input have on the clocking and distribution oscillators?
Attackman
chysn wrote:
starthief wrote:
what are the chances of adding more input to the Shift Register app? CV1/3 could be assigned to probability. Digital 1/3 could force 0% probability when high to freeze the pattern, and digital 2/4 could force 100% probability when high (overridden by the freeze)


With Enigma, you'll be able to curate a library of 40 Turing Machines and save them to memory or via sysex. So Organize in the applet will let you choose, via CV, from among a subset (or all, your choice) of your Turing Machine library.


Whoa. Amazing. Let this flow of constant goodness never end! : )
yghartsyrt
chysn wrote:
yghartsyrt wrote:
The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.


What effect does the rungler input have on the clocking and distribution oscillators?


It's a stepped CV control. Just like controlling both oscillator frequency with a S&H.
So as an applet it would be CV1-In for the signal to be sampled (distribution) and CV2-In for clock. CV-Out would be the the analog output of last 3 bits as a CV. One trigger-in for freezing the buffer.
chysn
yghartsyrt wrote:
chysn wrote:
yghartsyrt wrote:
The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.


What effect does the rungler input have on the clocking and distribution oscillators?


It's a stepped CV control. Just like controlling both oscillator frequency with a S&H.
So as an applet it would be CV1-In for the signal to be sampled (distribution) and CV2-In for clock. CV-Out would be the the analog output of last 3 bits as a CV. One trigger-in for freezing the buffer.


So there's no internal feedback into another input? And what's the other CV output? Would it be a different 3 bits? Like a multi-tap rungler?
yghartsyrt
chysn wrote:
yghartsyrt wrote:
chysn wrote:
yghartsyrt wrote:
The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.


What effect does the rungler input have on the clocking and distribution oscillators?


It's a stepped CV control. Just like controlling both oscillator frequency with a S&H.
So as an applet it would be CV1-In for the signal to be sampled (distribution) and CV2-In for clock. CV-Out would be the the analog output of last 3 bits as a CV. One trigger-in for freezing the buffer.


So there's no internal feedback into another input? And what's the other CV output? Would it be a different 3 bits? Like a multi-tap rungler?


there's no other output in the original rungler. but using three other bits would be a possible option. multi-tap rungler sounds like a fitting idea in that case.


In the original rungler design you can crossmodulate the two oscillators (clock and distribution). But that has nothing to to with the rungler itself.
chysn
yghartsyrt wrote:
chysn wrote:
yghartsyrt wrote:
chysn wrote:
yghartsyrt wrote:
The 8-bit shift register can be looped.
The 3 bits at the ouput form the rungler output. Since both the clocking and the distribution oscillator have rungler inputs it can used for a more complex feedback-chain.


What effect does the rungler input have on the clocking and distribution oscillators?


It's a stepped CV control. Just like controlling both oscillator frequency with a S&H.
So as an applet it would be CV1-In for the signal to be sampled (distribution) and CV2-In for clock. CV-Out would be the the analog output of last 3 bits as a CV. One trigger-in for freezing the buffer.


So there's no internal feedback into another input? And what's the other CV output? Would it be a different 3 bits? Like a multi-tap rungler?


there's no other output in the original rungler. but using three other bits would be a possible option. multi-tap rungler sounds like a fitting idea in that case.


In the original rungler design you can crossmodulate the two oscillators (clock and distribution). But that has nothing to to with the rungler itself.


A key part of this is still whooshing over my head. Every pulse of the clock reads the distribution oscillator, shifts the register left, and sets or leaves unset bit 0 based on that read. So when does the rotation come in? It seems like you're always using the most recent three reads. Why does the length of the register (in excess of three bits) even matter?
starthief
From what I gather in this thread https://www.youtube.com/watch?v=gls5eug9P9M it's the *oldest* three reads that make the CV out.


Here's Rob Hordijk giving a demo/talk on the Rungler:

[video]https://www.youtube.com/watch?v=gls5eug9P9M[/video]

Unfortunately it's about 8 minutes of general talk about his inspiration for it, and when he starts to get into the explanation of how it works he quickly runs out of time very frustrating
thelizard
yghartsyrt wrote:
chysn wrote:
Has anybody ever written a software rungler?


https://cycling74.com/forums/sharing-is-rungling-stepped-havoc-in-gen/

and there is a rungler in michael Hetricks Reaktor Blocks.
https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/9093/

evene more background:
http://electro-music.com/forum/viewtopic.php?t=38081


thanks for taking interest


Hey, thanks for sharing!

As a heads up, I don't have a hardware Rungler to compare against. I made the Euro Reakt Rungler based on my understanding of the description and schematics. I'm decently confident about it, and it's at least fun to use (I especially like how the Rungler Oscillator Block turned out). Still, if someone more familiar with the Rungler has any feedback about how my Block compares, please PM me! I'd be happy to hear your experiences.
chysn
thelizard wrote:
yghartsyrt wrote:
chysn wrote:
Has anybody ever written a software rungler?


https://cycling74.com/forums/sharing-is-rungling-stepped-havoc-in-gen/

and there is a rungler in michael Hetricks Reaktor Blocks.
https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/9093/

evene more background:
http://electro-music.com/forum/viewtopic.php?t=38081


thanks for taking interest


Hey, thanks for sharing!

As a heads up, I don't have a hardware Rungler to compare against. I made the Euro Reakt Rungler based on my understanding of the description and schematics. I'm decently confident about it, and it's at least fun to use (I especially like how the Rungler Oscillator Block turned out). Still, if someone more familiar with the Rungler has any feedback about how my Block compares, please PM me! I'd be happy to hear your experiences.


It's an amusing situation, isn't it? Nobody has the real thing, and there's no really definitive description, so we all just sort of build caricatures.

My first shot at it is in GitHub (RunglBook). Digital 1/3 is the clock, CV 1/3 is the signal, and the outputs are Rungler outputs using bits 0-2 (A/C) and 5-7 (B/D). And that's it. Bit 0 is set when the voltage of the signal goes over 3V.

So there's the starting point of my particular caricature.
yghartsyrt
chysn wrote:
thelizard wrote:
yghartsyrt wrote:
chysn wrote:
Has anybody ever written a software rungler?


https://cycling74.com/forums/sharing-is-rungling-stepped-havoc-in-gen/

and there is a rungler in michael Hetricks Reaktor Blocks.
https://www.native-instruments.com/en/reaktor-community/reaktor-user-l ibrary/entry/show/9093/

evene more background:
http://electro-music.com/forum/viewtopic.php?t=38081


thanks for taking interest


Hey, thanks for sharing!

As a heads up, I don't have a hardware Rungler to compare against. I made the Euro Reakt Rungler based on my understanding of the description and schematics. I'm decently confident about it, and it's at least fun to use (I especially like how the Rungler Oscillator Block turned out). Still, if someone more familiar with the Rungler has any feedback about how my Block compares, please PM me! I'd be happy to hear your experiences.


It's an amusing situation, isn't it? Nobody has the real thing, and there's no really definitive description, so we all just sort of build caricatures.

My first shot at it is in GitHub (RunglBook). Digital 1/3 is the clock, CV 1/3 is the signal, and the outputs are Rungler outputs using bits 0-2 (A/C) and 5-7 (B/D). And that's it. Bit 0 is set when the voltage of the signal goes over 3V.

So there's the starting point of my particular caricature.


Actually, I have a Rungler (aka Benjolin). Do I need to compile the applet?
I can check if the behavior is as expected. The Reaktor Rungler and the Max-Rungler are pretty close to the hardware - in terms of CV generation. Audiowise they sound a little different, but since sounding the same woud also mean emulating the adc, this isn't much of a surprise.
chysn
yghartsyrt wrote:
Do I need to compile the applet?
I can check if the behavior is as expected.


Yes, the next hex file release will be Sept. 28.
yghartsyrt
chysn wrote:
yghartsyrt wrote:
Do I need to compile the applet?
I can check if the behavior is as expected.


Yes, the next hex file release will be Sept. 28.


ah ok thanks. I'll check if I manage to compile it.
READYdot
Just chiming in to give my thanks to the developper of the Hemisphere... outstanding work!
naturarerum
hi chysn, would it be possible to expand on the o_c References app and build a precision adder as part of Hemisphere? there are not many intuitive adders in eurorack (disting is annoying to use, all I can think of is beast's chalkboard) and having a screen to control what's happening would be really nice.
Shledge
I think there is already an applet that pretty much does that.
chysn
READYdot wrote:
Just chiming in to give my thanks to the developper of the Hemisphere... outstanding work!


You're most welcome!

naturarerum wrote:
hi chysn, would it be possible to expand on the o_c References app and build a precision adder as part of Hemisphere? there are not many intuitive adders in eurorack (disting is annoying to use, all I can think of is beast's chalkboard) and having a screen to control what's happening would be really nice.


Hemisphere Suite 1.5 (release date of 9/28) will include a transposing quantizer, which will have a v/oct adder as one of its inputs. So you'll be able to take a signal, add (or subtract) with voltage into the second input, and get a quantized note from the output.

Why not a regular precision adder? Because the O_C's ADCs are simply not up to the task, and all I'd ever hear about it is that it's not accurate. The DACs, on the other hand, are really accurate, so going through the quantizer compensates for the ADCs in some ways.

The transposing quantizer will fill the role of a precision adder in many use cases.

Shledge wrote:
I think there is already an applet that pretty much does that.


The Voltages applet can be used in conjunction with an external unity mixer for this purpose.
naturarerum
it's strange, I thought o_c was a super precise device, but I guess that's just the DACs? anyway it's cool that the quantizer gets transposition. smile
chysn
If anybody would like to try Enigma before its release on Sept. 28, please PM me for a link to the hex file. You are, of course, welcome to compile it yourself.

The current documentation is here: https://github.com/Chysn/O_C-HemisphereSuite/wiki/Enigma
mcpepe
Excuse me if this has been asked before.
Do we have any app that could do envelopes like vactrol gates, or physical modelled plucks, or like the Natural Gate, to make bongos easier?
Shledge
You do realise a LPG is more than the vactrol response?
chysn
mcpepe wrote:
Excuse me if this has been asked before.
Do we have any app that could do envelopes like vactrol gates, or physical modelled plucks, or like the Natural Gate, to make bongos easier?


DSP is pretty much off Ornament and Crime's table. The sample rate is just too damn low, among various other well-trod limitations.

Expect some form of simple drum synthesis in an upcoming release, though.
mcpepe
I know the low pass gate affects the vca and the filter, but I can multiple the envelope from Hemisphere to modulate my vca and my filter.
Maybe I did not formulate my question correctly. What I am asking for is only an envelope, maybe with only one paremeter (decay), that emulates the curve of acoustic instruments.

I know the hardware has its limitations, so it may not be posible. Nevertheless, I am very happy with all the options Hemisphere is giving us and how easy is to use it!

This simple drum synthesis sound very good...
bemushroomed
I'd focus 100% on non DSP-things for this module tbh... those algos (turing machine type of stuff etc) is why people got it in the first place.

Enigma sounds incredible, can't wait hyper
chysn
mcpepe wrote:
I know the low pass gate affects the vca and the filter, but I can multiple the envelope from Hemisphere to modulate my vca and my filter.
Maybe I did not formulate my question correctly. What I am asking for is only an envelope, maybe with only one paremeter (decay), that emulates the curve of acoustic instruments.


Oh, I get what you're saying now.

I don't have the mathematical background to model the decay of vactrols or plucked instruments.
pld
chysn wrote:
I don't have the mathematical background to model the decay of vactrols or plucked instruments.

A place to start digging could be the Plaits or Streams source. Neither will be immediately useable since the former uses float, but the latter might just need rebuilding the LUTs and extracting of the relevant parts (although it might be easier to reimplement).
4eyez
Very excited about Enigma, I keep reading the manual over and over. Ha!
jonkers303
Something that would be really cool as an applet would be a random acid pattern generator.
That generates simple randomized 16 step (or so) acid lines with slides and all at the push of a button. With CV and gate out.

Something like this:
https://www.ableton.com/en/blog/sting-free-acid-pattern-generator-skin nerbox/

A version of the density knob as in above mentioned sting would be extra icing on the cake.

I have no idea how complex the algorithm would have to be to generate musical results thou.

Would be a unique functionality in eurorack i think?
Zymos
I was just wishing for that exact thing! I have one for M4L, and would love to be able to do the same thing without needing the laptop, MIDI-CV converter, etc...
dianusindustrial
Just wanted to share a cool Hemisphere experiment I tried today. I've been wanting to test out TheSlowGrowth's VCO tuning app (https://github.com/TheSlowGrowth/VCOTuner) for a while now but I don't own a midi->cv module. Up until now I've been thinking about grabbing a Yarns or something similar but then I had a thought, hmmn I could probably do this with Hemisphere! I tried it out today and it works great! I was able to tune a couple of my oscillators so now they all track 1v per oct very accurately. Give it a shot!

Picture file
chysn
jonkers303 wrote:
Something that would be really cool as an applet would be a random acid pattern generator.
***
I have no idea how complex the algorithm would have to be to generate musical results thou.


It's interesting. The algorithm is the secret sauce, of course. It doesn't sound like purely random notes.

dianusindustrial wrote:
Just wanted to share a cool Hemisphere experiment I tried today. I've been wanting to test out TheSlowGrowth's VCO tuning app (https://github.com/TheSlowGrowth/VCOTuner) for a while now but I don't own a midi->cv module. Up until now I've been thinking about grabbing a Yarns or something similar but then I had a thought, hmmn I could probably do this with Hemisphere! I tried it out today and it works great! I was able to tune a couple of my oscillators so now they all track 1v per oct very accurately. Give it a shot!


Cool! Who needs Yarns?
Patch77
Hi wiggles,

Ive just flashed hemisphere and the screen came out upside down.. wasn't so in OG firmware.

Any help?

Best

Luke
chysn
Patch77 wrote:
Hi wiggles,

Ive just flashed hemisphere and the screen came out upside down.. wasn't so in OG firmware.

Any help?


Hey, Luke, I distribute the firmware in flipped and unflipped versions. Make sure you get a file that doesn't say FLIP180.

https://github.com/Chysn/O_C-HemisphereSuite/releases/download/HSv1.4C  /HemisphereSuite1_4C.hex
bemushroomed
jonkers303 wrote:
Something that would be really cool as an applet would be a random acid pattern generator.


That would be fantastic hyper
mazzyboy
jonkers303 wrote:
Something that would be really cool as an applet would be a random acid pattern generator.
That generates simple randomized 16 step (or so) acid lines with slides and all at the push of a button. With CV and gate out.
Would be a unique functionality in eurorack I think?


Yes, please! That would compliment the evolving randomness of the turing machines nicely!

FYI - I think you can kinda do this with erica's Pico SEQ. You can choose how many octaves to cover and what scale to work in, then press randomise. Don't think you can control density or number of slides though?
DJ_JITTER
mazzyboy wrote:
jonkers303 wrote:
Something that would be really cool as an applet would be a random acid pattern generator.
That generates simple randomized 16 step (or so) acid lines with slides and all at the push of a button. With CV and gate out.
Would be a unique functionality in eurorack I think?


Yes, please! That would compliment the evolving randomness of the turing machines nicely!

FYI - I think you can kinda do this with erica's Pico SEQ. You can choose how many octaves to cover and what scale to work in, then press randomise. Don't think you can control density or number of slides though?


There's no option to control the density/number of slides for random pattern generation, it's just assigned for each step when you push the RND button. However it is possible to set all notes to slide or to set it to a random chance of sliding on each new step, as a separate function to the random sequence generator.
chysn
There are already a zillion ways to generate random melodies. It sounds like what the world needs is a slew sequencer. On each clock pulse, a different slew rate could be applied, with the option to randomize the sequence.

Or a single rate, but with probability-based sliding. A Turing Machine, but for slew!
rayultine
I cooked something like this up with my disting, o_C Sequins, the Studio Electronics 3003 VCF, and a pulse osc. I have no idea how because I never write down patch notes. D:

https://soundcloud.com/noiserpuss/051617-slewratelimiter-acid
hinterlands303
chysn wrote:
There are already a zillion ways to generate random melodies. It sounds like what the world needs is a slew sequencer. On each clock pulse, a different slew rate could be applied, with the option to randomize the sequence.

Or a single rate, but with probability-based sliding. A Turing Machine, but for slew!


I use a slew limiter, a VC switch and a trigger/gate sequencer to do this by switching between the original and the slewed cv - but you just gave me the idea to use branches between the gate and the switch to add probability. Could be fun!
chysn
Good evening. Hemisphere Suite version 1.5 is now available as a hex file at https://github.com/Chysn/O_C-HemisphereSuite/releases . The change log is here:

* Significantly changed the operation of Screen Saver mode. The calibration option now sets the number of minutes, not seconds, until the screen saver activates. The screen saver is now "screen blanking," and always simply darkens the screen. You can blank the screen at any time by long-pressing the Up button.

* Added full app Enigma, a shift register (a.k.a. "Turing Machine") workstation that allows you to curate a library of up to 40 Turing Machine-like registers and build four track songs by chaining them together. The philosophy involves the use of shift registers as repeatable compositional elements rather than ephemeral improvisational elements. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Enigma)

* LowerRenz: Fixed issue that potentially caused a crash when both hemispheres had LowerRenz selected.

* Carpeggio: Standardized tuning and note values, to A=440Hz, C4=0V, C4=MIDI Note #60

* Added applet Enigma Jr., which allows real-time composition (via CV) using your Enigma library. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Enigma,-Jr.)

* Fixed issue that caused quantizer latency to be higher than it should have been.

* Added applet RunglBook, a chaotic shift-register modulation based on Rob Hordijk's rungler circuit. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/RunglBook)

* Added applet Squanch, a Shifting Quantizer/Adder. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Squanch---Shifting- Quantizer)

* Added applet "A"SR, an analog shift register delay. In one hemisphere, it has two outputs; when selected in two hemispheres, the buffer is linked and uses all four outputs. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/%22A%22SR)

* Added applet ShiftGate, a dual shift register-based aleatoric gate or trigger sequencer (https://github.com/Chysn/O_C-HemisphereSuite/wiki/ShiftGate)

* Removed References, the last of the original Ornament and Crime apps.

* Envelope Follower: Added per-channel configuration for Follow and Duck functions.

* MIDI In/MIDI Out: Log is now accessed with button push rather than screen saver activation (because screen saver is no longer a thing; see above).

* ShiftReg: Increased low range to 0V for better support of digital oscillators.

* Sequence5: Same as above

* Minor font changes to improve readability

Thanks to everybody for your suggestions! The next release will be on October 26.
bemushroomed
Excellent grin

That page is a 404 though..

here's a direct link to the .hex:

https://github.com/Chysn/O_C-HemisphereSuite/releases/download/HSv1.5/ HemisphereSuite1_5.hex

or https://github.com/Chysn/O_C-HemisphereSuite/releases/download/HSv1.5/ HemisphereSuite1_5_FLIP_SCREEN.hex (flipped)
chysn
bemushroomed wrote:
That page is a 404 though..


Thanks, fixed! The forum software molested my URL.
bemushroomed
Why not have gate outs for Darkest Timeline so that we can trigger a synth voice properly using these outs? On e.g a x0x Heart you will not hear a sound apart from a "kr.." because this 6ms trigger is too short for the synth.. if they were gates i could use one out to trigger the voice and another for accent (a trigger is too short to make proper accent).

I tried some other ones (Engima + a few more) and trigger outs seems to be the choice? Seems a bit useless to me except if you want to use them to trigger e.g drum modules. But there are better algos for that purpose i think.
chysn
bemushroomed wrote:
Why not have gate outs for Darkest Timeline so that we can trigger a synth voice properly using these outs?


That's my West Coast bias showing, I guess. I almost always trigger envelopes rather than gate them.

At this point, it would be more of an interface problem. I should be able to find space on the Setup screen for a gate option that would try to find a time that's around 90% of the clock period.

Quote:
I tried some other ones (Engima + a few more) and trigger outs seems to be the choice? Seems a bit useless to me except if you want to use them to trigger e.g drum modules. But there are better algos for that purpose i think.


Enigma has both trigger and gate output types. For the gate type, the gate goes high when the low bit is 1, and stays there until the bit becomes 0.
bemushroomed
Great, i'll dwell a bit deeper into Enigma then smile
aaooaa



So I think I've got a bug but it's my first time using Hemisphere so I might be wrong. When using Enigma Jr in the right Hemisphere the register is displayed on the left hand side and covers whatever is loaded there. Seems to work fine when using it on the left though.
Supervillain
Downloaded, will try it this weekend !
Thanks
chysn
aaooaa wrote:



So I think I've got a bug but it's my first time using Hemisphere so I might be wrong. When using Enigma Jr in the right Hemisphere the register is displayed on the left hand side and covers whatever is loaded there. Seems to work fine when using it on the left though.


Yes indeed, thanks!
chysn
For those who compile in Arduino IDE, the Enigma, Jr. display is fixed, and I've added a gate percentage to The Darkest Timeline. There's one more gate option I want to try to add, but I'm not yet sure if it'll work well. I'm planning the 1.5A hex file release of these items sometime this weekend.
jonkers303
Great release, thanks!
chysn
So, looking forward now, the October 26 (1.6) release's theme will be time-based stuff.

I wrote a system that I call a Vector Oscillator, which is basically an oscillator that uses a list of up to 12 level/time segments to represent a waveform, and scales that waveform to a specific frequency. The main use of this will be for complex user-defined LFOs and envelopes.

But the Vector Oscillator also goes into the audio range. The O_C's lowish sample rate limits how good the tuning can be, but it does allow me to do some types of tuning-agnostic synthesis, like drum synthesis, and I've got a pretty decent proof-of-concept bass drum using one Vector Oscillator for pitch, and another one for a volume envelope.

The Hemisphere project is necessarily winding down as I'm pushing the program memory's 85% mark. I think it's about time that I focused on some LFOs before I'm done, and I'm pretty excited about the flexibility of this sort of modulation framework.

In this spirit of "time-based stuff," there's a Metronome, which is a sort of master clock that keeps running, even when you select a different applet, so you no longer need to provide an external clock for sequencers, etc.
Misk
Incredible update, and your vector oscillator idea just seems badass. I can't wait to check that out—in the meantime, i've got a LOT of shift register learning to do smile
djd_oz
Is there an app for a precision adder or equivalent for hemisphere?
starthief
chysn wrote:
I wrote a system that I call a Vector Oscillator, which is basically an oscillator that uses a list of up to 12 level/time segments to represent a waveform, and scales that waveform to a specific frequency. The main use of this will be for complex user-defined LFOs and envelopes.


Excellent, that's sort of like what I thought Xaoc Zadar was going to be.

chysn wrote:
In this spirit of "time-based stuff," there's a Metronome, which is a sort of master clock that keeps running, even when you select a different applet, so you no longer need to provide an external clock for sequencers, etc.


Nice smile
djd_oz
chysn wrote:


The Hemisphere project is necessarily winding down as I'm pushing the program memory's 85% mark.


How easy or difficult is it to customise what apps are included in the hex file? i.e. we could have 100 apps available and I only want to select 25 apps that I want to be loaded into my o_c.
chysn
starthief wrote:
Excellent, that's sort of like what I thought Xaoc Zadar was going to be.


Funny you mention that, because I got the idea for the Vector Oscillator by imagining how Zadar might work behind-the-scenes. There's something ingenious about just having a bunch of musically-relevant preset curves.

I'm going to provide a waveform editor, though, so once you build these waveforms, you can go into applets and use them as envelopes or LFOs. But I've been trying to conserve variable memory so that I can provide a lot of presets, too.

jdj_oz wrote:
How easy or difficult is it to customise what apps are included in the hex file? i.e. we could have 100 apps available and I only want to select 25 apps that I want to be loaded into my o_c.


For Hemisphere, it's really easy. Each of the Hemisphere applets is a single file and a single line in hemisphere_config.h. So, remove the file, remove the corresponding config file line, and change the HEMISPHERE_AVAILABLE_APPLETS value, and that's all you need to do.
rayultine
chysn,
If you have a weird/crappy lo-fi looper and you are into making lo-fi oscillators, what would o_C lo-fi audio rate delay or reverb sound like? smile
chysn
rayultine wrote:
chysn,
If you have a weird/crappy lo-fi looper and you are into making lo-fi oscillators, what would o_C lo-fi audio rate delay or reverb sound like? smile


It depends on the settings. Lo-Fi Tape down-samples data to 8 bits, and samples at 2kHz. It's all about saving memory there.

The Vector Oscillator, on the other hand, has a dynamic range of 13 bits, give or take, and a sample rate of 16kHz. Things can sound much better because the signal is calculated rather than stored.

If you want to hear the O_C's full potential w/r/t sampling, you can put an audio signal through the GatedVCA applet. That pretty much passes signal through at the highest dynamic range, and as fast as the O_C can get it. It could sound worse, but the more important barrier is memory.
4eyez
chysn wrote:
I think it's about time that I focused on some LFOs before I'm done, and I'm pretty excited about the flexibility of this sort of modulation framework.


Excited about LFOs! Just going to recast my vote for a clock syncable LFO, something along the lines of a Peaks' tap mode would be amazing, but it sounds like you're already heading that way. Great work!
MuffBob
I recently got an O_C and like the idea of hemispheres a lot. Think 2 pairs of different applets will be more useful to me than 4 of 1.

However, wondering which version would be better for me to install - the original hemispheres which keeps many of the original O_C applets or the newer suite.

I seem to recall reading that the suite was essentially aimed at those with multiple O_Cs. And it includes none of the original applets.

Swinging towards the original hemispheres but the fact that it is now deprecated puts me off a bit.

Advice from those who have been in a similar situation would be most appreciated! thanks
chysn
MuffBob wrote:
I recently got an O_C and like the idea of hemispheres a lot. Think 2 pairs of different applets will be more useful to me than 4 of 1.

However, wondering which version would be better for me to install - the original hemispheres which keeps many of the original O_C applets or the newer suite.

I seem to recall reading that the suite was essentially aimed at those with multiple O_Cs. And it includes none of the original applets.

Swinging towards the original hemispheres but the fact that it is now deprecated puts me off a bit.

Advice from those who have been in a similar situation would be most appreciated! thanks


The original Hemisphere firmware isn't going to suddenly stop working. It's just that it can no longer be updated. So if you like what's there, go for it.

At this point, I see the O_C/HS choice more like a choice between two different modules. They cover much of the same ground, because they both cover a lot of ground. But it's a lot like choosing between Function and Dixie II+, in that they're both the same size and have some overlap.

Also, you can try Hemisphere Suite out, and go back to Ornament and Crime whenever you want. With Hemisphere Suite, you can save everything you do via MIDI, so you don't even need to risk losing your work if you re-change your mind later.

I've flashed my Teensy hundreds of times, they can take it.
monokinetic
Chysn,
version 1.5 sounds exciting, more excellent work by the looks of it, but I haven't upgraded yet because of:

Quote:
* Removed References, the last of the original Ornament and Crime apps.


I used this regularly so far, so that I could run the closed loop tuning thangy to set up my VCOs accurately. It worked nicely as far as I could tell (I assume it was working, right?). Is there any way now to do that in Hemisphere? If not is there any chance to readd it somehow? I'm fine to compile the s/w myself....
chysn
monokinetic wrote:
Chysn,
version 1.5 sounds exciting, more excellent work by the looks of it, but I haven't upgraded yet because of:

Quote:
* Removed References, the last of the original Ornament and Crime apps.


I used this regularly so far, so that I could run the closed loop tuning thangy to set up my VCOs accurately. It worked nicely as far as I could tell (I assume it was working, right?). Is there any way now to do that in Hemisphere? If not is there any chance to readd it somehow? I'm fine to compile the s/w myself....


I haven't really decided what to do with this feature yet, whether to ignore it, or replace it somehow. I was strongly leaning toward ignoring it, since the traditional workaround is "calibrate the VCO." It seemed like one of those 80/20 rule features whose absence would go largely unnoticed. I guess not!

If you want References, and don't mind compiling it yourself, all you need to do is replace the APP_REFS.ino file from the O_C repository, and add the line

DECLARE_APP('R','F', "References", REFS, REFS_isr)

to the OC_apps.ino file. Note that you only need to add this line in one place, not two, as in the O_C firmware. There are no "boring names" in Hemisphere Suite.

If you need additional details on where to put this line, or where to get the file, just let me know.
rayultine
yeah, I use References as a tuner mostly and am sad to see it go. Would it be easy to increase the range of the Tuner applet to go down to C0 (16.35 Hz) or C1? Would an enhanced Tuner in one Hemisphere and Voltage in the other reliable enough to replace References? smile
Timmy
monokinetic wrote:
Chysn,
version 1.5 sounds exciting, more excellent work by the looks of it, but I haven't upgraded yet because of:

Quote:
* Removed References, the last of the original Ornament and Crime apps.


I used this regularly so far, so that I could run the closed loop tuning thangy to set up my VCOs accurately. It worked nicely as far as I could tell (I assume it was working, right?). Is there any way now to do that in Hemisphere? If not is there any chance to readd it somehow? I'm fine to compile the s/w myself....


chysn
rayultine wrote:
yeah, I use References as a tuner mostly and am sad to see it go. Would it be easy to increase the range of the Tuner applet to go down to C0 (16.35 Hz) or C1? Would an enhanced Tuner in one Hemisphere and Voltage in the other reliable enough to replace References? smile


Tuner goes down to C-1, the point where a Make Noise STO almost sounds like an LFO. It's actually just transplanted code from References, so aside from a different look, it's the same thing with the same range.

And Voltage uses the same 128-steps-per-semitone formula that guides all Ornament and Crime pitch calculations so it, too, will be just as accurate as References.

So this was the idea behind removing References: Basically, it takes up memory that I want, and if I can accomplish some of what it does elsewhere, then bonus. I don't feel too much compulsion to keep anything as it was. After all, Ornaments and Crime are small and inexpensive.
rayultine
From the wiki, I was reading

"The A4 frequency has a range of 400-500Hz"

as

"The Tuner frequency has a range of 400-500Hz"

which is a vast difference. very frustrating
chysn
rayultine wrote:
"The Tuner frequency has a range of 400-500Hz"

which is a vast difference. very frustrating


Worst. Tuner. Ever! smile
rayultine
hahahahaha! I feel like that's the fifth or sixth time I have done that in this thread. I has reading comprehension skillz
Zymos
All you really need a tuner for is to get the A right, then everything else just falls into place...
monokinetic
chysn wrote:

I haven't really decided what to do with this feature yet, whether to ignore it, or replace it somehow. I was strongly leaning toward ignoring it, since the traditional workaround is "calibrate the VCO." It seemed like one of those 80/20 rule features whose absence would go largely unnoticed. I guess not!


Heheh, always the way! Truth be told I've just been testing the apps out, and you're adding features faster than I can learn them, so not entirely sure if I'll miss the autotune feature yet. (and anyway my 2OC has 2x O_C in it, so I can run both firmwares for maximum fun easily!)

chysn wrote:

If you want References, and don't mind compiling it yourself, all you need to do is replace the APP_REFS.ino file from the O_C repository, and add the line

DECLARE_APP('R','F', "References", REFS, REFS_isr)

to the OC_apps.ino file. Note that you only need to add this line in one place, not two, as in the O_C firmware. There are no "boring names" in Hemisphere Suite.

If you need additional details on where to put this line, or where to get the file, just let me know.


Great thanks for the info, will try that over the weekend.
Dogma
Timmy wrote:
monokinetic wrote:
Chysn,
version 1.5 sounds exciting, more excellent work by the looks of it, but I haven't upgraded yet because of:

Quote:
* Removed References, the last of the original Ornament and Crime apps.


I used this regularly so far, so that I could run the closed loop tuning thangy to set up my VCOs accurately. It worked nicely as far as I could tell (I assume it was working, right?). Is there any way now to do that in Hemisphere? If not is there any chance to readd it somehow? I'm fine to compile the s/w myself....




How do you like the hemisphere suite? smile

Personally i think its brilliant but hey, i just use these things
Shledge
I think I noticed a bug with 1.5.

When I use the right hemisphere (often used for clock skipper), the o_C will crash if I go to certain apps on the left hemisphere.
Shledge
Also, regarding the original hemisphere app, is there any plans to put any bugfixes towards that? Some of the updates seen on the suite are on applets that are also in the original hemisphere app, and would by handy considering some people want to still use original o_C apps.
chysn
Shledge wrote:
I think I noticed a bug with 1.5.

When I use the right hemisphere (often used for clock skipper), the o_C will crash if I go to certain apps on the left hemisphere.


Thanks! Do you notice any patterns as to which applets cause this?

Quote:
Also, regarding the original hemisphere app, is there any plans to put any bugfixes towards that?


No, I don't plan to ever touch that again. It's not that I don't want to, it's that there's no room to do anything. I had one build that came in at literally 4 bytes under the memory limit. Also, I don't want to.
Shledge
I'll confirm by tomorrow when I get more free time. I've made it happen a couple of times so it should be easily repeatable.
Mashmore
Greetings! I’m on the hunt for a basic sequential switch. I would like to send a couple patterns to trig 1 and trig 2 and switch between them by sending cv to cv 1. I’m digging through the wiki now just wanted to see if this is doable already or could be added?
Something like every time CV A receives a signal above a certain voltage it toggles between trig 1 and trig 2.

Cheers!
chysn
Mashmore wrote:
Greetings! I’m on the hunt for a basic sequential switch. I would like to send a couple patterns to trig 1 and trig 2 and switch between them by sending cv to cv 1. I’m digging through the wiki now just wanted to see if this is doable already or could be added?
Something like every time CV A receives a signal above a certain voltage it toggles between trig 1 and trig 2.

Cheers!


It sounds like a description of the Switch applet. Except it's the other way around; the CV ins are the signals (which certainly can be patterns of triggers, if you like), and the Digital ins switch between them (either sequentially, or with a gate, depending on which input you use).

If you want to trigger the switch based on a voltage threshold, you can pair Switch with the Schmitt Trigger.

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Switch
https://github.com/Chysn/O_C-HemisphereSuite/wiki/Schmitt-Trigger
Mashmore
chysn wrote:
Mashmore wrote:
Greetings! I’m on the hunt for a basic sequential switch. I would like to send a couple patterns to trig 1 and trig 2 and switch between them by sending cv to cv 1. I’m digging through the wiki now just wanted to see if this is doable already or could be added?
Something like every time CV A receives a signal above a certain voltage it toggles between trig 1 and trig 2.

Cheers!


It sounds like a description of the Switch applet. Except it's the other way around; the CV ins are the signals (which certainly can be patterns of triggers, if you like), and the Digital ins switch between them (either sequentially, or with a gate, depending on which input you use).

If you want to trigger the switch based on a voltage threshold, you can pair Switch with the Schmitt Trigger.

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Switch
https://github.com/Chysn/O_C-HemisphereSuite/wiki/Schmitt-Trigger


Right on. Thanks so much. I looked into that but for some reason I read cv and completely wrote off feeding them triggers. Perfect!
Ebotronix
I miss a Doepfer/ Flame style Quantizer App
with clock in and a trigger out
whenever a new (microtonal) quatisation happened!
chysn
Ebotronix wrote:
I miss a Doepfer/ Flame style Quantizer App
with clock in and a trigger out
whenever a new (microtonal) quatisation happened!


I'm not sure I follow completely.

If I understand correctly, you need to know when a quantization change has happened. You can try using Squanch in conjunction with Trending. Squanch has two outputs for a single quantizer channel, so patch one output to your oscillator, and the other output to Trending's CV in, and Trending can send a trigger when the value changes.

Edit: Now that I think about it more, Trending only triggers when the CV changes direction. It might be valuable to have a Trending setting that triggers on any value change.
Ebotronix
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.

I would be better if the Quantizer does the trigger.
so I can use two Quantizers with one OC Module
chysn
Ebotronix wrote:
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.

I would be better if the Quantizer does the trigger.
so I can use two Quantizers with one OC Module


Once the update to Trending is done*, you'd be able to send each channel of Dual Quantizer into a channel of Trending. As long as the droop from the mults isn't too bad, it should work okay.

But, honestly, I'm pretty much quantizered-out.

* Meaning, adding a Change Value option and a dual-channel option
Ebotronix
half the MetaQ app (Single Q) with the clock in (trigger 1 in)
option would be very good!
the trigger in Meta Q Function tranposes (octaves) the the Quant Out.

it would be better if the CV 2 input does the transpose of the scale
(all scale notes) to select the different scale modes,
if the scale has 7 notes it would possible to select 7 different scales (x4)
and the trigger 2 input to select one of the 4 scales.
mxmxmx
Ebotronix wrote:
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.


I'm pretty sure this does work with an external clock (= triggered mode, as opposed to continuous mode ), it's just not that terribly useful in as much as it's basically just duplicating said external clock on the aux/secondary output.
Ebotronix
mxmxmx wrote:
Ebotronix wrote:
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.


I'm pretty sure this does work with an external clock (= triggered mode, as opposed to continuous mode ), it's just not that terribly useful in as much as it's basically just duplicating said external clock on the aux/secondary output.


the trigger in Meta Q app changes the octave of the quantisation!
mxmxmx
Ebotronix wrote:
mxmxmx wrote:
Ebotronix wrote:
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.


I'm pretty sure this does work with an external clock (= triggered mode, as opposed to continuous mode ), it's just not that terribly useful in as much as it's basically just duplicating said external clock on the aux/secondary output.


the trigger in Meta Q app changes the octave of the quntisation!


yeah, in "continuous" mode. that's the point, it could be mapped to some other parameter, but generally speaking there's no use for the external trigger input in that case
Ebotronix
[/quote]

yeah, in "continuous" mode. that's the point, it could be mapped to some other parameter, but generally speaking there's no use for the external trigger input in that case[/quote]

Doepfer A 156 and Flame Tame Machine/ Chord Machine 2 Quantizer Mode does it but they don't generat microtonal scales.
I like the clock in / Quant trigger out Function and I use it a lot!
mxmxmx
Ebotronix wrote:

Doepfer A 156 and Flame Tame Machine/ Chord Machine 2 Quantizer Mode does it but they don't generat microtonal scales.
I like the clock in / Quant trigger out Function and I use it a lot!



I haven't used the A 156 in a long time, so can't remember what it's doing, but whatever it's doing... the Meta Q quantizer has two basic modes, both put out a trigger/gate signal: sample and hold, in which case you need an external clock, obviously. and continuous quantization, in which case there's no need for an external clock, and the gate output tracks the variation of the input/CV signal rather than some clock signal.
Ebotronix
mxmxmx wrote:
Ebotronix wrote:

Doepfer A 156 and Flame Tame Machine/ Chord Machine 2 Quantizer Mode does it but they don't generat microtonal scales.
I like the clock in / Quant trigger out Function and I use it a lot!



I haven't used the A 156 in a long time, so can't remember what it's doing, but whatever it's doing... the Meta Q quantizer has two basic modes, both put out a trigger/gate signal: sample and hold, in which case you need an external clock, obviously. and continuous quantization, in which case there's no need for an external clock, and the gate output tracks the variation of the input/CV signal rather than some clock signal.


I use the Quantizer trigger out function with the ribbon controller.
if the quantizer is clocked the ribbon is ever in time.
zerodivide
I'm a melodic guy when it comes to Ornament and Crime and man, I would so love a Hemisphere app that replaces the last 4 apps in the original O_C (ping pong, viznutzcracker, acid curds, references) with some of the newest applets like Enigma and Darkest Timeline. How difficult would it be to hack the 2 source codes to make some hybrid of all the apps I'm interested in? Trying to get the best of both worlds. Meta-Q for me is essential and I really dont want to keep flipping between 2 firmwares. Hopefully this post doesn't come across as too specific on a selfish level. great work!
chysn
zerodivide wrote:
How difficult would it be to hack the 2 source codes to make some hybrid of all the apps I'm interested in? Trying to get the best of both worlds.


Hey, zerodivide! It's a good question, definitely one worth answering.

Moving Hemisphere Suite apps into the Classic O_C codebase is something I'd put at the "daunting" level of difficulty.

All of the apps I write support MIDI, so you'd need to enable MIDI support in your O_C codebase; that means you might have to take even more stuff out, because MIDI support has some memory overhead.

Once you clear the MIDI memory hurdle, you need to take out as much or more than you put in. Ping Pong and Viznutcracker Sweet are both tiny applications and won't gain you much by taking them out. I'd suggest moving out O_C apps one at a time, because there are a few dependencies, and you'll need to clear up some compile-time errors as you go.

Also important: For Enigma, I took the memory that was dedicated to user chords and used it for the user Turing Machine library instead. So all the code that supports that (in OC_apps.ino) would have to be moved over. At that point, any Classic O_C apps that reference that chord library would cease to compile. If I recall, this was only Acid Curds, though.

There are also a few places where I've changed some things in the main O_C framework. In this case, it's probably just a matter of addressing compile-time errors as they come up. I don't think I did anything that makes the O_C apps fundamentally unusable (except where noted above), but there might be a few Gotchas.

Anyway, it'll be a pain in the ass but not impossible. My best advice would be to make changes slowly and compile often to make sure things aren't getting out of hand.
Kinan
zerodivide wrote:
I'm a melodic guy when it comes to Ornament and Crime and man, I would so love a Hemisphere app that replaces the last 4 apps in the original O_C (ping pong, viznutzcracker, acid curds, references) with some of the newest applets like Enigma and Darkest Timeline. How difficult would it be to hack the 2 source codes to make some hybrid of all the apps I'm interested in? Trying to get the best of both worlds. Meta-Q for me is essential and I really dont want to keep flipping between 2 firmwares. Hopefully this post doesn't come across as too specific on a selfish level. great work!


I have to admit that I would also be very interested in a new hybrid version. I kept the last hybrid version of Hemisphere and didn't follow the last developpement of the app because I couldn't resign myself to drop Quadraturia and Quantermain

Yet I understood that it was not the intention of the programmer to go in this direction and totaly understand it.
chysn
It may or may not be easier to go the other way (that is, starting with the Hemisphere Suite codebase and bringing O_C apps back in). But I removed a lot more than the app files, so you'll need to replace all of the supporting files as well. I was fairly thorough in removing stuff I didn't need.

But it might be a little easier that way. In either case, just give the compiler what it asks for when it complains.
Shledge
Inversely, you could try putting some of the new applets for the original hemisphere firmware which includes o_C apps. May have to remove a few of the main apps to make space, but it could work (unless I'm not aware of any changes to the framework for various applets in hemisphere suite).
REVIVER
BootsNCats...very cool and thoughful addition. Works great with the Trigger Sequencer app.

Would there be a way to "normal" these two particular apps? I.E. When this particular combination is used (maybe Trig on the left, BnC on the right), the two trigger sequencer patterns can trigger the BD/SD accordingly without patching at the front...aside from a clock source to the Trigger Sequence, of course?

Guinness ftw!
chysn
REVIVER wrote:
BootsNCats...very cool and thoughful addition. Works great with the Trigger Sequencer app.

Would there be a way to "normal" these two particular apps? I.E. When this particular combination is used (maybe Trig on the left, BnC on the right), the two trigger sequencer patterns can trigger the BD/SD accordingly without patching at the front...aside from a clock source to the Trigger Sequence, of course?

Guinness ftw!


I’m kind of reluctant to do that sort of thing, where you need to remember exactly some combination. I don’t think it’s much hardship to self patch it.

Another good friend of Boots ‘n Cats is Annular Fusion.

Note that when you start Metronome, it will continue to run when you select something else, and will drive clocks at Digital 1, so you don’t even need an external clock anymore.
REVIVER
chysn wrote:
Note that when you start Metronome, it will continue to run when you select something else, and will drive clocks at Digital 1, so you don’t even need an external clock anymore.


Ah! That's excellent. Thank you!
rayultine
I started the following thread thinking there would be a billion cool ideas.
Where's the love?!
https://www.muffwiggler.com/forum/viewtopic.php?t=208360
qbassa@gmail.com
Is there something like a144 in Hemisphere algorithms?
http://www.doepfer.de/a144.htm
One applet would allow generate peak function on two outputs.
Having two applets simultaneous could allow additional mode of 4 outputs linked (but without loosing the ability to for two independent applets). I see many application of such algorithm. Here is one exemplary use: https://youtu.be/4IffwamiO0A?t=361
Best!
chysn
qbassa@gmail.com wrote:
Is there something like a144 in Hemisphere algorithms?
http://www.doepfer.de/a144.htm
One applet would allow generate peak function on two outputs.
Having two applets simultaneous could allow additional mode of 4 outputs linked (but without loosing the ability to for two independent applets). I see many application of such algorithm. Here is one exemplary use: https://youtu.be/4IffwamiO0A?t=361
Best!


Ooh, I like that! It could be a lot of fun in conjunction with The Bends.
qbassa@gmail.com
I can imagine version with different shapes of the peak (gaussian, linear) or even dual peak function (don't know the use of this yet wink
Shledge
Is there a possibility of a dual unipolar LFO applet? Wouldn't have skew or end triggers, but could have selectable basic waveforms instead. Encoder to select between them and set their speed. CV ins for speed control of both, and trigger ins for reset.
insoul8
Is it possible to have a quantizer on the ornament and crime that has a gate out and operates similar to a channel on the ADDAC207 where without a clock signal present, gates are generated based on note changes with control over gate length? Someone please tell me if I have that wrong.
chysn
Shledge wrote:
Is there a possibility of a dual unipolar LFO applet? Wouldn't have skew or end triggers, but could have selectable basic waveforms instead. Encoder to select between them and set their speed. CV ins for speed control of both, and trigger ins for reset.


https://github.com/Chysn/O_C-HemisphereSuite/wiki/VectorLFO

VectorLFO isn't exclusively a uni-polar LFO, but there will be some uni-polar waveshapes built-in, and you'll be able to create your own very easily.

It's a dual LFO with a twist; the second output, by default, is an 50/50 mix of both LFOs.
chysn
insoul8 wrote:
Is it possible to have a quantizer on the ornament and crime that has a gate out and operates similar to a channel on the ADDAC207 where without a clock signal present, gates are generated based on note changes with control over gate length? Someone please tell me if I have that wrong.


I'm not familiar with the ADDAC207, but you can achieve a similar result in Hemisphere Suite using Squanch and Trending. Trending has been updated with a change detector mode, in which it sends a trigger on a voltage change. So you'd patch one output of Squanch to your VCO, and the other output to the change detector in Trending. There's no control of gate length here, it's just a trigger, but there are plenty of ways to make that conversion elsewhere.
insoul8
chysn wrote:
insoul8 wrote:
Is it possible to have a quantizer on the ornament and crime that has a gate out and operates similar to a channel on the ADDAC207 where without a clock signal present, gates are generated based on note changes with control over gate length? Someone please tell me if I have that wrong.


I'm not familiar with the ADDAC207, but you can achieve a similar result in Hemisphere Suite using Squanch and Trending. Trending has been updated with a change detector mode, in which it sends a trigger on a voltage change. So you'd patch one output of Squanch to your VCO, and the other output to the change detector in Trending. There's no control of gate length here, it's just a trigger, but there are plenty of ways to make that conversion elsewhere.


Cool! Thanks. I'll check it out.
chysn
insoul8 wrote:
Cool! Thanks. I'll check it out.


Please note that the update to Trending will be in the 1.6 release, which is scheduled to be released tomorrow*. If you can compile the firmware yourself, the new version is available in the repo now.

* But except my hard drive crashed last night, and there were a few things left to do. So I might have to push it back to Sunday. Not sure yet.
jahreed
this project is the most exciting thing in euro IMHO

O_C was a default must have before the hemisphere suite which has expanded the capabilities 4x

kudos!!!
chysn
jahreed wrote:
this project is the most exciting thing in euro IMHO


Very kind of you to say, thank you.

I'm sorry to announce that I've decided to delay the release of Hemisphere Suite 1.6 by a few days. I had planned to wrap the code yesterday but, instead, my hard drive crashed. So the time that I wanted to devote to finishing things up had to go into troubleshooting and making sure my backups are in order. They are, but I still have to find the time to go buy a new hard drive and restore.

I could release what's in the repo now, but I really wouldn't feel comfortable half-assing it. There are some things that must be done.

So now, I've got a target of Monday, to give me time to restore my Macbook and put the finishing touches on the release.
chysn
qbassa@gmail.com wrote:
I can imagine version with different shapes of the peak (gaussian, linear) or even dual peak function (don't know the use of this yet wink


To make the delay up to you, I'm adding a new applet that does this. It's essentially an LFO, but with no frequency, so you can control its phase in degrees via CV. And you'll be able to link up to four of them together for morph control like the A-144. Since it's based on the Vector Oscillator that's used for all the other modulators in this release, you can create your own waveshapes, rather than being tied to a triangle.
chysn
https://youtu.be/WrENiAQHhq0
Leonard de Leonard
So cool, can t wait to test the Vector stuff (and also the Boots ‘n Cats )
Thanks a lot for all your work!
chysn
I'm pleased to announce the release of Hemisphere Suite 1.6. Hex files for normal and 180-degree flipped screen are available here:

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.6

Updates for the release are as follows:

* Added Metronome applet, a master clock system that can run in the background to clock other applets, so that Hemisphere Suite sequencers (and other applets) no longer require an external clock. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Metronome)
* Tuner: Now allows Tuner to function in the right hemisphere when the screen is flipped.
* Added Vector Waveform Editor, a full app that allows you to build and audition Vector Waveforms. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Waveform-Editor)
* Added VectorLFO, a dual LFO applet that uses custom and preset Vector Waveforms. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/VectorLFO)
* Added VectorEG, a dual uni-polar envelope generator applet that uses custom and preset Vector Waveforms. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/VectorEG)
* Added VectorMod, a dual bi-polar modulator that uses custom and preset Vector Waveforms as either one-shot or cycling modulation sources. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/VectorMod)
* Added VectMorph, a morph controller applet that uses custom and preset Vector Waveforms. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/VectMorph)
* Added Boots 'n Cats (BootsNCat), a simple bass/snare drum synthesizer with stereo audio outputs. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/BootsNCat)
* Added Setup/About app, which allows easy access to calibration and EEPROM reset functions, and checking the firmware version. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Setup)
* Removed bootup splash screen so that the module is immediately ready for use on power-up.
* Trending: Added dual operation, so it now accepts input on CV 1 and CV 2, and each operation and output is independent.
* Trending: Added change detection, which emits a trigger when a signal has changed.
* TrigSeq/TrigSeq16: Fixed issue that caused the first step to not play
* "A"SR: Fixed scale selection buffer overread crash

Since this release reconfigures EEPROM memory, any saved data will be removed. Please make sure to save anything you want to keep via sysex.

Thank you all for your continued support.
4eyez
Exciting stuff! Really excited about the syncable LFO! Many thanks!
starthief
A VectorLFO driving a VectorMorph can get pretty crazy SlayerBadger!
AnalogAssailant
Quote:
I'm pleased to announce the release of Hemisphere Suite 1.6.


awesome, installing this tonight. Thank you!
chysn
starthief wrote:
A VectorLFO driving a VectorMorph can get pretty crazy SlayerBadger!


A good insight, I'll have to try that.
tFunk
I just installed Hemisphere Suite v1.6 on my O&C, and I might have found a bug, or I'm overlooking something.

These are the steps to reproduce it:

- Clock divider app (on left hemisphere)
- Clock from Batumi square wave to TR1 IN

Issue: OUT A doens't output anything when clock multiplication is set. Clock division works fine.

Great job chysn btw!
bemushroomed
Nice update, thanks! w00t
chysn
tFunk wrote:
Issue: OUT A doens't output anything when clock multiplication is set. Clock division works fine.


Thank you for reporting this. Yes, it looks like I broke the clock multiplier during a refactoring spree. I'll have a new hex file out today.

Update: If you compile from Arduino IDE, this fix is available on GitHub now. I'm going to see if any other issues arise today, and then I'll update the hex files this evening. Note that this issue also affects VectorLFO sync.
qbassa@gmail.com
1.6 is realy a major upgrade! Chapeau bas for developement work!. One idea - would it be possible to change one output of VectorXXX to end of cycle trigger after every envelope stage?
qbassa@gmail.com
And thanks for great implementation of a144 based morph - it is truly unbelievable considering request for this feature was submitted 4 days ago...
StateAzure
Just loaded Hemisphere onto my uo_C via method a, and it seems to be fine but sometimes it goes into calibration mode? My module was already calibrated, so I'm wondering why it's doing this. If I reboot the system it goes back to normal operation.
matzemoerk
Same here
qbassa@gmail.com
I am afraid I've found a bug also. TrigSeq seems to add one step to both sequences (tried in right panel, clock is taken from Metronome). Best!
qbassa@gmail.com
And it is the same with TrigSeq16
chysn
Thanks for the bug reports! 1.6A corrects these things:

* The clock multiplier and sync wasn't working
* Trigger Sequencers were adding an extra phantom step at the end

Additionally, 1.6A reduces (or possibly eliminates) the probability of getting a random intermittent calibration screen on bootup. I haven't experienced this, but I did see an intermittent EEPROM reset screen, and slightly increasing the wait time for button presses seemed to eliminate it. So give it a try, and let me know if it keeps happening. Since calibration can now be done at any time from the Setup app, the worst-case scenario is that I just remove the encoder press.

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.6

qbassa@gmail.com wrote:
1.6 is realy a major upgrade! Chapeau bas for developement work!. One idea - would it be possible to change one output of VectorXXX to end of cycle trigger after every envelope stage?


With only two outputs for a dual LFO/modulator/morpher/EG, I'm not sure how to elegantly achieve that. But I'm considering doing a more full-featured single LFO, so this sort of feature might find its way in.
qbassa@gmail.com
That feature I was thinking about was something like Single VectorLFO/EG with EOS trigger (instead of second channel). Best!
lukimb
my cv in's stopped working, globally.
no signal in scope, pong, asr, quant
1,6A
Is it just me?
lukimb
flashed orginal oc firmware and cv in works fine
chysn
lukimb wrote:
my cv in's stopped working, globally.
no signal in scope, pong, asr, quant
1,6A
Is it just me?

flashed orginal oc firmware and cv in works fine


Okay, that's interesting. Did you use Hemisphere Suite prior to 1.6? If not, can you please try out 1.5 or earlier and let me know what happens?

Also: Yeah, anybody else?

I need to mull this one over.
lukimb
1.5 worked fine for me
chysn
lukimb wrote:
1.5 worked fine for me


Thanks. Try resetting the EEPROM (note: this will not wipe out your calibration). There are two ways to do this:

* Hold the Up and Down buttons during power-up, or
* Go to the Setup/About app (bottom of the main menu)

Do the inputs work after reset?
chysn
Hemisphere Suite 1.6B is now available. It makes a couple fixes to VectorEG:

(1) When a waveform was changed, the output would be the level of the final segment rather than 0V.

(2) The release stage for Square/Pulse waveforms was not handled correctly, and the waveform would never reach the final segment.
starthief
chysn wrote:
Hemisphere Suite 1.6B is now available. It makes a couple fixes to VectorEG:

(1) When a waveform was changed, the output would be the level of the final segment rather than 0V.

(2) The release stage for Square/Pulse waveforms was not handled correctly, and the waveform would never reach the final segment.


thumbs up

I'm not seeing the hex file?
swhic
hex? thanks
REVIVER
Not sure if it's a bug or if it's me, but I can't seem to get the tuner to respond to the input on Digital 2 unless I navigate to the tuner app on the left side first.



hmmm.....
chysn
swhic wrote:
hex? thanks


Sorry, the upload didn't take last night for some reason. It's here now:

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.6B
chysn
REVIVER wrote:
Not sure if it's a bug or if it's me, but I can't seem to get the tuner to respond to the input on Digital 2 unless I navigate to the tuner app on the left side first


Thanks! This is fixed in the 1.6B hex file.
REVIVER
Yes it is! Thanks so much for the fast turnaround on that.

BTW, the About page labels it as 1.6A still...just an FYI.
chysn
REVIVER wrote:
Yes it is! Thanks so much for the fast turnaround on that.

BTW, the About page labels it as 1.6A still...just an FYI.


Okay, thanks! I probably won't sweat that unless something else serious comes up.
lukimb
chysn wrote:
lukimb wrote:
my cv in's stopped working, globally.
no signal in scope, pong, asr, quant
1,6A
Is it just me?

flashed orginal oc firmware and cv in works fine


Okay, that's interesting. Did you use Hemisphere Suite prior to 1.6? If not, can you please try out 1.5 or earlier and let me know what happens?

Also: Yeah, anybody else?

I need to mull this one over.



problem disappered after flashing 1.6B wink
chysn
lukimb wrote:
chysn wrote:
lukimb wrote:
my cv in's stopped working, globally.
no signal in scope, pong, asr, quant
1,6A
Is it just me?

flashed orginal oc firmware and cv in works fine


Okay, that's interesting. Did you use Hemisphere Suite prior to 1.6? If not, can you please try out 1.5 or earlier and let me know what happens?

Also: Yeah, anybody else?

I need to mull this one over.



problem disappered after flashing 1.6B wink


Glad to hear it! I have exactly zero plausible theories about what was going on there.
sushiluv
the wiki is confusing, for example for the vector LFO you wrote

Quote:
Outputs: Output A/C is the signal for LFO 1. Output B/D is the signal for LFO 2 mixed with the signal for LFO 1, with the mix based on the input to CV 2.


but the hemisphere apps only have two outputs, i´m confused.

Thanks for the top work, love hemipshere.
chysn
sushiluv wrote:
the wiki is confusing, for example for the vector LFO you wrote

Quote:
Outputs: Output A/C is the signal for LFO 1. Output B/D is the signal for LFO 2 mixed with the signal for LFO 1, with the mix based on the input to CV 2.


but the hemisphere apps only have two outputs, i´m confused.

Thanks for the top work, love hemipshere.


Sorry for the confusion. The terms "Output A/C" and "Output B/D" are conventions used throughout the documentation.

The Ornament and Crime has four outputs, traditionally labelled A, B, C, and D. The left hemisphere uses Outputs A and B, and the right hemisphere uses Outputs C and D. So "Output A/C" is just a shortcut for saying "Output A if the applet is in the left hemisphere or Output C if the applet is in the right hemisphere."
rayultine
chysn,
would it be useful to have an applet that could provide noise up to the highest freq capable by the o_C hardware? It might be fun to add a simple lp/bp/hp filter (resonant?) Or can the random function in Calculate give white noise when driven at fast enough rates?
chysn
rayultine wrote:
chysn,
would it be useful to have an applet that could provide noise up to the highest freq capable by the o_C hardware? It might be fun to add a simple lp/bp/hp filter (resonant?) Or can the random function in Calculate give white noise when driven at fast enough rates?


The O_C's interrupt runs at about 16.6kHz, so that's the highest frequency the O_C firmware is capable of. Hemisphere inherits that system, and the random source in Calculate runs at that rate (as long as it's not being clocked), and it's totally possible to use it as a white noise audio signal.
qbassa@gmail.com
How CV recorder goes from one recorded value to another. Is it linear transition?
chysn
qbassa@gmail.com wrote:
How CV recorder goes from one recorded value to another. Is it linear transition?


It just steps from one value to the next. There’s no slew applied.
Shledge
I just updated, really liking the vector LFO. I may tweak it that the 2nd output doesn't initially mix both waveforms, but other than that it's brilliant.
chysn
Shledge wrote:
I just updated, really liking the vector LFO. I may tweak it that the 2nd output doesn't initially mix both waveforms, but other than that it's brilliant.


Yeah, that one was a tough call. I went in favor of the somewhat unconventional, assuming that people usually have basic LFOs covered pretty well. It's actually kind of nice to have a "complex" LFO available without extra patching.

You can reverse it by changing HEM_VectorLFO.ino, line 53, to:

atten1 = HEMISPHERE_MAX_CV - constrain(atten1, 0, HEMISPHERE_MAX_CV);
rayultine
chysn wrote:
rayultine wrote:
chysn,
would it be useful to have an applet that could provide noise up to the highest freq capable by the o_C hardware? It might be fun to add a simple lp/bp/hp filter (resonant?) Or can the random function in Calculate give white noise when driven at fast enough rates?


The O_C's interrupt runs at about 16.6kHz, so that's the highest frequency the O_C firmware is capable of. Hemisphere inherits that system, and the random source in Calculate runs at that rate (as long as it's not being clocked), and it's totally possible to use it as a white noise audio signal.


But would a dedicated noise applet be useful to you and others instead of tying up Calculate and an external audio source or "H"FO? To be able to add a simple filter to that would be sublime. Maybe two differently filtered Noise applets on each side would make a weird hihat for someone.
chysn
rayultine wrote:
chysn wrote:
rayultine wrote:
chysn,
would it be useful to have an applet that could provide noise up to the highest freq capable by the o_C hardware? It might be fun to add a simple lp/bp/hp filter (resonant?) Or can the random function in Calculate give white noise when driven at fast enough rates?


The O_C's interrupt runs at about 16.6kHz, so that's the highest frequency the O_C firmware is capable of. Hemisphere inherits that system, and the random source in Calculate runs at that rate (as long as it's not being clocked), and it's totally possible to use it as a white noise audio signal.


But would a dedicated noise applet be useful to you and others instead of tying up Calculate and an external audio source or "H"FO? To be able to add a simple filter to that would be sublime. Maybe two differently filtered Noise applets on each side would make a weird hihat for someone.


Useful to me, personally? Not really. Noise is one thing that I don't have a problem getting, nor am I very picky about it when I do need it. I'm not much of a noise connoisseur, is what I'm saying.
rayultine
gotcha!
qbassa@gmail.com
Do you consider adding linear mode? It is not the same as having slew limiter with constant change speed afterwards and seams quite easy to implement. Quadratic / Polynominal curve that goes through recorded points would be more complicated. What i think about as an adavnced feature would be saving point data / recordings in the form of VectorXXX presets and possible to use as VectorEg/VectorMod. Best!
chysn
qbassa@gmail.com wrote:
What i think about as an adavnced feature would be saving point data / recordings in the form of VectorXXX presets and possible to use as VectorEg/VectorMod. Best!


That's an interesting idea. Translating real-time movement into level/time segments would be a real challenge, but it's something to think about.
adh82
Been really enjoying Hemisphere.
Thanks so much for the hard work @chysn, much appreciated!

Got a couple of questions.

1. Going through the apps I couldn't see a quantizer that operates like the meta q quantizer on o&c, meaning a quantizer that spits out a gate every time it quantises. Is there any apps that can do this?

2. Coming from the o&c I love seeing the graphics on the screen and the graphics in Hemisphere are awesome! Seeing there is no screen saver, is there an option to stop the screen turning off or change the turn off time. Would really love to see the graphics as a reference while patching.
chysn
adh82 wrote:
1. Going through the apps I couldn't see a quantizer that operates like the meta q quantizer on o&c, meaning a quantizer that spits out a gate every time it quantises. Is there any apps that can do this?


This would be done with a combination of Squanch and Trending. One output of Squanch to your VCO, and the other to one channel of Trending, set to ChgValue. If you want to do this on two channels, you can use Dual Quantizer and stackables instead of Squanch.

Quote:
2. Coming from the o&c I love seeing the graphics on the screen and the graphics in Hemisphere are awesome! Seeing there is no screen saver, is there an option to stop the screen turning off or change the turn off time. Would really love to see the graphics as a reference while patching.


Go to the calibration utility (by holding down the left encoder button on startup, or from the Setup/About app) and scroll through to the end to set the screen blank time in minutes, up to two hours.
hinterlands303
I'm not sure if this is a bug or not - but in Trending I'm getting a lot of triggers when there is no value change in ChgValue mode. Even with nothing plugged into the VC input it sends out what seems to be random bursts of triggers (although usually at a regular tempo stream, which then either increases or decreases based on the sensitivity). Is there something I'm missing here? Even with sensitivity at it's lowest I'm still getting a ton of triggers when there is no change.

I'm trying to use it as mentioned above to get the doepfer a-156 and meta q behavior.
chysn
hinterlands303 wrote:
I'm not sure if this is a bug or not - but in Trending I'm getting a lot of triggers when there is no value change in ChgValue mode. Even with nothing plugged into the VC input it sends out what seems to be random bursts of triggers (although usually at a regular tempo stream, which then either increases or decreases based on the sensitivity). Is there something I'm missing here? Even with sensitivity at it's lowest I'm still getting a ton of triggers when there is no change.

I'm trying to use it as mentioned above to get the doepfer a-156 and meta q behavior.


This could be several things. Your ADCs might be getting some noise that's triggering the change. I find that it's a lot more reliable if I have a signal that's not too close to 0V.

One thing you can do is check the Scope applet's CV 2 input. Trending detects changes of 1/8 semitone, or about .01V. So if the Scope's volt reading is fluctuating a lot, then you've got some noise going on in the ADC, or maybe RF interference, or not-so-great cables, or something.

I have a lot of RF interference around me, due to a 50kW AM station less than a mile away, so I get behavior similar to what you describe when nothing is plugged into the input. It's intermittent, though.

Update: I've been thinking about making the change threshold .02V, which would filter out a good deal of noise. I have to see how it affects other things, too.
qbassa@gmail.com
And what about this linear transition thing? wink With linear transmission CV recorder could be used as smooth multitrack (4 track on one O_C) automation tool where you set 4 voltages - lets say tuning timbre of the synth voice with some FM, when it is fine you send a trigger to record a "point" which uses 2 x "Cv recorder", then you set another 4 voltages and when it sounds nice you record another point. When the sequence is complete and playback starts linear transitions "animates" between points on every track.
chysn
qbassa@gmail.com wrote:
And what about this linear transition thing? wink With linear transmission CV recorder could be used as smooth multitrack (4 track on one O_C) automation tool where you set 4 voltages - lets say tuning timbre of the synth voice with some FM, when it is fine you send a trigger to record a "point" which uses 2 x "Cv recorder", then you set another 4 voltages and when it sounds nice you record another point. When the sequence is complete and playback starts linear transitions "animates" between points on every track.


I'll probably review the CV recorder again at some point. There's always a bit of a gulf between what can be done and what might get too complex from an interface standpoint, so it would likely by a separate "smoothed" CV applet.

It will depend on how much space is left after the Neural Network is finished. Memory is getting pretty tight in there now, and I think that there are already plenty of ways to smooth out modulation sources.

I think the probability of me adding linear interpolation to some form of CV recorder is 90%.
hinterlands303
chysn wrote:


This could be several things. Your ADCs might be getting some noise that's triggering the change. I find that it's a lot more reliable if I have a signal that's not too close to 0V.

One thing you can do is check the Scope applet's CV 2 input. Trending detects changes of 1/8 semitone, or about .01V. So if the Scope's volt reading is fluctuating a lot, then you've got some noise going on in the ADC, or maybe RF interference, or not-so-great cables, or something.

I have a lot of RF interference around me, due to a 50kW AM station less than a mile away, so I get behavior similar to what you describe when nothing is plugged into the input. It's intermittent, though.

Update: I've been thinking about making the change threshold .02V, which would filter out a good deal of noise. I have to see how it affects other things, too.


That could be it - it does sound like random triggers so noise would make sense. I’ll test it with the scope applet later and see. I guess I could try it in a different case as well and see. Thanks for your help!
qbassa@gmail.com
chysn wrote:
qbassa@gmail.com wrote:
And what about this linear transition thing? wink With linear transmission CV recorder could be used as smooth multitrack (4 track on one O_C) automation tool where you set 4 voltages - lets say tuning timbre of the synth voice with some FM, when it is fine you send a trigger to record a "point" which uses 2 x "Cv recorder", then you set another 4 voltages and when it sounds nice you record another point. When the sequence is complete and playback starts linear transitions "animates" between points on every track.


I'll probably review the CV recorder again at some point. There's always a bit of a gulf between what can be done and what might get too complex from an interface standpoint, so it would likely by a separate "smoothed" CV applet.

It will depend on how much space is left after the Neural Network is finished. Memory is getting pretty tight in there now, and I think that there are already plenty of ways to smooth out modulation sources.

I think the probability of me adding linear interpolation to some form of CV recorder is 90%.


Great! Thank you
hinterlands303
hinterlands303 wrote:
chysn wrote:


This could be several things. Your ADCs might be getting some noise that's triggering the change. I find that it's a lot more reliable if I have a signal that's not too close to 0V.

One thing you can do is check the Scope applet's CV 2 input. Trending detects changes of 1/8 semitone, or about .01V. So if the Scope's volt reading is fluctuating a lot, then you've got some noise going on in the ADC, or maybe RF interference, or not-so-great cables, or something.

I have a lot of RF interference around me, due to a 50kW AM station less than a mile away, so I get behavior similar to what you describe when nothing is plugged into the input. It's intermittent, though.

Update: I've been thinking about making the change threshold .02V, which would filter out a good deal of noise. I have to see how it affects other things, too.


That could be it - it does sound like random triggers so noise would make sense. I’ll test it with the scope applet later and see. I guess I could try it in a different case as well and see. Thanks for your help!


Update - i was able to fix it by offsetting the incoming voltage. Like you said it seems that when it's around zero volts it gets unreliable. Of course doing this also transposes the pitch when using it to generate arpeggios in combination with the quantizer.
adh82
Thanks so much for that. Works a treat!
Just had a play on the cv recorder, really fun with my Koma kommander.
Wondering if the loop length could longer? I turned up the clock speed to give it some more resolution but then of course the loop length is reduced.
Also having the output quantized would be amazing for looping up melodic sequences.
chysn
adh82 wrote:
Thanks so much for that. Works a treat!
Just had a play on the cv recorder, really fun with my Koma kommander.
Wondering if the loop length could longer? I turned up the clock speed to give it some more resolution but then of course the loop length is reduced.
Also having the output quantized would be amazing for looping up melodic sequences.


There probably will be some upgrades to the CV recorder. There's a request outstanding for value interpolation, and I'm not entirely happy with the interface (e.g., having to record both channels at once).

The development roadmap is that the November 30 release will be the final release with significant new additions, and then I'll revisit and refine things for a while. Some of these refinements will involve memory usage, which should free up more space for things like the CV recorder.
matzemoerk
I was running two shift registers using master clock forwarding. Just an Idea for a extension: Clock division / multiplication for one of the two hemispheres.

That would have saved me one hemisphere of the second o_c for further usage ;-). What do you think? Possible?
chysn
hinterlands303 wrote:
Update - i was able to fix it by offsetting the incoming voltage. Like you said it seems that when it's around zero volts it gets unreliable. Of course doing this also transposes the pitch when using it to generate arpeggios in combination with the quantizer.


The 1.6C hex file is available now, and this increases the change detection threshold to .02V (1/4 semitone), which should reduce the effect of noise in the ADC. Hopefully you'll be able to use it now without sacrificing your pitch.
indifference point
As always thanks immensely for all your work on this.

Would it be possible to save matrixes in the Carpeggio?

Also I'm sure this has been asked before but can I give you money somewhere?
chysn
matzemoerk wrote:
I was running two shift registers using master clock forwarding. Just an Idea for a extension: Clock division / multiplication for one of the two hemispheres.


It's a good idea, but I think the interface elements are pretty much used up. That is, I can't think of a way to provide an interface for that sort of feature without it being clunky and/or menu-divey.

indifference point wrote:
Would it be possible to save matrixes in the Carpeggio?


Unfortunately, it's not possible with the Hemisphere architecture. Since the app storage memory has to be allocated in advance, I only allocated 32 bits for each active hemisphere for EEPROM storage. A saved matrix would need at least 96 bits. That's the same reason that bigger things like CV recorder and Lo-Fi Tape data don't get saved.

I think Carpeggio really deserves a whole module to itself, although Rene II is almost there with its Z-axis.

Quote:
Also I'm sure this has been asked before but can I give you money somewhere?


I appreciate the thought, but it's not necessary, and I don't have any mechanisms set up for that. It's a labor of love and/or obsession. You can help by reporting bugs or offering your insights. Thanks, though!
chysn
matzemoerk wrote:
I was running two shift registers using master clock forwarding. Just an Idea for a extension: Clock division / multiplication for one of the two hemispheres.


Pro tip: If you're running two shift registers anyway, you might as well use Enigma instead. You'll get four Turing Machines, the ability to set clock division per track, and access to all the fancy output types (trigger, gate, etc.).
matzemoerk
Quote:
Pro tip: If you're running two shift registers anyway, you might as well use Enigma instead. You'll get four Turing Machines, the ability to set clock division per track, and access to all the fancy output types (trigger, gate, etc.).


Thanks! I will have a look.
uebl
This is awesome, and it keeps getting better and better. Yesterday night I made my own custom Scala scale, converted it using BeigeMaze's web-based converter and send the resulting SysEx file to the Scale Editor app on my uO_C without any problems. Thank you so much for this!
chysn
uebl wrote:
This is awesome, and it keeps getting better and better. Yesterday night I made my own custom Scala scale, converted it using BeigeMaze's web-based converter and send the resulting SysEx file to the Scale Editor app on my uO_C without any problems. Thank you so much for this!


I'm glad someone besides me enjoys that feature. The project really got me interested in alternate tunings.

It's worth mentioning here that Manuel Op de Coul gave me permission to redistribute the Scala archive in Hemisphere Suite SysEx format. I converted all 3000 or so of the scales that are compatible with the O_C's tuning system, so you just need to pick a .syx file and dump it over to Scale Editor:

http://www.beigemaze.com/downloads/HS_Scala_Archive.zip

Random other note: Last night I learned that Sysex Base for iPad is better than any librarian software for desktop MacOS.
uebl
Yeah, I already got that archive when you first posted it, but 3,000 scales is really overwhelming. I therefore prefer to just define my "own" nanners
swhic
Is there a way to interface a midi keyboard with Hemisphere (no computer)?
chysn
swhic wrote:
Is there a way to interface a midi keyboard with Hemisphere (no computer)?


Yeah, you can use a MIDI host box, like this one:

https://www.amazon.com/midiplus-USB-MIDI-Host-Controller/dp/B074NGTGHN

Kenton also makes one, and you can also use Oplab if you can find one.

As long as the host device is class-compliant, everything should work pretty smoothly.
swhic
Thanks, do u know if a keyboard>iphone/ipad>o_c will work?
chysn
swhic wrote:
Thanks, do u know if a keyboard>iphone/ipad>o_c will work?


iPad's driver is class-compliant, and I do use iPad with Sysex Base for Hemisphere Suite system exclusive storage. That works great.

I also have a KMI Q-Nexus, which works fine with my iPad.

However, I'm not sure how the keyboard > iPad > O_C connection would work, physically, since the iPad has only a single port. If anyone has ideas about that, it would be most welcome.
chysn
I want to take a moment to thank Misk for building a µO_C for me, to replace my regular one. He did an incredible job on the build, nice and clean, and works great! I'm humbled by his generosity.

If you need a µO_C built, hit up Misk!



The panel is a custom design from Front Panel Express.[/img]
Godphaser
chysn wrote:
The panel is a custom design from Front Panel Express.[/img]


Sorry to hijack the thread real quick but how much did it cost you?
chysn
Godphaser wrote:
chysn wrote:
The panel is a custom design from Front Panel Express.[/img]


Sorry to hijack the thread real quick but how much did it cost you?


Front Panel Express's prices are obscene. That one was about $45 USD, whereas you can get a Magpie sandblasted panel for $28 or a PCB panel for as little as $10.
pines
45 bucks is a bit steep, but that is one handsome panel.
a100user
Shay sent me my 1uO_c loaded with Hemispheres - very cool.

My first exposure to Hemispheres and all I can say is, Wonderful

chysn
a100user wrote:
Shay sent me my 1uO_c loaded with Hemispheres - very cool.

My first exposure to Hemispheres and all I can say is, Wonderful



Thanks! That's a cool module. I keep thinking about building a new case with a 1U row, but I'm always intimidated by the fact that there doesn't seem to be much of a standard size. Building a case is a summertime activity, anyway, for me. Maybe next year.

I see you've got Bow Tie Sequential running. Do you have a Boss Bow Tie?
a100user
chysn wrote:
a100user wrote:
Shay sent me my 1uO_c loaded with Hemispheres - very cool.

My first exposure to Hemispheres and all I can say is, Wonderful



Thanks! That's a cool module. I keep thinking about building a new case with a 1U row, but I'm always intimidated by the fact that there doesn't seem to be much of a standard size. Building a case is a summertime activity, anyway, for me. Maybe next year.

I see you've got Bow Tie Sequential running. Do you have a Boss Bow Tie?


Not yet, this just how the module powered up.

Two of my Euro cases have 1U rails and they are full except the gap seen in the picture. So many great utilities available to choose from.

Have a look on modular grid here for an idea of what is available and then you can decide it it is worth building 1U into your case. Not that there are two 1U standards, the Pulp Logic version which I have and the Intellijel size.

Link
rayultine
chysn wrote:
I keep thinking about building a new case with a 1U row, but I'm always intimidated by the fact that there doesn't seem to be much of a standard size.


Not to turn the Hemispheres thread into another annoying thread about 1U sizes, but there's effectively two sizes - Intellijel and everything else. Intellijel's specs are slightly smaller, so if you build a 1U row to the bigger specification, you can fit all 1U modules in. Though Intellijel will leave a small gap, which can be bad if you have wiggler OCD. Personally, I haven't been super interested in the Intellijel offerings. I have 120hp of PulpLogic and it's sooooo damn useful.
Ebotronix
chysn wrote:
Ebotronix wrote:
because I need two quantizers,
I had tried it with the OC Meta Q app
but it doesn't work with external clock in.

I would be better if the Quantizer does the trigger.
so I can use two Quantizers with one OC Module


Once the update to Trending is done*, you'd be able to send each channel of Dual Quantizer into a channel of Trending. As long as the droop from the mults isn't too bad, it should work okay.

But, honestly, I'm pretty much quantizered-out.

* Meaning, adding a Change Value option and a dual-channel option


thanks for the update ,now the Tending does the Trigger.
where can I find the Quantizer scale intervals?
chysn
Ebotronix wrote:
thanks for the update ,now the Tending does the Trigger.
where can I find the Quantizer scale intervals?


If you mean scale masks, the closest thing in Hemisphere is in Scale Duet, which lets you set two independent masks for a semitone scale.
Ebotronix
chysn wrote:
Ebotronix wrote:
thanks for the update ,now the Tending does the Trigger.
where can I find the Quantizer scale intervals?


If you mean scale masks, the closest thing in Hemisphere is in Scale Duet, which lets you set two independent masks for a semitone scale.


I mean the steps of the scales
example
C major
C,D,E,F,G,A,B
there are some scales with other division than half or whole tones.
chysn
Ebotronix wrote:

I mean the steps of the scales
example
C major
C,D,E,F,G,A,B
there are some scales with other division than half or whole tones.


Gotcha. Dual Quantizer has settings for both the scale and the root.
Ebotronix
chysn wrote:
Ebotronix wrote:

I mean the steps of the scales
example
C major
C,D,E,F,G,A,B
there are some scales with other division than half or whole tones.


Gotcha. Dual Quantizer has settings for both the scale and the root.


yes I use this app but where can I find the tones from the different scales?
there are more than 70 scale , does a list exist ?
( in cent, 100 cent is a halftone)
chysn
Ebotronix wrote:
chysn wrote:
Ebotronix wrote:

I mean the steps of the scales
example
C major
C,D,E,F,G,A,B
there are some scales with other division than half or whole tones.


Gotcha. Dual Quantizer has settings for both the scale and the root.


yes I use this app but where can I find the tones from the different scales?
there are more than 70 scale , does a list exist ?
( in cent, 100 cent is a halftone)


Here you go: http://ornament-and-cri.me/predefined_scales/
Ebotronix
thanks chysn!
you are the best!
sushiluv
hi chysn,

would it be possible to build an applet which delays CV instead of a trigger or gate signal?

Madrona Labs Aalto, a softsynth, has the possibility to delay the output of it´s sequencer, it´s very nice for interesting effects.
chysn
sushiluv wrote:
hi chysn,

would it be possible to build an applet which delays CV instead of a trigger or gate signal?

Madrona Labs Aalto, a softsynth, has the possibility to delay the output of it´s sequencer, it´s very nice for interesting effects.


There are two CV delay-type options in Hemisphere Suite.

(1) Try out the "A"SR applet, which is an analog shift register implementation. That's probably the best way to do a CV delay on a device with such limited memory. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/%22A%22SR)

(2) For higher sampling rate (1 second at 2kHz), try LoFi Tape, which works with CV even better than it works with audio. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/LoFi-Tape)
sushiluv
thanks, great to know!

i hesitate to ask, but how would i set either of them up to delay incoming cv so i can control the delay? smile
chysn
sushiluv wrote:
thanks, great to know!

i hesitate to ask, but how would i set either of them up to delay incoming cv so i can control the delay? smile


For “A”SR, this is the index value. It can be set from the panel, or modulated with CV2.
sushiluv
chysn wrote:
sushiluv wrote:
thanks, great to know!

i hesitate to ask, but how would i set either of them up to delay incoming cv so i can control the delay? smile


For “A”SR, this is the index value. It can be set from the panel, or modulated with CV2.


Thank you, works perfect!
vantablack
chysn wrote:
swhic wrote:
Thanks, do u know if a keyboard>iphone/ipad>o_c will work?


iPad's driver is class-compliant, and I do use iPad with Sysex Base for Hemisphere Suite system exclusive storage. That works great.

I also have a KMI Q-Nexus, which works fine with my iPad.

However, I'm not sure how the keyboard > iPad > O_C connection would work, physically, since the iPad has only a single port. If anyone has ideas about that, it would be most welcome.


uum.. would usb connection from ipad>"camera"connector adapter>usb hub>o_c work?
chysn
vantablack wrote:
uum.. would usb connection from ipad>"camera"connector adapter>usb hub>o_c work?


Maybe. I have a USB hub around somewhere, but I can't find it right now.
cafesinleche
been loving hemisphere immensely and really opens up the o_c in a wonderful way.

I've just started running into a little interesting problem. I'm running a couple CV sources into two doepfer precision adders each of which i plug into one of the dual quantizers. However if I change the cv from one of the precision adders it affects both sides of the dual quantizer. If i were to plug cv into the dual quantizer straight, I dont have this problem and the signals into the quantizers are kept seperate. it only happens when it's run through the precision adders.

any tips on what's going on anyone?
chysn
cafesinleche wrote:
been loving hemisphere immensely and really opens up the o_c in a wonderful way.

I've just started running into a little interesting problem. I'm running a couple CV sources into two doepfer precision adders each of which i plug into one of the dual quantizers. However if I change the cv from one of the precision adders it affects both sides of the dual quantizer. If i were to plug cv into the dual quantizer straight, I dont have this problem and the signals into the quantizers are kept seperate. it only happens when it's run through the precision adders.

any tips on what's going on anyone?


Does this affect only adjacent inputs? Let's say you put the Dual Quantizer in both hemispheres and use the most physically-distant two inputs (input 1 of the left hemisphere and input 2 of the right hemisphere). Is there still this crosstalk?

Do you observe this with applets other than the quantizer? (for example, is it something you can investigate if you use Scope?)
cafesinleche
chysn wrote:
cafesinleche wrote:
been loving hemisphere immensely and really opens up the o_c in a wonderful way.

I've just started running into a little interesting problem. I'm running a couple CV sources into two doepfer precision adders each of which i plug into one of the dual quantizers. However if I change the cv from one of the precision adders it affects both sides of the dual quantizer. If i were to plug cv into the dual quantizer straight, I dont have this problem and the signals into the quantizers are kept seperate. it only happens when it's run through the precision adders.

any tips on what's going on anyone?


Does this affect only adjacent inputs? Let's say you put the Dual Quantizer in both hemispheres and use the most physically-distant two inputs (input 1 of the left hemisphere and input 2 of the right hemisphere). Is there still this crosstalk?

Do you observe this with applets other than the quantizer? (for example, is it something you can investigate if you use Scope?)


Thanks for getting back to me!

It does not cross hemispheres. It still occurs when using scope.


edit: i actually now think it's a problem with my precision adders bleeding into each other. going to try to trouble shoot this...

thanks!
pld
cafesinleche wrote:
edit: i actually now think it's a problem with my precision adders bleeding into each other. going to try to trouble shoot this...

thanks!

IIRC the precision adders route the output to the CV bus (one of the extra signals in the 12 pin cables) so if you have two they would be shorting. There should be a jumper to disable it for one of them...
chysn
pld wrote:
cafesinleche wrote:
edit: i actually now think it's a problem with my precision adders bleeding into each other. going to try to trouble shoot this...

thanks!

IIRC the precision adders route the output to the CV bus (one of the extra signals in the 12 pin cables) so if you have two they would be shorting. There should be a jumper to disable it for one of them...


OK... thanks, cafesinleche, thanks, Patrick. The closest thing I have to a precision adder is Links, and it's not messing anything up for me. At this point, I don't have any satisfying theories on how it could be a software problem, so please let us know what you find out.
Timmy
cafesinleche wrote:
I've just started running into a little interesting problem. I'm running a couple CV sources into two doepfer precision adders each of which i plug into one of the dual quantizers. However if I change the cv from one of the precision adders it affects both sides of the dual quantizer. If i were to plug cv into the dual quantizer straight, I dont have this problem and the signals into the quantizers are kept seperate. it only happens when it's run through the precision adders.

any tips on what's going on anyone?


As discussed, it sounds like the problem is with or due to the precision adder before your O&C.

However, be aware that there was a minor hardware design flaw with versions of O&C prior to v2.e - basically all versions with 4 trimpots on the back (v2.e removed those and substituted in-firmware calibration instead). The problem resulted in very minor CV bleed between adjacent ADC input channels. When I say very minor, I mean very, very minor! If more than 4 or 5V was input into, say, channel 1 CV in, then channel 2 would get a few millivolts of bleed from the channel 1 voltage. Only adjacent channels were affected, and the bleed was minuscule - if the input voltage was less than 4V or so, the cross-channel bleed was unmeasurable. The problem was completely corrected in v2.e of the hardware design. If your O&C doesn't have any trimpots on it, then it is unaffected.

I doubt this is the cause of the problem described by the OP, but I mention here for the sake of completeness.

Edit: I forgot to mention that there is a post hoc fix available for anyone with a v2.d or earlier O&C. It involves removing the trimpots and then jumpering them, and replacing, iirc, 4 resistors. Easy to do, especially with a hot air gun to desolder the relevant parts, takes about 10 or 20 minutes. That said, I still have several early version O&Cs which I have never bothered to modify - the problem is so small it isn't really worth the effort.
chysn
The Hemisphere Suite 1.7 Release Candidate is available on GitHub (https://github.com/Chysn/O_C-HemisphereSuite) for those of you who compile from source. The official hex file will be released on Friday, Nov. 30, but it probably won't be materially different from this.

* Added Neural Net, a highly-configurable logic processor full app (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Neural-Net)

* Updated CV Recorder with a completely-rewritten applet. The new CV Recorder has 384 steps per track on two tracks. Each track can now be recorded independently. A "Smooth" (linear interpolation) option is now available for playback. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/CV-Recorder)

* Added Backup/Restore app, a utility that allows you to transfer complete sets of app and/or calibration data to and from your module via MIDI. (https://github.com/Chysn/O_C-HemisphereSuite/wiki/Backup-and-Restore)

* Fixed an issue with the MIDI quantizer, which caused inaccurate MIDI notes to be generated during CV-to-MIDI operations (MIDI Out, Captain MIDI) with negative voltage.

* Minor updates to improve consistency in icon usage.

* Added main menu checkbox icons, to show the active app more clearly.

* "A"SR: Added a CV indicator for the index, so you can see what the index is with modulation.
adh82
Amazing work on the cv recorder revision!

Just winding what've would be the best app to get some downsampling or decimating tones?
A dedicated app would Be amazing if you ever considered it. Like the alright devices t-wrex but I know it has to be possible with what's already available.
Suggestions?
chysn
adh82 wrote:
Just winding what've would be the best app to get some downsampling or decimating tones?
A dedicated app would Be amazing if you ever considered it. Like the alright devices t-wrex but I know it has to be possible with what's already available.
Suggestions?


If you're talking about processing audio, try running audio through the Gated VCA. That will downsample your signal to 16K. Also, try RunglBook and LoFi Tape with audio for some gnarly stuff.
Naenyn
So, I’d been avoiding this firmware because I haven’t had my uO_c that long and am still getting used to its capabilities. However, tonight I decided to read up on the wiki and see what Hemisphere was all about. And.. just.. wow.
At the start of this thread, it was mentioned that one mode of the original firmware was removed to make room for Hemisphere. Is that still accurate? It looks like Hemisphere has grown a lot since then. And yet, everything I’m reading seems to indicate that most if not all original functionality is maintained.
Could someone clarify?
I’m definitely going to have to try this at some point... Nice work, chysn!
Osk
I know that the beta kept most of the original O_c apps, does the current release do the same?
Shledge
All o_C apps had to be removed to make room for the full suite.
Naenyn
Shledge wrote:
All o_C apps had to be removed to make room for the full suite.

Ahh that’s a bummer. At least switching firmware doesn’t seem that painful. Thanks for the info!
Osk
Shledge wrote:
All o_C apps had to be removed to make room for the full suite.


On this note, how do the quantizer apps stack up to quantermain?
Shledge
You can edit and convert scala files to scales for hemisphere to use, can handle microtonal scales etc.

chysn
Osk wrote:
Shledge wrote:
All o_C apps had to be removed to make room for the full suite.


On this note, how do the quantizer apps stack up to quantermain?


Quantermain is the heart of a lot of people's setup. Nothing in Hemisphere Suite tries to be that comprehensive. I'd say the functionality of the Hemisphere quantizers lies somewhere between a Doepfer A-156 and a µScale II, with subsets of functionality being distributed among several applets, and other functionality missing all together.

Here are a few of the differences:

* The Hemisphere quantizers' I/O mapping is not changeable. Dual Quantizer always maps CV 1 to Out A, etc.

* For the most part, I ignore masking scales. An exception is Scale Duet, which has two scale masks for the semitone scale, switchable via CV. If you like to do lots of work with masks, you'll probably be annoyed.

* Hemisphere has pretty rigid definitions for the buttons, so there aren't the octave up/down performance controls.

* There's no fine-tuning control.

* Hemisphere Suite abandons the oscillator-specific calibration concept (although it'll still work if you saved such calibrations in memory prior to installation).

* The Hemisphere quantizer apps have only external voltage sources (although, of course, you can patch in output from the opposite hemisphere).

* The Turing Machines are separate applets in Hemisphere, and they have their own quantizers.

* The Hemisphere Suite scale editor is an external app. I think the extra space makes it more usable and opens up the option to save scales with MIDI. There are nearly 4000 scales from the Scala Archive that can be dumped to Hemisphere Suite via MIDI.

It's best to think of Ornament and Crime and Hemisphere Suite as totally different modules that happen to run on the same hardware.
Naenyn
chysn wrote:
It's best to think of Ornament and Crime and Hemisphere Suite as totally different modules that happen to run on the same hardware.

Thanks for explaining! Makes me want two of them. screaming goo yo
aragorn23
I may very well have missed this given how vast this project has become, but is there a ratcheting app anywhere in Hemisphere?
chysn
aragorn23 wrote:
I may very well have missed this given how vast this project has become, but is there a ratcheting app anywhere in Hemisphere?


There's not an applet with traditional ratcheting behavior built-in. But there are countless ways to roll your own ratcheting if you're up to exploring them.

One example: Use TrigSeq (the dual 8-step trigger sequencer) in the left hemisphere, and Burst in the right. Generate two parallel rhythm lines with TrigSeq, and patch one of those lines to Burst. Activate clock forwarding to sync the bursts to the main clock.

If you want more complexity, try ShiftGate instead of TrigSeq. Again, one channel for a main rhythm, and one for ratcheted accents with Burst.
chysn
adh82 wrote:
Just winding what've would be the best app to get some downsampling or decimating tones?
A dedicated app would Be amazing if you ever considered it.


Yeah, this was too good an idea to pass up, and it'll be in Friday's release. Thanks!

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Dr.-Crusher
aragorn23
chysn wrote:
aragorn23 wrote:
I may very well have missed this given how vast this project has become, but is there a ratcheting app anywhere in Hemisphere?


There's not an applet with traditional ratcheting behavior built-in. But there are countless ways to roll your own ratcheting if you're up to exploring them.

One example: Use TrigSeq (the dual 8-step trigger sequencer) in the left hemisphere, and Burst in the right. Generate two parallel rhythm lines with TrigSeq, and patch one of those lines to Burst. Activate clock forwarding to sync the bursts to the main clock.

If you want more complexity, try ShiftGate instead of TrigSeq. Again, one channel for a main rhythm, and one for ratcheted accents with Burst.


Thanks - some really good tips there :-)
rayultine
chysn wrote:
adh82 wrote:
Just winding what've would be the best app to get some downsampling or decimating tones?
A dedicated app would Be amazing if you ever considered it.


Yeah, this was too good an idea to pass up, and it'll be in Friday's release. Thanks!

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Dr.-Crusher


I could kiss you
starthief
I have a simple app idea: a "random-access selector"

In the UI, you can set four values. Each is associated with an input.

A trigger in any of the inputs will set the output to the associated value -- in other words, you are selecting the value by sending a trigger.

The second output is just a boolean "or" of the four inputs.

If you run the app on both sides, they can be linked so it's an 8-way selection.


This is something that few Eurorack modules can do: the Verbos Sequence Selector, Sputnik 5-Step, and SynthWerks PGM 4x4 are all I can think of, or it could be programmed in Teletype. Most sequencers don't have gate inputs to select arbitrary steps even if they have manual buttons to do it.

But I think this sort of value selection by gates is useful enough that I'm surprised Doepfer or someone else hasn't made a dedicated module to do it, and I think it'd make a handy Hemisphere app.
chysn
starthief wrote:
I have a simple app idea: a "random-access selector"

In the UI, you can set four values. Each is associated with an input.

A trigger in any of the inputs will set the output to the associated value -- in other words, you are selecting the value by sending a trigger.

The second output is just a boolean "or" of the four inputs.

If you run the app on both sides, they can be linked so it's an 8-way selection.


This is something that few Eurorack modules can do: the Verbos Sequence Selector, Sputnik 5-Step, and SynthWerks PGM 4x4 are all I can think of, or it could be programmed in Teletype. Most sequencers don't have gate inputs to select arbitrary steps even if they have manual buttons to do it.

But I think this sort of value selection by gates is useful enough that I'm surprised Doepfer or someone else hasn't made a dedicated module to do it, and I think it'd make a handy Hemisphere app.


It is an interesting idea. It actually seems like it would be better as a full app, because then you'd be able to devote inputs to clock and reset.

At the moment, I probably don't have room for new things, as I'm at about 98%. But I'm hoping to make some more space by rewriting the EEPROM code. Right now, I'm using the O_C storage, which is fine, but I also do SysEx storage for everything. My plan is to unify SysEx and EEPROM storage formats, which will reduce the amount of code I have to devote to serializing and un-serializing objects in different formats. This will also allow me to do auto-saves so that you don't need to explicitly save the module's state. TLDR: I might be able to find space for this sort of app or applet in the future.
pld
chysn wrote:
At the moment, I probably don't have room for new things, as I'm at about 98%. But I'm hoping to make some more space by rewriting the EEPROM code. Right now, I'm using the O_C storage, which is fine, but I also do SysEx storage for everything. My plan is to unify SysEx and EEPROM storage formats, which will reduce the amount of code I have to devote to serializing and un-serializing objects in different formats. This will also allow me to do auto-saves so that you don't need to explicitly save the module's state. TLDR: I might be able to find space for this sort of app or applet in the future.

FWIW I did a quick rewrite of parts of the Temps Utile storage to be able to load/save slots (code here).
That implements a read/write to a buffer (rather than direct to EEPROM) but still based off the same ol' SettingsBase as o_C since I didn't want to uproot to much. There's a bonus feature of being able to define dependent settings which can reduce the amount of storage space needed, but the tradeoff is more space in flash meh
starthief
chysn wrote:
It is an interesting idea. It actually seems like it would be better as a full app, because then you'd be able to devote inputs to clock and reset.


My thought was, those incoming selection gates are already going to be sequenced or performed, so the app itself shouldn't need to be clocked... but having that option would make it more like the sequencers I mentioned, so that makes sense smile
coolshirtdotjpg
Every time I check in this gets more amazing. For some reason a lot of multifunction modules tend to do the same things (basic synth building blocks, and certain kinds of functionality always linked together), and this really solves a lot of problems that I had honestly spent a lot of money trying to solve in my system, especially the midi stuff.

With that said, I have one feature request; I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses. The idea being that, let's say you want a bassline to enter for 12 bars and then stop for 4, you could set how many pulses you want it to stay on for, and how many you want it to stay off for, and then hook this gate output to a VCA to make it happen. Doing this without a dedicated app is a pain in the behind, and it's something I want to do in most patches. It's funny because the code for it should be fairly simple but it requires quite a bit of patching to do with any granularity, especially if you want to wait for a lot of pulses (say, in the hundreds). This could be easily accomplished by an app with the use of an encoder.

I post about this problem a lot, so forgive me if I've mentioned it in this thread, but I really do think this would change the way a lot of people patch.
chysn
coolshirtdotjpg wrote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


I like that! It could actually be a sequencer rather than just a gate, where you make a table of values

0V for 64 clocks
5V for 16 clocks
0V for 24 clocks
etc...

Is that kind of the idea?

Like I said earlier, Hemisphere Suite 1.x is pretty much complete now until I revamp some stuff behind the scenes to reclaim some memory, so new ideas need to go in the icebox for now. I'm not sure how much space I'll be able to wring out of the system, but I've got some other ideas brewing, too.

pld wrote:
FWIW I did a quick rewrite of parts of the Temps Utile storage to be able to load/save slots (code here).
That implements a read/write to a buffer (rather than direct to EEPROM) but still based off the same ol' SettingsBase as o_C since I didn't want to uproot to much. There's a bonus feature of being able to define dependent settings which can reduce the amount of storage space needed, but the tradeoff is more space in flash meh


OK, I'll take a look, thanks!

My issue has more to do with data formatting. It's hard to use SettingsBase because I'm usually dealing with objects that need to be saved, rather than just a list of parameters. So I have to serialize the objects, and parameterize the bytes for SettingsBase, and then do it all backwards to get stuff back.

Meanwhile, each app has a system exclusive format in which everything is serialized and sent as a packets of 48 (or so) bytes, and I have to go backwards for that, too. So each app sort of does the same thing twice, and objects are in memory twice (once as values_[] ints, and once as the working objects).

If I store things in flash exactly as they'd come in via SysEx, that inefficiency would go away.
chysn
Hemisphere Suite 1.7 is now available.

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.7

Project roadmap: At this point, 1.x is complete, feature-wise. Version 1.8 (ETA not determined) will focus on clearing up any issues for the Northern Lights Buchla format 2OC, and after that, the roadmap includes some pretty significant low-level changes for what will be called Hemisphere 2. I'm expecting to do away with the "menu" system and build a framework where everything is a Hemisphere applet, but with some applets (e.g., The Darkest Timeline, Captain MIDI, etc.) taking up both hemispheres.
sleepgardens
I clearly lost track of how big this got. The last time i looked into it there were maybe 15 apps? Wow. Installing this on my O_c right away!
magneteyez
sleepgardens wrote:
I clearly lost track of how big this got. The last time i looked into it there were maybe 15 apps? Wow. Installing this on my O_c right away!

+1. Amazing stuff! Thnx
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


I like that! It could actually be a sequencer rather than just a gate, where you make a table of values

0V for 64 clocks
5V for 16 clocks
0V for 24 clocks
etc...

Is that kind of the idea?

Like I said earlier, Hemisphere Suite 1.x is pretty much complete now until I revamp some stuff behind the scenes to reclaim some memory, so new ideas need to go in the icebox for now. I'm not sure how much space I'll be able to wring out of the system, but I've got some other ideas brewing, too.


Yes, that's precisely it! I'm glad you'd consider it for a later version. I was planning on writing it myself, but I'm a pretty poor coder all things considered, and very busy with a newborn/finishing up my Ph.D. If you are able to do this, I think it would be a really great meta app for creating more complex generative pieces with hemispheres.
Ebotronix
hi chisn!
there is something wrong with the Dual Quantizer App.
it was not possible to select a scale with the first Quantizer
if I select Dorian or another scale the quantization is always chromatic.
the other bug is the transpose function from C to B
it doesn't change the between the possible scale steps.
the second Quantizer is working ,
I can select a scale and the quantisation is ok.
but the transpose function doesn't work at both Quantizers.
Firmware 1.65b
chysn
Ebotronix wrote:
hi chisn!
there is something wrong with the Dual Quantizer App.
it was not possible to select a scale with the first Quantizer
if I select Dorian or another scale the quantization is always chromatic.
the other bug is the transpose function from C to B
it doesn't change the between the possible scale steps.
the second Quantizer is working ,
I can select a scale and the quantisation is ok.
but the transpose function doesn't work at both Quantizers.
Firmware 1.65b


I just tested the dual quantizer (1.7, though), and it's working correctly.

Please note that the Dual Quantizer does not have a transposition setting. The setting there is for root note, which is a bit different, and may or may not change the note that's currently being quantized.
Ebotronix
it works if you choose a new scale but if you save the scales and restart
the module the first quantizer does the chromatic scale.

is it possible to add a scale transpose funtion to get the scale modes
of every scale?

something like the greecescales
ionian , dorian, phrygian, lydian, mixolydian,aeolian ,locrian
the next scale startet at the next step of the pre scale
chysn
Ebotronix wrote:
it works if you choose a new scale but if you save the scales and restart
the module the first quantizer does the chromatic scale.


Oh, gotcha. That's a different issue. Thanks!
MuffBob
I just installed the latest hemisphere suite 1.7 and I can't seem to find the Enigma applet? My menu goes from Dual Quantizer to Enigma Jr to Envelope Follower - no main Enigma... can anyone enlighten me as to why that might be please or if I have to do anything to enable it? thanks.
behndy
chysn wrote:
coolshirtdotjpg wrote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


I like that! It could actually be a sequencer rather than just a gate, where you make a table of values

0V for 64 clocks
5V for 16 clocks
0V for 24 clocks
etc...

Is that kind of the idea?



this would be AWESOME. i do something similar a lot as well, and it ties up 2 VCA's and an extra Pam's out.

also, i can't get the tuner to work. i load it in the right side, then plug an osc out into the 2nd digital in, and i see it moving the app on the left side that expects data at Digi In 02, but the tuner doesn't show any note data. whaaaaaat am i doing wrong?

as always, thanks so much for such an AWESOME device.
chysn
MuffBob wrote:
I just installed the latest hemisphere suite 1.7 and I can't seem to find the Enigma applet? My menu goes from Dual Quantizer to Enigma Jr to Envelope Follower - no main Enigma... can anyone enlighten me as to why that might be please or if I have to do anything to enable it? thanks.


Enigma is not an applet, it's a full app. So you get to it from the main menu. Hold down the right button for about two seconds and release it to escape to the main menu.
MuffBob
chysn wrote:
MuffBob wrote:
I just installed the latest hemisphere suite 1.7 and I can't seem to find the Enigma applet? My menu goes from Dual Quantizer to Enigma Jr to Envelope Follower - no main Enigma... can anyone enlighten me as to why that might be please or if I have to do anything to enable it? thanks.


Enigma is not an applet, it's a full app. So you get to it from the main menu. Hold down the right button for about two seconds and release it to escape to the main menu.


Aha! - thank you - much to learn!
missingtwin
Quote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


this would be an excellent utility...found myself wanting this exactly wiggling just now.
soggybag
I just ordered a 1u o_C with Hemispheres. Seemed like it would be suited to small setup.
Shledge
behndy wrote:
chysn wrote:
coolshirtdotjpg wrote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


I like that! It could actually be a sequencer rather than just a gate, where you make a table of values

0V for 64 clocks
5V for 16 clocks
0V for 24 clocks
etc...

Is that kind of the idea?



this would be AWESOME. i do something similar a lot as well, and it ties up 2 VCA's and an extra Pam's out.

also, i can't get the tuner to work. i load it in the right side, then plug an osc out into the 2nd digital in, and i see it moving the app on the left side that expects data at Digi In 02, but the tuner doesn't show any note data. whaaaaaat am i doing wrong?

as always, thanks so much for such an AWESOME device.


You should already be able to do this with the clock divider eg. 1/64 requires 64 triggers to turn on or off.
coolshirtdotjpg
Shledge wrote:
behndy wrote:
chysn wrote:
coolshirtdotjpg wrote:
I would like an app that stays at gate high and gate low for an arbitrary number of clock pulses


I like that! It could actually be a sequencer rather than just a gate, where you make a table of values

0V for 64 clocks
5V for 16 clocks
0V for 24 clocks
etc...

Is that kind of the idea?



this would be AWESOME. i do something similar a lot as well, and it ties up 2 VCA's and an extra Pam's out.

also, i can't get the tuner to work. i load it in the right side, then plug an osc out into the 2nd digital in, and i see it moving the app on the left side that expects data at Digi In 02, but the tuner doesn't show any note data. whaaaaaat am i doing wrong?

as always, thanks so much for such an AWESOME device.


You should already be able to do this with the clock divider eg. 1/64 requires 64 triggers to turn on or off.


You seem to be missing the fact that that on/off cycles would be arbitrary numbers of clock cycles, separate from each other. That's what makes what I'm describing different from a standard clock divider. The ER-101 can do this, but then you use up an entire track and two CV outs if you just want this to make meta chances to your patch. It also eats up available notes.
behndy
sooooo.... Tuner app? i can not get it to work, thought i was following the GitHub instructions correctly.

moop? any advices?
chysn
behndy wrote:
sooooo.... Tuner app? i can not get it to work, thought i was following the GitHub instructions correctly.

moop? any advices?


It works okay, but here are a couple tips:

(1) Make sure that you're using the right hemisphere's second digital input (TR 4 for most O_Cs). This seems counter-intuitive because those inputs are usually just for high or low signals, but the FTM pins are connected to that input, and Tuner swaps the function while Tuner is selected.

(2) If you have a signal connected to the right input and it's still not responding, push the right hemisphere's encoder button. This will re-initialize the frequency capture system, and ensure that it's on.

(3) The signal has to be pretty hot. That is, it's mainly for tuning Eurorack oscillators and not, say, guitars.
hemeroscopium
Dunno if it's already done or not, but gonna ask anyway - is there a trigger/gate delay app per one channel ? (4ins 4outputs)
behndy
chysn wrote:
behndy wrote:
sooooo.... Tuner app? i can not get it to work, thought i was following the GitHub instructions correctly.

moop? any advices?


It works okay, but here are a couple tips:

(1) Make sure that you're using the right hemisphere's second digital input (TR 4 for most O_Cs). This seems counter-intuitive because those inputs are usually just for high or low signals, but the FTM pins are connected to that input, and Tuner swaps the function while Tuner is selected.

(2) If you have a signal connected to the right input and it's still not responding, push the right hemisphere's encoder button. This will re-initialize the frequency capture system, and ensure that it's on.

(3) The signal has to be pretty hot. That is, it's mainly for tuning Eurorack oscillators and not, say, guitars.


coooool. thank you! i"ll try that when i get home tonight.
chysn
erenod wrote:
Dunno if it's already done or not, but gonna ask anyway - is there a trigger/gate delay app per one channel ? (4ins 4outputs)


Yes! Gate Delay is the traditional dual delay (modeled after Ladik S-189): https://github.com/Chysn/O_C-HemisphereSuite/wiki/Gate-Delay

And see also "A"SR, which can do some of the same stuff in a different way: https://github.com/Chysn/O_C-HemisphereSuite/wiki/%22A%22SR

And, of course, CV Recorder, which can do some of the same stuff in an even differenter way: https://github.com/Chysn/O_C-HemisphereSuite/wiki/CV-Recorder
sushiluv
Can´t say enough how much i enjoy Hemisphere, thanks!

however, i have a feature suggestion for enigma/jr.
Would it be possible to add an internal sample&hold in those applets?
So when using output b for triggers, the pitch only updates when a trigger occurs. smile
chysn
sushiluv wrote:
So when using output b for triggers, the pitch only updates when a trigger occurs. smile


Which pitch are you referring to?
aaooaa
rayultine wrote:
I'm a nerd that likes to print manuals and user guides. I made a pdf of chysn's applet descriptions:

www (dot) sendspace.com/file/ibjd9o

Don't get a virus from sendspace ads please.


How did you end up creating a pdf from the wiki? I'd like to make and print an updated version but can't figure it out apart from an ugly epub copy or pasting each individual page into a text doc. Seems to be a lot of command line tools but I'd like to avoid that for now if possible.
behndy
annnnnd that's it. thanks so much!
rayultine
aaooaa wrote:
rayultine wrote:
I'm a nerd that likes to print manuals and user guides. I made a pdf of chysn's applet descriptions:

www (dot) sendspace.com/file/ibjd9o

Don't get a virus from sendspace ads please.


How did you end up creating a pdf from the wiki? I'd like to make and print an updated version but can't figure it out apart from an ugly epub copy or pasting each individual page into a text doc. Seems to be a lot of command line tools but I'd like to avoid that for now if possible.


pasting each individual page into a text doc. I didn't feel like doing my actual job for about 15 minutes, so... smile
selfdestroyer
rayultine wrote:
aaooaa wrote:
rayultine wrote:
I'm a nerd that likes to print manuals and user guides. I made a pdf of chysn's applet descriptions:

www (dot) sendspace.com/file/ibjd9o

Don't get a virus from sendspace ads please.


How did you end up creating a pdf from the wiki? I'd like to make and print an updated version but can't figure it out apart from an ugly epub copy or pasting each individual page into a text doc. Seems to be a lot of command line tools but I'd like to avoid that for now if possible.


pasting each individual page into a text doc. I didn't feel like doing my actual job for about 15 minutes, so... smile


I did the exact same thing.... at work. lol
aaooaa
I've made an updated, printable version of the wiki. There's two versions, one with every applet on it's own page and if you want to save on paper the other one has them one after the other.

www (dot) sendspace.com/file/rxqvl0
vantablack
aaooaa wrote:
I've made an updated, printable version of the wiki. There's two versions, one with every applet on it's own page and if you want to save on paper the other one has them one after the other.

www (dot) sendspace.com/file/rxqvl0


wow thank you!
Destinova
aaooaa wrote:
I've made an updated, printable version of the wiki. There's two versions, one with every applet on it's own page and if you want to save on paper the other one has them one after the other.

www (dot) sendspace.com/file/rxqvl0


That's great! Thanks so much for sharing.
miles_macquarrie
I have a question that has probably already been answered, but I don't want to have to go through all 30 pages of this thread.

Can Hemisphere Suite and all the apps from the most current O_c firmware be used simultaneously, meaning do I need to flash different firmware each time I want to switch between?

Thanks,
Miles
Naenyn
miles_macquarrie wrote:
do I need to flash different firmware each time I want to switch between?
Alas, you do indeed need to flash to switch between Hemisphere and the original firmware. Thankfully, the process is painless.
chysn
Naenyn wrote:
miles_macquarrie wrote:
do I need to flash different firmware each time I want to switch between?
Alas, you do indeed need to flash to switch between Hemisphere and the original firmware. Thankfully, the process is painless.


One thing I'll add is that you'll lose your app settings (but not calibration) when you change the firmware. Prior to replacing Hemisphere Suite with classic O_C, I'd recommend doing an app data backup with the Backup/Restore app, so you can get your stuff back if you re-install Hemisphere Suite.
aaooaa
Destinova wrote:
That's great! Thanks so much for sharing.


No worries! It took me ages to figure out so there's bound to be a few mistakes, let me know if anything is missing or can be improved... nanners
miles_macquarrie
chysn wrote:
Naenyn wrote:
miles_macquarrie wrote:
do I need to flash different firmware each time I want to switch between?
Alas, you do indeed need to flash to switch between Hemisphere and the original firmware. Thankfully, the process is painless.


One thing I'll add is that you'll lose your app settings (but not calibration) when you change the firmware. Prior to replacing Hemisphere Suite with classic O_C, I'd recommend doing an app data backup with the Backup/Restore app, so you can get your stuff back if you re-install Hemisphere Suite.


Thank you both for the info.
4eyez
I was going through the manual on github and noticed this:

Quote:
Vector Oscillators are used in four Hemisphere applets:

Boots 'n Cats, as an audio oscillator and drum envelope generators


Is this correct? If so, how do we access the vector waveforms in BootsNCat?
Ebotronix
hi chysn!
can you add a chord app that is easy to program with all available scales?
selectable in a list with up to 32 four voice chords (for 4 VCO's)
something like this...
examples
C major
steps 0, 4, 7, 12
C minor7
steps 0,3,7, 10
chysn
4eyez wrote:
I was going through the manual on github and noticed this:

Quote:
Vector Oscillators are used in four Hemisphere applets:

Boots 'n Cats, as an audio oscillator and drum envelope generators


Is this correct? If so, how do we access the vector waveforms in BootsNCat?


Well, yeah, Boots 'n Cats produces its sound with a Vector Oscillator, and controls the amplitude with another. But there's no interface available for changing the audio waveform from a triangle. It might be an interesting thing to try some day, though.

Ebotronix wrote:
can you add a chord app that is easy to program with all available scales?


Hemisphere Suite is now complete, app-wise. There's just no room for anything else.

Over the next few weeks, I'm going to be evaluating Spoken Into The Void, an open-source hardware replacement board by Tom Moxon that basically upgrades a µOC to Teensy 3.5/3.6, gives it way more memory, SD storage, audio outputs, and various other things. It's likely that Hemisphere 2 will be built on the µOC+SitV platform, and I'll be able to do a lot more stuff.
Naenyn
chysn wrote:
Over the next few weeks, I'm going to be evaluating Spoken Into The Void, an open-source hardware replacement board by Tom Moxon that basically upgrades a µOC to Teensy 3.5/3.6, gives it way more memory, SD storage, audio outputs, and various other things. It's likely that Hemisphere 2 will be built on the µOC+SitV platform, and I'll be able to do a lot more stuff.
Woooo that is super-exciting! I hope things work out and you're able to give it a shot. Good luck!
coolshirtdotjpg
chysn wrote:
4eyez wrote:
I was going through the manual on github and noticed this:

Quote:
Vector Oscillators are used in four Hemisphere applets:

Boots 'n Cats, as an audio oscillator and drum envelope generators


Is this correct? If so, how do we access the vector waveforms in BootsNCat?


Well, yeah, Boots 'n Cats produces its sound with a Vector Oscillator, and controls the amplitude with another. But there's no interface available for changing the audio waveform from a triangle. It might be an interesting thing to try some day, though.

Ebotronix wrote:
can you add a chord app that is easy to program with all available scales?


Hemisphere Suite is now complete, app-wise. There's just no room for anything else.

Over the next few weeks, I'm going to be evaluating Spoken Into The Void, an open-source hardware replacement board by Tom Moxon that basically upgrades a µOC to Teensy 3.5/3.6, gives it way more memory, SD storage, audio outputs, and various other things. It's likely that Hemisphere 2 will be built on the µOC+SitV platform, and I'll be able to do a lot more stuff.


Does this mean that those who have older O+C can upgrade by swapping in the newer teensy? Or is it only compatable with the micro version
Naenyn
coolshirtdotjpg wrote:
Does this mean that those who have older O+C can upgrade by swapping in the newer teensy? Or is it only compatable with the micro version

I also wonder if the firmware would need to be recompiled for the different teensy...
chysn
coolshirtdotjpg wrote:
Does this mean that those who have older O+C can upgrade by swapping in the newer teensy? Or is it only compatable with the micro version


My understanding is that it's only compatible with the micro version. Or, at least it was designed with µOC in mind. But for the µOC, it's a simple replacement: take out the Teensy 3.2, and put the Spoken Into The Void board (which includes Teensy 3.5 or 3.6, in addition to some other stuff) in its place.

Naenyn wrote:
I also wonder if the firmware would need to be recompiled for the different teensy...


Yes, absolutely. Some of the low-level guts need to be re-arranged a little bit. Tom has already got classic O_C and Hemisphere Suite running with the Teensy 3.5.
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
Does this mean that those who have older O+C can upgrade by swapping in the newer teensy? Or is it only compatable with the micro version


My understanding is that it's only compatible with the micro version. Or, at least it was designed with µOC in mind. But for the µOC, it's a simple replacement: take out the Teensy 3.2, and put the Spoken Into The Void board (which includes Teensy 3.5 or 3.6, in addition to some other stuff) in its place.

Naenyn wrote:
I also wonder if the firmware would need to be recompiled for the different teensy...


Yes, absolutely. Some of the low-level guts need to be re-arranged a little bit. Tom has already got classic O_C and Hemisphere Suite running with the Teensy 3.5.


That's interesting. This means that a fairly small segment of O+C owners will be able to use it (uO+C owners who are capable of modifying their devices, even if it's a simple switch, I'd venture that most of the folks who bought pre-made O+C won't do that) correct?

To my mind, that sort-of defeats the purpose of it being alternate firmware a bit. That's totally fine, but I'm wondering, since you're interesting in moving towards new hardware, why not partner with someone to come up a hemispheres module with hardware that supports all the things you are hoping to accomplish in hemispheres 2.0? You could maintain everything great about the O+C version, while fixing all the stuff that isn't (for example, some of the audio rate stuff you are doing is probably hampered a bit by the ADCs O+C uses).

I also say this because, I hate the interface of micro modules, and hemispheres is already a bit small on a module with a full-size screen. One other possibility would be to make some of the new modes backwards compatible so that we could compile them in place of apps we already have covered in our system.

I am really grateful for the work you've done, these are just suggestions that I hope will be helpful.
chysn
coolshirtdotjpg wrote:
That's interesting. This means that a fairly small segment of O+C owners will be able to use it (uO+C owners who are capable of modifying their devices, even if it's a simple switch, I'd venture that most of the folks who bought pre-made O+C won't do that) correct?


Sort of. Note that it's the world's easiest "modification." The brains of an Ornament and Crime module, the Teensy 3.2, is the small circuit board on the bottom layer. It's in a 26-pin socket. Spoken Into The Void requires you to remove the Teensy 3.2, and plug the SitV board where the Teensy 3.2 used to be. It's literally as hard as plugging a power cable into a module, but without the risk of blowing up your whole system if you do it wrong.

Quote:
To my mind, that sort-of defeats the purpose of it being alternate firmware a bit. That's totally fine, but I'm wondering, since you're interesting in moving towards new hardware, why not partner with someone to come up a hemispheres module with hardware that supports all the things you are hoping to accomplish in hemispheres 2.0?


Because laziness. I already know the O_C framework, and I already have 50+ apps written for it. I don't want to go anywhere near producing a new module.

Note that nothing is really decided yet. Tom is sending me a SitV board, and I'll be able to evaluate it and see what it offers. It has to be something that will be immediately useful if people buy or build one, and also offer good value in the form of future potential. In these respects, I think it looks promising.

Quote:
I also say this because, I hate the interface of micro modules, and hemispheres is already a bit small on a module with a full-size screen.


Yeah, I think 14HP was about the right size for the module. It spreads things out nicely. At 8HP, though, it's not practically harder to use, but the jacks tend to become more congested.
mcpepe
So wouldnt the normal O+C users be able to simply change the Teensy to upgrade the module?
chysn
mcpepe wrote:
So wouldnt the normal O+C users be able to simply change the Teensy to upgrade the module?


I'm interpreting "change the Teensy" as "replace the Teensy 3.2 with a Teensy 3.5." If this is your question, the answer is no. The later Teensy boards have more pins and wouldn't fit the socket.
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
That's interesting. This means that a fairly small segment of O+C owners will be able to use it (uO+C owners who are capable of modifying their devices, even if it's a simple switch, I'd venture that most of the folks who bought pre-made O+C won't do that) correct?


Sort of. Note that it's the world's easiest "modification." The brains of an Ornament and Crime module, the Teensy 3.2, is the small circuit board on the bottom layer. It's in a 26-pin socket. Spoken Into The Void requires you to remove the Teensy 3.2, and plug the SitV board where the Teensy 3.2 used to be. It's literally as hard as plugging a power cable into a module, but without the risk of blowing up your whole system if you do it wrong.

Quote:
To my mind, that sort-of defeats the purpose of it being alternate firmware a bit. That's totally fine, but I'm wondering, since you're interesting in moving towards new hardware, why not partner with someone to come up a hemispheres module with hardware that supports all the things you are hoping to accomplish in hemispheres 2.0?


Because laziness. I already know the O_C framework, and I already have 50+ apps written for it. I don't want to go anywhere near producing a new module.

Note that nothing is really decided yet. Tom is sending me a SitV board, and I'll be able to evaluate it and see what it offers. It has to be something that will be immediately useful if people buy or build one, and also offer good value in the form of future potential. In these respects, I think it looks promising.

Quote:
I also say this because, I hate the interface of micro modules, and hemispheres is already a bit small on a module with a full-size screen.


Yeah, I think 14HP was about the right size for the module. It spreads things out nicely. At 8HP, though, it's not practically harder to use, but the jacks tend to become more congested.


I hear you, that makes sense. I think if you ever do, there will very likely be people who would be willing to do that work for you because I think it's pretty brilliant. In fact, since O+C is open source, I think someone my well do a revised O+C that allows for this adaptation. I know that it's super easy to just swap in what is essentially a socketed IC, I'm just going to guess that most people won't do that, which is a shame since I'm certain Hemisphere 2.0 is going to be incredible. Either way, thanks a thousand times over, I'm loving hemispheres.
Timmy
chysn wrote:
mcpepe wrote:
So wouldnt the normal O+C users be able to simply change the Teensy to upgrade the module?


I'm interpreting "change the Teensy" as "replace the Teensy 3.2 with a Teensy 3.5." If this is your question, the answer is no. The later Teensy boards have more pins and wouldn't fit the socket.


No, that's not quite correct. The Teensy 3.5 and 3.6 do indeed have more pins, but they are also backwardly pin-compatible with the Teensy 3.1 and 3.2. Thus you can just plug the matching Teensy 3.5 or 3.6 pins into a Teensy 3.2/2 socket. Firmware may need adjustment (in the case of O&C, it definitely needs adjustment), but all the matching pins are backwardly compatible.

However on the 14HP O&C, the eurorack power cable interferes with the Teensy 3.5/6 boards, which are longer, but turn out to be exactly 14HP wide and they just fit behind the 14HP O&C panel widthwise, but they need an ungainly set of socket "stilts" to lift the Teensy board away from the main PCB far enough to clear the power. It's an ugly, fragile and unsatisfactory arrangement.

What is needed is an adaptor PCB which fits into the existing horizontal Teensy socket on the O&C underneath, and on top provides a socket for the Teensy 3.5/6 mounted vertically, with the pins connected suitably through the adaptor PCB to the teensy 3.1/2 socket underneath. Such an adaptor should be able to fit both the 14HP O&C and the 8HP uO&C. And, of course, the firmware needs to be tweaked to make it all work. In theory it should be just a matter of remapping the pin declarations and tweaking the clock speed and clock dividers a bit. I started to fiddle with that a year ago, but had to move house and somehow misplaced the box with all my Teensy 3.6 boards and the jury-rigged socket adaptor (just a bunch of flying leads soldered up as proof-of-concept), so I never finished it. I got as far as the Teensy 3.6 booting O&C firmware, but the display and DAC SPI timings weren't right so no display or output.

The SitV adaptor appears to want to expand the O&C hardware capabilities, not just upgrade the processor and storage, and its author said he plans to re-write the firmware framework from scratch, as I understand it. That's fine, but that makes it a different project, I think.
coolshirtdotjpg
Timmy wrote:
chysn wrote:
mcpepe wrote:
So wouldnt the normal O+C users be able to simply change the Teensy to upgrade the module?


I'm interpreting "change the Teensy" as "replace the Teensy 3.2 with a Teensy 3.5." If this is your question, the answer is no. The later Teensy boards have more pins and wouldn't fit the socket.


No, that's not quite correct. The Teensy 3.5 and 3.6 do indeed have more pins, but they are also backwardly pin-compatible with the Teensy 3.1 and 3.2. Thus you can just plug the matching Teensy 3.5 or 3.6 pins into a Teensy 3.2/2 socket. Firmware may need adjustment (in the case of O&C, it definitely needs adjustment), but all the matching pins are backwardly compatible.

However on the 14HP O&C, the eurorack power cable interferes with the Teensy 3.5/6 boards, which are longer, but turn out to be exactly 14HP wide and they just fit behind the 14HP O&C panel widthwise, but they need an ungainly set of socket "stilts" to lift the Teensy board away from the main PCB far enough to clear the power. It's an ugly, fragile and unsatisfactory arrangement.

What is needed is an adaptor PCB which fits into the existing horizontal Teensy socket on the O&C underneath, and on top provides a socket for the Teensy 3.5/6 mounted vertically, with the pins connected suitably through the adaptor PCB to the teensy 3.1/2 socket underneath. Such an adaptor should be able to fit both the 14HP O&C and the 8HP uO&C. And, of course, the firmware needs to be tweaked to make it all work. In theory it should be just a matter of remapping the pin declarations and tweaking the clock speed and clock dividers a bit. I started to fiddle with that a year ago, but had to move house and somehow misplaced the box with all my Teensy 3.6 boards and the jury-rigged socket adaptor (just a bunch of flying leads soldered up as proof-of-concept), so I never finished it. I got as far as the Teensy 3.6 booting O&C firmware, but the display and DAC SPI timings weren't right so no display or output.

The SitV adaptor appears to want to expand the O&C hardware capabilities, not just upgrade the processor and storage, and its author said he plans to re-write the firmware framework from scratch, as I understand it. That's fine, but that makes it a different project, I think.



I'm not an expert, but that adapter seems like it would be a 10 minute CAD project for someone who has made a PCB before. Wondering if anyone reading this might be interested? I could talk to some of the engineering students at my university who are interested in eurorack, to see if there's any interest.
pld
coolshirtdotjpg wrote:
I'm not an expert, but that adapter seems like it would be a 10 minute CAD project for someone who has made a PCB before. Wondering if anyone reading this might be interested? I could talk to some of the engineering students at my university who are interested in eurorack, to see if there's any interest.

Yes, it should be super simple (if somewhat ugly) and mostly a mechanical puzzle. Not sure about the power consumption of the 3.6 or if it'd maybe be useful to include an addtional vreg.
Heck, it only took a few days to layout an adapter with an stm32f405.

All of these approaches are only partial solutions. You still have the DAC/screen sharing a port which is an inherent limitation of the main PCB, so if all you wanted was more space for apps (or rather, more flexibility which are loaded/used) there'd be less hardware-intensive approaches.
aragorn23
Is there anywhere one can read more on these SitV boards?
pld
aragorn23 wrote:
Is there anywhere one can read more on these SitV boards?

This post (and some follow-ups) in the o_C thread.
aragorn23
pld wrote:
aragorn23 wrote:
Is there anywhere one can read more on these SitV boards?

This post (and some follow-ups) in the o_C thread.


Cool, thanks!
selfdestroyer
aragorn23 wrote:
Is there anywhere one can read more on these SitV boards?


You may also want to watch his Github repository on the project:

https://github.com/PatternAgents/TheVoid
qbassa@gmail.com
Is there an algorithm that can output two additional pitches to the root note pitch on input to complete selected type of the chord? Changing chord type could be second CV input. Not sure about the need to use digital inputs, like quentizers do to trigger the moment output pitches change. Best!
chysn
qbassa@gmail.com wrote:
Is there an algorithm that can output two additional pitches to the root note pitch on input to complete selected type of the chord? Changing chord type could be second CV input. Not sure about the need to use digital inputs, like quentizers do to trigger the moment output pitches change. Best!


No, nothing remotely close to that in Hemisphere Suite. In fact, I ditched the user chords so I'd have space for user Turing Machines.
4eyez
Crazy idea: having Hemisphere apps (Enigma, Darkest, etc) sending data over the Teensy's i2c ports to things like monome Ansible, ER-301, TXo, etc or receiving data from the new 16n fader controller. Pretty niche and I'm sure more of a pain that I would know, but something that came to mind to throw out there.
chysn
4eyez wrote:
Crazy idea: having Hemisphere apps (Enigma, Darkest, etc) sending data over the Teensy's i2c ports to things like monome Ansible, ER-301, TXo, etc or receiving data from the new 16n fader controller. Pretty niche and I'm sure more of a pain that I would know, but something that came to mind to throw out there.


I don't think the SCL and SDA pins are actually connected to anything on the Ornament and Crime. I could be wrong. In any case, I don't have any of that hardware, so I wouldn't get too far.

As far as hardware enhancements go, I'm getting ready to explore the Spoken Into The Void board, which is on my desk, but I've been taking a bit of a holiday-season break. The board does a lot of stuff, but my focus will probably be on MIDI functions. There will be a MIDI DIN breakout module available, and the board has potential to be a MIDI host, which is potentially interesting.
pld
chysn wrote:
4eyez wrote:
Crazy idea: having Hemisphere apps (Enigma, Darkest, etc) sending data over the Teensy's i2c ports to things like monome Ansible, ER-301, TXo, etc or receiving data from the new 16n fader controller. Pretty niche and I'm sure more of a pain that I would know, but something that came to mind to throw out there.


I don't think the SCL and SDA pins are actually connected to anything on the Ornament and Crime. I could be wrong. In any case, I don't have any of that hardware, so I wouldn't get too far.

There should be unused SCL/SDA on the bottom, which is fairly inconvenient -- back when the TELEX{io} were first released bpcmusic pinged us with the same idea, and that was one of the main sticking points for not pursuing in further. That and, given the nature of i2c communications, it'd also need a bunch of framework work to integrate; one would have to decide if o_C is the master (reading values from TXi, pushing values to TXo etc.) or slave (being queried/driven from TT), which can change the timing requirements significantly. As master there's a dependence on the number of additional devices being addressed, and as slave you need to be timely in responding. I'm not sure of the current state of multi-master with those devices but it can be a huge PITA.
While it does generally work (heck, I made a clone of TELEXi) I'm a bit of a sceptic of i2c in this context, but apparently I'm in the minority and practice beats theory smile
4eyez
Yeah, not surprised to hear i2c is a PITA in this context.

Excited to hear about potentially more MIDI options for Hem with SITV board. I spent the holidays rejigging my studio and one of the things was getting a USB cable permanently running from my uOC to my computer so it can talk to everything else.
Mdashdotdashn
First of all, i'd like to chime my love for Hemisphere. First of all because it's packed of utilities. Second, as a developer, it makes developing apps a breeze which is something that was a huge pain in the original O_C. I've got tons of ideas to implement which I'll do, time permitting.

4eyez wrote:
Crazy idea: having Hemisphere apps (Enigma, Darkest, etc) sending data over the Teensy's i2c ports to things like monome Ansible, ER-301, TXo, etc or receiving data from the new 16n fader controller. Pretty niche and I'm sure more of a pain that I would know, but something that came to mind to throw out there.


If there's indeed a possibility to upgrade O_C to a teensy 3.6, there's also the possibility to implement usb midi host and connect various devices to it. See the little-scale's video here:

Mini launchpad and teensy 3.6

That's be double awesome.
pld
Mdashdotdashn wrote:
Second, as a developer, it makes developing apps a breeze which is something that was a huge pain in the original O_C. I've got tons of ideas to implement which I'll do, time permitting.

No wonder, it wasn't intended to be an SDK since, well, time wasn't permitting. Making apps was possible and fairly simple, but I don't think there was ever a claim it was supposed to be easy smile
Mdashdotdashn
pld wrote:

No wonder, it wasn't intended to be an SDK since, well, time wasn't permitting. Making apps was possible and fairly simple, but I don't think there was ever a claim it was supposed to be easy smile


Don't get me wrong, the O_C is one of the most exciting thing a developer can get its hands one, you guys have my eternal gratitude for it. It's just that now making an app is a breeze and I love focusing on doing something else than frameworky stuff grin
uebl
I guess this could be patched using the clock skipper app plus some mixer/gate-or module, but what do you think of an app that emulates (a single "channel" of) Noise Engineering's Confundo Funkitus, i.e. a probabilistic rhythm cross fader?
chysn
uebl wrote:
I guess this could be patched using the clock skipper app plus some mixer/gate-or module, but what do you think of an app that emulates (a single "channel" of) Noise Engineering's Confundo Funkitus, i.e. a probabilistic rhythm cross fader?


I don't understand Confundo Funkitus well enough to have a good response. For one thing, I don't really grasp, at the moment, the concept of "crossfading" as applied to a set of trigger outputs. A few minutes on YouTube didn't really enlighten me. But it sounds intriguing.
Mdashdotdashn
chysn wrote:
uebl wrote:
I guess this could be patched using the clock skipper app plus some mixer/gate-or module, but what do you think of an app that emulates (a single "channel" of) Noise Engineering's Confundo Funkitus, i.e. a probabilistic rhythm cross fader?


I don't understand Confundo Funkitus well enough to have a good response. For one thing, I don't really grasp, at the moment, the concept of "crossfading" as applied to a set of trigger outputs. A few minutes on YouTube didn't really enlighten me. But it sounds intriguing.


Maybe crossfading is the probablility factor between the two inputs ? Like a bernouilli gate fed to a switcher ?
uebl
Yes, the crossfader sets the probabilities of the 2 trigger inputs to "go through" to the output. Confundo Funkitus has 3 modes:

Mode 1: if crossfader is centered, both input 1's trigger and input 2's trigger go through 100% of the time. If crossfader moves left of center, input 2's probability decreases. If crossfader moves right of center, input 1's probability decreases.

Mode 2: if crossfader is centered, both input 1's trigger and input 2's trigger go through 50% of the time. If crossfader moves left of center, input 1's probability increases and input 2's probability decreases. If crossfader moves right of center, input 2's probability increases and input 1's probability decreases.

Mode 3: if crossfader is centered, neither input 1's trigger nor input 2's trigger goes through. If crossfader is fully right, 100% of input 1's triggers go through. If crossfader is fully left, 100% of input 2's triggers go through.
Mdashdotdashn
uebl wrote:
Yes, the crossfader sets the probabilities of the 2 trigger inputs to "go through" to the output. Confundo Funkitus has 3 modes:

Mode 1: if crossfader is centered, both input 1's trigger and input 2's trigger go through 100% of the time. If crossfader moves left of center, input 2's probability decreases. If crossfader moves right of center, input 1's probability decreases.

Mode 2: if crossfader is centered, both input 1's trigger and input 2's trigger go through 50% of the time. If crossfader moves left of center, input 1's probability increases and input 2's probability decreases. If crossfader moves right of center, input 2's probability increases and input 1's probability decreases.

Mode 3: if crossfader is centered, neither input 1's trigger nor input 2's trigger goes through. If crossfader is fully right, 100% of input 1's triggers go through. If crossfader is fully left, 100% of input 2's triggers go through.


But it's not clocked right ? So basically it is two channels with a probability of pass-through except the probabilities are linked through the crossfader.
uebl
Mdashdotdashn wrote:
uebl wrote:
Yes, the crossfader sets the probabilities of the 2 trigger inputs to "go through" to the output. Confundo Funkitus has 3 modes:

Mode 1: if crossfader is centered, both input 1's trigger and input 2's trigger go through 100% of the time. If crossfader moves left of center, input 2's probability decreases. If crossfader moves right of center, input 1's probability decreases.

Mode 2: if crossfader is centered, both input 1's trigger and input 2's trigger go through 50% of the time. If crossfader moves left of center, input 1's probability increases and input 2's probability decreases. If crossfader moves right of center, input 2's probability increases and input 1's probability decreases.

Mode 3: if crossfader is centered, neither input 1's trigger nor input 2's trigger goes through. If crossfader is fully right, 100% of input 1's triggers go through. If crossfader is fully left, 100% of input 2's triggers go through.


But it's not clocked right ? So basically it is two channels with a probability of pass-through except the probabilities are linked through the crossfader.


Yes, exactly. The not-being-clocked thing is what makes the Confundo Funkitus a bit hard to understand at first.
Mdashdotdashn
Here's a quick attempt in the direction, just for the kicks and because I had a couple of hours free.

Not sure I got it right though.
https://www.instagram.com/p/BslV8djDqaj/

(if someone knows why instagram likes randomly tilting videos, I'm all ears smile
uebl
Cool! What exactly did you put into the inputs?
Mdashdotdashn
uebl wrote:
Cool! What exactly did you put into the inputs?


Clock triggers at different rates. I'm not sure what is the best use of the underlying principle. I'm also not sure what is supposed to happen when gates overlap (I'm assuming they are or'ed) but at least it show it is somehow easy to achieve. I would need to iron things out to get it right.
uebl
I'd play around with one input being "straight" (e.g. 4/4) and the other being "broken"/"shuffled" and see what a "crossfader position" between 0.20 and 0.8 (assuming 0 fully left, 1 fully right) would give me. And yes, I assume overlapping gates are or'ed.
Mashmore
just snagged an OP-Z and scratching my head on a couple captain midi related issues.

First i must say i am still wrapping my head around the op-z when it come to midi. So this could most def be user error.

I am getting note and gates and trigs from the opz to the captain no problem so far. I am even clocking pams new workout via 24ppqn with pretty good accuracy. but as soon as i sequence something from the opz the clock on pams becomes unstable. is the captain picking up spurious noise and causing this?

Also, would it be possible to just send a clock from pams and have the captain send midi time clock information to the opz? I could also send a high value from pams to run/stop the transport.
Any help would be fantastic. CHEERS!
chysn
Mashmore wrote:
but as soon as i sequence something from the opz the clock on pams becomes unstable. is the captain picking up spurious noise and causing this?


The short answer is, no, it couldn't really be "noise." The Clock type won't do anything unless MIDI clock is received, and the setting basically just sends a CV trigger based on some multiple of that clock, up to 24ppqn.

Quote:
Also, would it be possible to just send a clock from pams and have the captain send midi time clock information to the opz? I could also send a high value from pams to run/stop the transport.


Captain MIDI only receives MIDI clock, it can't send it. This is a limitation of the Teensy 3.2 USB MIDI library.
Ebotronix
chysn wrote:
Ebotronix wrote:
it works if you choose a new scale but if you save the scales and restart
the module the first quantizer does the chromatic scale.


Oh, gotcha. That's a different issue. Thanks!


don't forget to fix the save function!
chysn
Ebotronix wrote:
chysn wrote:
Ebotronix wrote:
it works if you choose a new scale but if you save the scales and restart
the module the first quantizer does the chromatic scale.


Oh, gotcha. That's a different issue. Thanks!


don't forget to fix the save function!


This fix is already released as 1.7A: https://github.com/Chysn/O_C-HemisphereSuite/releases
seanrieger
So when using the USB port on back for class compliant midi device, the unit lists itself as "hemisphere". Is there an easy way to change that per device? I have 3 of these units now, and would love to be able to differentiate (Hemisphere 1, Hemisphere 2, Hemisphere 3 etc...) on my DAW. I'm assuming that's why only one will get signal at a time for me (when using Captain Midi), or is there more at play than just the name? What I'm trying to accomplish is have O_C 1 get gate and cv on midi channels 1 and 2, O_C 2 gets gate and cv from midi channel 3 and 4 and O_C 3 get gate and cv from midi channel 5 and 6. What happens is that whatever O_C is plugged in first gets midi and triggers gates and cv just fine, while the other two units don't get midi at all. In my DAW I see 3 instances of hemisphere showing up in the midi devices section.
chysn
seanrieger wrote:
I'm assuming that's why only one will get signal at a time for me (when using Captain Midi), or is there more at play than just the name?


I'm surprised that the OS doesn't sort them out, but I've never tried two at once. I've got a Northern Light 2OC in the house at the moment, so I'll play around with it today and see what's up.

Technical FYI:

If you're compiling in Arduino IDE, the name and ID are not set in the main codebase, but in the Teensy library files. Here's the path relative to the .app path for MacOS:

Contents/Java/hardware/teensy/avr/cores/teensy3/usb_desc.h

You'd just scroll down to the USB_MIDI section and it'll become apparent what to edit.
chysn
Update on the above post

In Ableton Live on MacOS, the second device comes up with the name "Hemisphere #2" and Ableton can access them independently. If your system isn't doing this, it's likely some issue with the OS or your DAW, and not a problem with the Teensy's USB MIDI implementation.
seanrieger
chysn wrote:
Update on the above post

In Ableton Live on MacOS, the second device comes up with the name "Hemisphere #2" and Ableton can access them independently. If your system isn't doing this, it's likely some issue with the OS or your DAW, and not a problem with the Teensy's USB MIDI implementation.


I just wanted to say thank you. Not just for the quick answer but for making an amazing firmware. I'll look deeper at my OS and go from there. You sir, rock.
chysn
seanrieger wrote:
chysn wrote:
Update on the above post

In Ableton Live on MacOS, the second device comes up with the name "Hemisphere #2" and Ableton can access them independently. If your system isn't doing this, it's likely some issue with the OS or your DAW, and not a problem with the Teensy's USB MIDI implementation.


I just wanted to say thank you. Not just for the quick answer but for making an amazing firmware. I'll look deeper at my OS and go from there. You sir, rock.


Thanks!
Ebotronix
Hi chysn!
I use the Dual Quantizer.
when I chose the last scale (Bohlen-Pierce l)on both sections
of the two quantizers
the module stoped working when I chose the next place from the second quantizer.
the next place (B-PI) from the first quantizer is user1,
the next place (B-PI) from the second quantizer is a free space.
if I select this free space the modul stoped working.
chysn
Ebotronix wrote:
Hi chysn!
I use the Dual Quantizer.
when I chose the last scale (Bohlen-Pierce l)on both sections
of the two quantizers
the module stoped working when I chose the next place from the second quantizer.
the next place (B-PI) from the first quantizer is user1,
the next place (B-PI) from the second quantizer is a free space.
if I select this free space the modul stoped working.


Thank you for the report. I'm able to reproduce this, and I'll correct it for a 1.7B release.
chysn
Update on the above post

The fix for this issue is available as 1.7B:

https://github.com/Chysn/O_C-HemisphereSuite/releases/tag/HSv1.7B[/b]

Thanks again!
rayultine
Good sleuthing Ebotronix! I hit this bug many versions ago but didn't really understand what had happened and couldn't reproduce!
Ebotronix
rayultine wrote:
Good sleuthing Ebotronix! I hit this bug many versions ago but didn't really understand what had happened and couldn't reproduce!


I was looking for the scales behind the Bohlen Pierce P scales.
I hope the missing scales will be integrated with the new Tennsy chip again!
These scales are more important to me than the many incomplete scales of the current selection
chysn
Ebotronix wrote:
rayultine wrote:
Good sleuthing Ebotronix! I hit this bug many versions ago but didn't really understand what had happened and couldn't reproduce!


I was looking for the scales behind the Bohlen Pierce P scales.
I hope the missing scales will be integrated with the new Tennsy chip again!
These scales are more important to me than the many incomplete scales of the current selection


What do you mean by "scales behind the Bohlen Pierce P scales?" All of the original O_C scales are in Hemisphere Suite, and are available to these quantizers.

If you need other scales, you can load any of the 4000 or so scales from the Scala archive to the user slots via MIDI, or use Scala or the Scale Editor.
Ebotronix
[/quote]

What do you mean by "scales behind the Bohlen Pierce P scales?" All of the original O_C scales are in Hemisphere Suite, and are available to these quantizers.

If you need other scales, you can load any of the 4000 or so scales from the Scala archive to the user slots via MIDI, or use Scala or the Scale Editor.[/quote]

since OC firmware 1.3.3 there are less scales available.
Retired scales
The following scales were removed in v1.3.3 to make more space for some additional programme code. Their definitions are still present in the source code, just commeneted out.

some of the original scales has more than 16 steps.
(exp: Quaterton scale = 24 steps,Mandelbaum 19steps)
chysn
Ah, gotcha! Yeah, Hemisphere 2 will likely use Teensy 3.5, and I’ll make sure to devote more space to scales.
naturarerum
my encoders randomly go in the reverse direction. right now my left encoder is reversed and the right one isn't. but a couple of power cycles ago the right one was also reversed. 1.7A

does the use of Teensy 3.5 mean we'll be able to pop on a new Teensy on our OCs? or will the breakout board be different?
chysn
naturarerum wrote:
my encoders randomly go in the reverse direction. right now my left encoder is reversed and the right one isn't. but a couple of power cycles ago the right one was also reversed. 1.7A


You can change the encoder direction via the calibration procedure, accessible via the Setup/About app, or by holding down the left encoder while powering up.

See these two pages:

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Setup
http://ornament-and-cri.me/calibration/

Quote:
does the use of Teensy 3.5 mean we'll be able to pop on a new Teensy on our OCs? or will the breakout board be different?


Hemisphere 2 will be built on µOC plus the Spoken Into The Void board by Tom Moxon:

https://github.com/PatternAgents/SitV

I'm actually not 100% sure whether I'll use 3.5 or 3.6 yet. Teensy 3.6 allows O_C to be a MIDI host, which would free us from computers. But I want to standardize Hemisphere 2 to one or the other, as I don't want to support a fragmented system.

In addition to just being generally "more," I'll be able to add actual oscillators, sampling, granular synthesis, and more advanced MIDI. And microSD card storage!

Currently, I'm looking at a beta release in the summer of 2019.
vantablack
chysn wrote:
I'm actually not 100% sure whether I'll use 3.5 or 3.6 yet. Teensy 3.6 allows O_C to be a MIDI host, which would free us from computers. But I want to standardize Hemisphere 2 to one or the other, as I don't want to support a fragmented system.

In addition to just being generally "more," I'll be able to add actual oscillators, sampling, granular synthesis, and more advanced MIDI. And microSD card storage!

Currently, I'm looking at a beta release in the summer of 2019.


woah Dead Banana applause
naturarerum
sorry if I'm dull, chysn, but I'm not seeing an option for the encoder direction when I go into calibration. I get pixel offset, then all the DAC voltages, the ADC stuff, CV scaling, screen blank. that's it.

that SitV stuff looks extremely promising. can't wait to see custom hemisphere panels to make this a properly different module nanners
chysn
naturarerum wrote:
sorry if I'm dull, chysn, but I'm not seeing an option for the encoder direction when I go into calibration. I get pixel offset, then all the DAC voltages, the ADC stuff, CV scaling, screen blank. that's it.


It's not a visible option, it's done with the Down button. Read "setting encoder directions" here:

http://ornament-and-cri.me/calibration/

Quote:
that SitV stuff looks extremely promising. can't wait to see custom hemisphere panels to make this a properly different module nanners


14HP and 8HP Hemisphere panels are in Front Panel Design format here:

https://github.com/Chysn/O_C-HemisphereSuite/tree/master/hardware

Front Panel Express does a good job, but they're not cheap.

Demo with 8HP panel: https://www.youtube.com/watch?v=K8Gv3Ht25Ao&t=77s
Demo with 14HP panel: https://www.youtube.com/watch?v=WrENiAQHhq0
Ebotronix
hi chsyn!
I use four OC Moduls, 3x with original Firmware 1.3.2
and one OC with Hemisphere.
there are much more scales (29) in the original 1.3.2 firmware,
is it possible to complete the Hemisphere Firmware with the 29 scales?
(Dual Quantizer App)
chysn
Ebotronix wrote:
hi chsyn!
I use four OC Moduls, 3x with original Firmware 1.3.2
and one OC with Hemisphere.
there are much more scales (29) in the original 1.3.2 firmware,
is it possible to complete the Hemisphere Firmware with the 29 scales?
(Dual Quantizer App)


There are 29 scales on the list that are commented out. This was done in an O_C commit in July of 2017 with the note "[DIET] Rationalise integer sequences and scales further so it fits with gcc 4.8 (sob!)"

I sort of interpret "fits with gcc 4.8 (sob!)" as meaning that gcc 4.8 may have unexpectedly increased the binary size of the firmware so some things had to go.

I uncommented these scales and compiled, and there's plenty of room for them in Hemisphere Suite, so I'll probably commit and release with these extra scales after doing some testing.
LazyCircuit
Quick question for Hemisphere users. Using metronome, do you get a BPM read out on screen?
chysn
LazyCircuit wrote:
Quick question for Hemisphere users. Using metronome, do you get a BPM read out on screen?


You'll see the selected tempo* in the Metronome applet. If the metronome is running but not selected, you'll see a little metronome icon, but not the numeric tempo. Does that answer your question?

________________
* BPM = tempo x multiplier
Shledge
Did you get around to sorting out it handling MIDI clock? I remember you said you had issues with it previously.
LazyCircuit
chysn wrote:
LazyCircuit wrote:
Quick question for Hemisphere users. Using metronome, do you get a BPM read out on screen?


You'll see the selected tempo* in the Metronome applet. If the metronome is running but not selected, you'll see a little metronome icon, but not the numeric tempo. Does that answer your question?

________________
* BPM = tempo x multiplier


Yes, thanks. I just didn’t know how precise I could be with my clocking.
coolshirtdotjpg
One feature that would be great, and I wonder how hard it would be to implement is: saving tempo in the metronome app, or even just another parameter with common tempos.

How imagine it is, some button combo allows you to save the tempo, it then appears in a list next to the tempo's you are currently scrolling through which can be scrolled through in the same way.

I often record stuff, occasionally forget to write down the tempo, and then it's sort of a nightmare to find the tempo I recorded at so I can add more to it later. Anyway, it's nothing major, but if you're working on 2.0, I think this would significantly improve the app.
chysn
Shledge wrote:
Did you get around to sorting out it handling MIDI clock? I remember you said you had issues with it previously.


It's a limitation of the USB MIDI Library for Teensy 3.2. The library can listen for MIDI clock and I can send CV triggers in response (MIDI to CV), but I can't listen to triggers and send MIDI clock with the library (CV to MIDI). That is, there's no way to construct a MIDI clock message.
chysn
coolshirtdotjpg wrote:
One feature that would be great, and I wonder how hard it would be to implement is: saving tempo in the metronome app, or even just another parameter with common tempos.


It's a good idea. For Hemisphere 2, I plan to make the metronome more central to the system, by assigning it to a long-hold of the Down button. That will replace the category filtering system, which I don't think is all that useful. So some form of preset system would fit in pretty well.
Shledge
chysn wrote:
Shledge wrote:
Did you get around to sorting out it handling MIDI clock? I remember you said you had issues with it previously.


It's a limitation of the USB MIDI Library for Teensy 3.2. The library can listen for MIDI clock and I can send CV triggers in response (MIDI to CV), but I can't listen to triggers and send MIDI clock with the library (CV to MIDI). That is, there's no way to construct a MIDI clock message.


Asking due to the fact that when I tested the MIDI clock to trigger part of Captain MIDI, it tends to glitch. It sometimes sends double pulses and there seems to be some offset, at least from initial testing. Was wondering if that was part of the issues you came across.
chysn
Shledge wrote:
Asking due to the fact that when I tested the MIDI clock to trigger part of Captain MIDI, it tends to glitch. It sometimes sends double pulses and there seems to be some offset, at least from initial testing. Was wondering if that was part of the issues you came across.


I haven't observed that with Ableton for Mac, but it doesn't surprise me. I was deeply reluctant to support MIDI beat clock at all, because nobody ever talks about how rock-solid it is, and I knew I'd be turning its problems into my problems.
Shledge
This was straight from a BSP via my Mio10 interface - generally it's rock solid on there. Maybe the teensy doesn't work well with the Mio10?
chysn
Shledge wrote:
This was straight from a BSP via my Mio10 interface - generally it's rock solid on there. Maybe the teensy doesn't work well with the Mio10?


I couldn’t say. My experience has been that it responds to MIDI messages immediately.
Shledge
It responds to normal MIDI messages fine, it's just MIDI clock that's wonky. hmmm.....
chysn
Shledge wrote:
It responds to normal MIDI messages fine, it's just MIDI clock that's wonky. hmmm.....


What I mean is, there's nothing special about MIDI clock messages from the Teensy's point-of-view. They're formatted and received exactly like other MIDI messages.
Kroaton
https://www.youtube.com/watch?v=gmdAXjviXDk - could a version of this be a worthwhile successor to the O_c platform?
lamateur2wiggles
Following in the footsteps of Rayultine and aaooaa, here are three PDFs of the Hemisphere 1.7 guide in various formats.

o_C-Hemisphere-1.7-Tablet.pdf is designed for tablets. Every app guide is on its own page and you can navigate to any page from any page.

o_C-Hemisphere-1.7-BookLetter.pdf and BookA4.pdf will print as nice little booklets, fairly convenient when away from the screen. BookLetter is to be printed on Letter size paper and BookA4 on, yes, A4 paper.

To print the booklets you'll need a duplex printer (prints on both side of the paper). Make sure to select the "Short Side Binding" option before you send the job or it's not going to work at all. The whole guide will use 16 pages in all.

Once printed, gather all the pages in the order they came out of the printer, fold the bunch in half and staple in the middle (or use some string or a weak elastic band). Et voila, there's your booklet...

Enjoy...
a100user
lamateur2wiggles wrote:
Following in the footsteps of Rayultine and aaooaa, here are three PDFs of the Hemisphere 1.7 guide in various formats.

o_C-Hemisphere-1.7-Tablet.pdf is designed for tablets. Every app guide is on its own page and you can navigate to any page from any page.

o_C-Hemisphere-1.7-BookLetter.pdf and BookA4.pdf will print as nice little booklets, fairly convenient when away from the screen. BookLetter is to be printed on Letter size paper and BookA4 on, yes, A4 paper.

To print the booklets you'll need a duplex printer (prints on both side of the paper). Make sure to select the "Short Side Binding" option before you send the job or it's not going to work at all. The whole guide will use 16 pages in all.

Once printed, gather all the pages in the order they came out of the printer, fold the bunch in half and staple in the middle (or use some string or a weak elastic band). Et voila, there's your booklet...

Enjoy...


Many thanks for that, a lot of work in there, look forward to running my Hemispheres against this as I know I have only scratched the surface.
chysn
Thank you!
lamateur2wiggles
chysn wrote:
Thank you!


Yes, thanks, but actually, thank YOU for the great firmware and the exhaustive documentation. What a treat...
ipassenger
lamateur2wiggles wrote:


Enjoy...


Thanks smile
Maximars
Hello there!

some little improvement proposals

- The Tuner should emit even voltages, like 2.000 and 3.000 volts.
- Schmittrigger should have a tuneable gate output like the Voltage app
(for instance, if a pitch-CV goes beyond a certain level, it emits -1V. In short, a bassline-cv-compressor hmmm..... )

I hope there is still a little free memory for this.

Thanks!
Inju
First, thank you, thank you, thank you for making Hemisphere. It is amazing. I had a request for an option to Volts. I currently have a Minibrute 2s where you can sequence specific volts out. The Minibrute 2s has four patch points you can sequence out from and thus each patch point is valuable to control my Eurorack.

I wanted to use only one patch point of the sequencer to trigger two devices by using voltage changes. My Kick at 4 volts and the Snare at 2 volts. Thus allowing me to sequence both the Kick and Snare from one patch point. However currently this will only work if I have the sequence of Kick on, then Snare off, because the Volt acts as a gate vs. a trigger. In other words once the gate is in the on position, I have to turn it off again in order for it to act as a trigger again, while a trigger quickly turns the gate on and off as a pulse.

Even with this solved there is the problem of triggering both at the same time. I believe I figured this out by using a third voltage, 7 Volts that would trigger both. I can do this if I use mult out to the second Volt and have that in a mult to the triggers of both the snare and drum. I believe there may also be a way to use the Maths for this but have not figgured it out yet.
chysn
Inju wrote:
First, thank you, thank you, thank you for making Hemisphere. It is amazing. I had a request for an option to Volts. I currently have a Minibrute 2s where you can sequence specific volts out. The Minibrute 2s has four patch points you can sequence out from and thus each patch point is valuable to control my Eurorack.

I wanted to use only one patch point of the sequencer to trigger two devices by using voltage changes. My Kick at 4 volts and the Snare at 2 volts. Thus allowing me to sequence both the Kick and Snare from one patch point. However currently this will only work if I have the sequence of Kick on, then Snare off, because the Volt acts as a gate vs. a trigger. In other words once the gate is in the on position, I have to turn it off again in order for it to act as a trigger again, while a trigger quickly turns the gate on and off as a pulse.

Even with this solved there is the problem of triggering both at the same time. I believe I figured this out by using a third voltage, 7 Volts that would trigger both. I can do this if I use mult out to the second Volt and have that in a mult to the triggers of both the snare and drum. I believe there may also be a way to use the Maths for this but have not figgured it out yet.


I read this a few times, and tried to wrap my head around what you're doing. If I understand correctly, you're using a single channel of a sequencer to send CV output, and you want to interpret a 4V voltage as a kick and a 2V voltage as a snare and 7V as both kick and snare.

I'm a little skeptical of this approach. You'd probably find it way more reliable to use separate tracks for snare and kick.

But I don't think you ever got around to saying what your request for Voltage is, exactly.
chysn
Maximars wrote:
- The Tuner should emit even voltages, like 2.000 and 3.000 volts.


I'd say there are plenty of ways to generate output voltages. I don't think there's a good reason to do it from a tuner.

Quote:
- Schmittrigger should have a tuneable gate output like the Voltage app
(for instance, if a pitch-CV goes beyond a certain level, it emits -1V. In short, a bassline-cv-compressor hmmm..... )


Like other logic modules, you can put its output through attenuation and/or inversion to get specific voltages.
Maximars
At the moment I’m struggling with the Schmitt Trigger. Can’t set the threshold of the second unit. No way to reach the settings of the second one. Both work always with the same setting, which makes no sense. Is it possibly a bug?

In my opinion, it would be a great help if the Tune applet emits Voltage and the outputs are not in use at the moment. The References App of o_C did emit even voltages for a good reason.

Whenever I tune my oscillators, I want to have a even voltage at their v/octave input, something like 2.000 volts. Most of the time patches are crowded and it is not so simple to allocate the oscillators straight voltages. So i would pull the 1v/oct cable, patch it into the CV output of the Tuner, tune it to C for instance, patch it back and you’re ready to go. Left jack 2V, right jack 3V, fixed, little thing, big improvement! One applet for all tuning purposes thumbs up
Inju
Hi Chysn,

I apologize for the confusion. I'm currently trying to figure out a way to easily sequence my drums using the Minibrute 2S. The Minibrute 2S has an amazing sequencer where you can simply depress 16 pads to send out one of several variables assignable to two outputs. These variables include;

• 1V, 2V, 5V, and 8V. These are referred to collectively as the xV type. (Each one of these can easily be changed up the voltage parameter you choose by turning the corresponding encoders which are above each of the 16 corresponding pads. In other words I can choose the pads to go up to 8 volts, but simply turning a nob above the pads, I can easily change their voltages from 1 to 8V. The increments is the voltage you select divided by 100)
• Env (for Envelope)
• Sin(e), Tri(angle), Saw Up, Saw Down, Sqr (Square), and SnH (Sample & Hold). These are referred to collectively as the LFO type.
• Pitch, Gate, Velocity and Pressure is also assignable to each pad and encoder.

The amazing thing about the Minibrute 2S is that for each of the 16 steps, I can assign a constant voltage to come out. Thus I can assign one step to 1.5 volts and another step to 7.5 volts, or any voltage that I want.

I am trying to figure out a way to trigger the Kick and Snare modules using only ONE of the TWO outputs. Currently I would need to assign each output channel, one to a Kick and the other to the Snare.

By allowing a trigger vs. a gate on the Volt in hemisphere, every time the step sends out a certain value, the Volt sends a gate signal to the trigger of my kick drum.

By allowing a trigger versus a gate, this has some amazing applications besides the use I wish to use it for. For instance, by using a mult to a KB out that sends out Voltz/Octave signals to your oscillator and the other to the Volts, I can program that every time a C2 note is pressed a kick drum would be triggered and if a C4 is pressed a snare drum would be triggered. However, since the volts being sent out by the keyboard or in my case sequencer/pads is a constant signal, the gate never closes unless it receives a different signal that is of a different voltage. In other words, every time a voltage is registered on the Volt, instead of it being a gate, allow it to simply be a trigger. Add a third variable where if the signal is above a certain amount, both triggers fire.

Hope that makes sense. Also, though no-one really talks about it. I love the scope widget you have in Hemisphere. I use it all the time and compare it to my scope on my laptop. The spinning of the encoder to get the resolution just right takes time for each wave, but still it is amazing! Once again thank you. If you have a Patreon or Paypal account, please pm it to me and I'll send you some funds. smile
chysn
Maximars wrote:
At the moment I’m struggling with the Schmitt Trigger. Can’t set the threshold of the second unit. No way to reach the settings of the second one. Both work always with the same setting, which makes no sense. Is it possibly a bug?


It's as designed. Both channels use the same threshold. See the documentation here: https://github.com/Chysn/O_C-HemisphereSuite/wiki/Schmitt-Trigger

Quote:
Whenever I tune my oscillators, I want to have a even voltage at their v/octave input, something like 2.000 volts. Most of the time patches are crowded and it is not so simple to allocate the oscillators straight voltages. So i would pull the 1v/oct cable, patch it into the CV output of the Tuner, tune it to C for instance, patch it back and you’re ready to go. Left jack 2V, right jack 3V, fixed, little thing, big improvement! One applet for all tuning purposes thumbs up


Yeah, I can understand that. It would be easy enough to do in a future release.
Mdashdotdashn
@chrsn: I'm curious if you found anyway to measure the performance/intensity of calculation of any app. I've currently have a terrain synthesis oscillator that's really nice but it is also close to brining the teensy to its knees.

One approach would be to bang a digital output with 0 at the start of the audio interrupt and 1 at the end and pass it through a slew limiter. But before messing up with hardware, I was wondering if there was any built-in solution you coul see.
chysn
Mdashdotdashn wrote:
@chrsn: I'm curious if you found anyway to measure the performance/intensity of calculation of any app.


While your app is running, return to the main menu by holding down the right encoder. While in the main menu, your app will continue running. From the main menu, press the left encoder button. This will bring you to a sort of diagnostic screen. The CORE line will tell you the processor's usage percentage.
Mdashdotdashn
chysn wrote:

While your app is running, return to the main menu by holding down the right encoder. While in the main menu, your app will continue running. From the main menu, press the left encoder button. This will bring you to a sort of diagnostic screen. The CORE line will tell you the processor's usage percentage.


Fantastic ! Thanks. Urg. 80%
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
One feature that would be great, and I wonder how hard it would be to implement is: saving tempo in the metronome app, or even just another parameter with common tempos.


It's a good idea. For Hemisphere 2, I plan to make the metronome more central to the system, by assigning it to a long-hold of the Down button. That will replace the category filtering system, which I don't think is all that useful. So some form of preset system would fit in pretty well.


That would be amazing. Thank you, thank you, thank you!
coolshirtdotjpg
chysn wrote:
Inju wrote:
First, thank you, thank you, thank you for making Hemisphere. It is amazing. I had a request for an option to Volts. I currently have a Minibrute 2s where you can sequence specific volts out. The Minibrute 2s has four patch points you can sequence out from and thus each patch point is valuable to control my Eurorack.

I wanted to use only one patch point of the sequencer to trigger two devices by using voltage changes. My Kick at 4 volts and the Snare at 2 volts. Thus allowing me to sequence both the Kick and Snare from one patch point. However currently this will only work if I have the sequence of Kick on, then Snare off, because the Volt acts as a gate vs. a trigger. In other words once the gate is in the on position, I have to turn it off again in order for it to act as a trigger again, while a trigger quickly turns the gate on and off as a pulse.

Even with this solved there is the problem of triggering both at the same time. I believe I figured this out by using a third voltage, 7 Volts that would trigger both. I can do this if I use mult out to the second Volt and have that in a mult to the triggers of both the snare and drum. I believe there may also be a way to use the Maths for this but have not figgured it out yet.


I read this a few times, and tried to wrap my head around what you're doing. If I understand correctly, you're using a single channel of a sequencer to send CV output, and you want to interpret a 4V voltage as a kick and a 2V voltage as a snare and 7V as both kick and snare.

I'm a little skeptical of this approach. You'd probably find it way more reliable to use separate tracks for snare and kick.

But I don't think you ever got around to saying what your request for Voltage is, exactly.


This is a feature of the stillson hammer mk1, and it's pretty cool, basic just mapping a threshold to trigger a gate output. Unfortunately it doesn't seem particularly well suited to hemispheres since it really only gets interesting when you have at least four outputs. This would an interesting idea for the original ornament and crime where X volts cause a gate at X output, or alternatively Y volts opens up inputs X through Y. This doesn't seem particularly hard to implement, but just doesn't seem that useful with only the two gate outputs for each applet.

[edit, I say it doesn't seem that hard to implement with precisely zero knowledge of programming for O+C so of course, forgive me if I'm being ignorant.]
mxmxmx
Mdashdotdashn wrote:
@chrsn: I'm curious if you found anyway to measure the performance/intensity of calculation of any app. I've currently have a terrain synthesis oscillator that's really nice but it is also close to brining the teensy to its knees.


fwiw, it's probably obvious but you generally don't want to use floats on this thing; MK20 doesn't have an FPU. extern/dspinst.h includes a bunch of things you might find useful ... (not sure which oscillator we're talking about, of course, but superficial sleuthing turns up this).
coolshirtdotjpg
Another thought, the O+C hardware seems perfectly suited to some sort of quadrophonic panning control, but again hemispheres might not be the right way to implement it.

I'm imagining this in combination with a quad VCA setup so that each of the 4 outputs of O+C are controlling each channel of the VCA to allow for quadrophonic panning/mixing.

Chysn, is there a way that the input of one app could be "normalled" to another, if so, you could have each side of a quadrophonic app sending voltage to the appropriate outputs with the right side receiving X and Y information from the left side?

If so that would be super interesting.

Also, I hope that my floating ideas is not seen as "gimme gimme gimme" feature requests. Since you seem to be really open to ideas from the community, I am just throwing out ideas and seeing what sticks, it is interesting to see what is feasible, what features seem useful to people other than myself etc. Like most folks my ideas far outstrip my ability to implement any of them in code.
chysn
coolshirtdotjpg wrote:
Chysn, is there a way that the input of one app could be "normalled" to another, if so, you could have each side of a quadrophonic app sending voltage to the appropriate outputs with the right side receiving X and Y information from the left side?


Possible, sure. That's vaguely how master clock forwarding works. The first digital input gets the real data, and the third digital input becomes a sort of "virtual input."

Realistically, there are interface limitations that make it hard to do that kind of thing elegantly and frustration-free, so I've kept fancy things to a minimum.

Quote:
Also, I hope that my floating ideas is not seen as "gimme gimme gimme" feature requests.


Not at all, I understand. At this point, memory being what it is, I sort of have the luxury of having little latitude to do anything else with the firmware. smile
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
Chysn, is there a way that the input of one app could be "normalled" to another, if so, you could have each side of a quadrophonic app sending voltage to the appropriate outputs with the right side receiving X and Y information from the left side?


Possible, sure. That's vaguely how master clock forwarding works. The first digital input gets the real data, and the third digital input becomes a sort of "virtual input."

Realistically, there are interface limitations that make it hard to do that kind of thing elegantly and frustration-free, so I've kept fancy things to a minimum.

Quote:
Also, I hope that my floating ideas is not seen as "gimme gimme gimme" feature requests.


Not at all, I understand. At this point, memory being what it is, I sort of have the luxury of having little latitude to do anything else with the firmware. smile


Haha, that sounds about right! Just out of curiousity, is it possible for the user to compile the firmware with only certain applets, so that if I ever do what I keep threatening to do and come up with an applet for it, could I swap out something I wouldn't use out to save on memory?
chysn
coolshirtdotjpg wrote:
Just out of curiousity, is it possible for the user to compile the firmware with only certain applets, so that if I ever do what I keep threatening to do and come up with an applet for it, could I swap out something I wouldn't use out to save on memory?


Yeah, of course. See the hemisphere_config.h file. Remove, or comment out, the unwanted applets from this file, and then remove the associated HEM_applet.ino file from the project.
Mdashdotdashn
chysn wrote:
coolshirtdotjpg wrote:
Just out of curiousity, is it possible for the user to compile the firmware with only certain applets, so that if I ever do what I keep threatening to do and come up with an applet for it, could I swap out something I wouldn't use out to save on memory?


Yeah, of course. See the hemisphere_config.h file. Remove, or comment out, the unwanted applets from this file, and then remove the associated HEM_applet.ino file from the project.


Also you can remove all applications but hemisphere and the configuration apps.
Mdashdotdashn
mxmxmx wrote:
Mdashdotdashn wrote:
@chrsn: I'm curious if you found anyway to measure the performance/intensity of calculation of any app. I've currently have a terrain synthesis oscillator that's really nice but it is also close to brining the teensy to its knees.


fwiw, it's probably obvious but you generally don't want to use floats on this thing; MK20 doesn't have an FPU. extern/dspinst.h includes a bunch of things you might find useful ... (not sure which oscillator we're talking about, of course, but superficial sleuthing turns up this).


Your search lead to the right area although it wasn't really in par with my current state. The oscillator is here and involves currently multiple evaluations of perlin noise. I'm totally aware of the lack of FPU, I've got used to fixed point back in the day the raspberry pi was in the same situation and it's used heavily in all my processing (hence the templated code). There's a fairly comprehensive class for it here

Edit: It seems a "blank" app selected in both side uses still 25% of the core. Is that the amount of time it takes to peek/poke data to the outside ?
mxmxmx
Mdashdotdashn wrote:


Your search lead to the right area although it wasn't really in par with my current state.


alright then.

Mdashdotdashn wrote:

Edit: It seems a "blank" app selected in both side uses still 25% of the core. Is that the amount of time it takes to peek/poke data to the outside ?


possible, idk. there's the DAC (~4 us, IIRC), ADC (dunno), display (mostly DMA, but not all of it), polling of inputs etc. o_C v1.35 uses DMA for the ADC now, too, i don't think hemispheres does; so that might help bring that number down a little, but probably not much. either way, (sorry i keep rehearsing this...) OC was designed to be *quantizer*, there's only one SPI bus on MK20, and so on; it'll be of limited use for DSP, even within the limits of what teensy 3.2 can/could theoretically do.

i have a teensy 4 for beta testing on my desk now, that might open up some options (if i/we/someone actually port the code) but some of the bottlenecks (1 bus for DAC + OLED) are set in stone, so there's that.
Mdashdotdashn
[quote="mxmxmx"]
Mdashdotdashn wrote:

either way, (sorry i keep rehearsing this...) OC was designed to be *quantizer*, there's only one SPI bus on MK20, and so on; it'll be of limited use for DSP, even within the limits of what teensy 3.2 can/could theoretically do


I'm complaining at all, the O_C is a wonderful platform and there's *tons* of applications that uses low dsp power that can be used. This was more of an experiment to see how far I could take it. But I've got a bag full of other things to try.
pld
Mdashdotdashn wrote:
Edit: It seems a "blank" app selected in both side uses still 25% of the core. Is that the amount of time it takes to peek/poke data to the outside ?

The time measured is the total execution time (in cycles) of the core interrupt handler, so that includes all the DAC, ADC filtering, framebuffer switcheroo, setup next DMA transfer to screen, etc. I only have an ancient 'scope shot that shows ~20us of pre-processing but that's probably gone down a bit.
All the app computations have to happen in the window where the screen DMA transfer takes place so there's little wiggle room since the whole ISR has to complete in < 60us, and it can't use 100% either since the screen redraws and UI happen in lower priority interrupts/asynchronously.
Ebotronix
hi chysn!
Trending dual slope detector app
ChgValue: The assigned output emits a trigger when the signal changes its value by more than 1/4 semitone (or about .02V)

is it possible to reduce the 0,2 V to get a trigger from the microtonal scales?
Timmy
Mdashdotdashn wrote:
I've currently have a terrain synthesis oscillator that's really nice but it is also close to brining the teensy to its knees.


Yes, Perlin noise is an intriguing possibility. Here is some discussion of implementing it on the MI Peaks and on O&C from about 4 years ago (neither implementations ever happened): https://forum.mutable-instruments.net/t/dead-mans-catch-modified-firmw are-for-peaks/7248/203 which was inspired by this blog post: https://dmachinery.wordpress.com/2014/08/15/perlin-pattern-generator/

Of course, that was as a modulation source operating in the time domain. I suspect it might be very useful as a source for quantisers. Not sure that perlin noise offers anything special as an audio noise source in the frequency domain, though, and in any case, with the 16.6kHz DAC sample rate on the O&C, it is hardly worth bothering unless you like Merzbow (although I am told by Merzbow connoisseurs that the Viznutcracker, sweet! app in O&C sounds nothing like Merzbow's best...oh well).

Doesn't Sonic Potions produce a module or modules that use Perlin noise?
chysn
Ebotronix wrote:
hi chysn!
Trending dual slope detector app
ChgValue: The assigned output emits a trigger when the signal changes its value by more than 1/4 semitone (or about .02V)

is it possible to reduce the 0,2 V to get a trigger from the microtonal scales?


Eh... You can change the value and recompile if you want, but that value is based on quite a bit of experimentation. I don't "smooth" incoming CV like the native firmware does (did?), so if you go much lower, the tradeoff is false triggers because of the ADC noise, or whatever.
coolshirtdotjpg
chysn wrote:
coolshirtdotjpg wrote:
Just out of curiousity, is it possible for the user to compile the firmware with only certain applets, so that if I ever do what I keep threatening to do and come up with an applet for it, could I swap out something I wouldn't use out to save on memory?


Yeah, of course. See the hemisphere_config.h file. Remove, or comment out, the unwanted applets from this file, and then remove the associated HEM_applet.ino file from the project.


That's great. Forgive my ignorance on that. Very exciting work that you've done here.
Mdashdotdashn
Timmy wrote:
. Not sure that perlin noise offers anything special as an audio noise source in the frequency domain, though, and in any case, with the 16.6kHz DAC sample rate on the O&C, it is hardly worth bothering unless you like Merzbow (although I am told by Merzbow connoisseurs that the Viznutcracker, sweet! app in O&C sounds nothing like Merzbow's best...oh well).


I really depends how you use perlin. Mine is an ever evolving low frequency oscillator with a parameter for harmonics. Although I'd like to stuff it in a more powerful cpu at some point, it sound pretty nice on the O_C already:

https://www.instagram.com/p/BuOCqVII2y_/

@chrysn: On another topic, I would like to have a "triggerable" shuffle where by shuffle amount is not changing at every clock but triggered by digital input 1. Is that something you'd see added to the current shuffle ? If so, I could do a PR rather than rolling my own version.
chysn
Mdashdotdashn wrote:
Although I'd like to stuff it in a more powerful cpu at some point, it sound pretty nice on the O_C already:

https://www.instagram.com/p/BuOCqVII2y_/


That's really cool!

Quote:
@chrysn: On another topic, I would like to have a "triggerable" shuffle where by shuffle amount is not changing at every clock but triggered by digital input 1. Is that something you'd see added to the current shuffle ? If so, I could do a PR rather than rolling my own version.


I don't completely understand the concept of a triggerable shuffle. Are you talking about suppressing the trigger delay unless a gate is high, or something like that?
Mdashdotdashn
chysn wrote:

I don't completely understand the concept of a triggerable shuffle. Are you talking about suppressing the trigger delay unless a gate is high, or something like that?


Basically have one delay if Gate is high, the other if gate is low. This mode wouldn't need a reset so it could use Dig1
qbassa@gmail.com
please delete this
qbassa@gmail.com
Hi, what happened with the idea of slope detector that was mentioned in this thread months ago. I am especially asking about gate to trigger converter (on both slopes) , but proposed algorithm had broader range.
Best regards!

Maximars wrote:
Hi chysn

Thank you for adding the AD Envelope!

Another idea: Slope Dedector App

Option to select one of 4 states on both outputs:

CV is rising, Gate
CV is falling, Gate
CV is steady, Gate
CV changes direction, Trigger

Maybe passing CV1 to the second seide of Hemisphere

Somehow like:

Ladik, J-110 Derivator
or

Elby Desings, Slope Detector

I do not know if the sensitivity controller is necessary with a digital module.

Eny chance for this app?

Tank You!
chysn
qbassa@gmail.com wrote:
Hi, what happened with the idea of slope detector that was mentioned in this thread months ago. I am especially asking about gate to trigger converter (on both slopes) , but proposed algorithm had broader range.


The Hemisphere slope detector applet is called "Trending":

https://github.com/Chysn/O_C-HemisphereSuite/wiki/Trending
Bath House
Is there an app that will give me a fixed voltage output/offset, that I can sweep manually?
chysn
Bath House wrote:
Is there an app that will give me a fixed voltage output/offset, that I can sweep manually?


https://github.com/Chysn/O_C-HemisphereSuite/wiki/Voltage
Bath House
chysn wrote:
Bath House wrote:
Is there an app that will give me a fixed voltage output/offset, that I can sweep manually?


https://github.com/Chysn/O_C-HemisphereSuite/wiki/Voltage


I read that one, but it sounds like it needs a gate to be active, right?
chysn
Bath House wrote:
chysn wrote:
Bath House wrote:
Is there an app that will give me a fixed voltage output/offset, that I can sweep manually?


https://github.com/Chysn/O_C-HemisphereSuite/wiki/Voltage


I read that one, but it sounds like it needs a gate to be active, right?


Nope. It depends on whether the channel is set to G-On or G-Off.
Ebotronix
hi chysn!
is it possible to program an unipolar inverter app?
positive voltage only
examples
input 5V output 0C
in 4,85V out 0,15V
in 3V out 2V
in 2,5V out 2,5V
in 1,06V out 3,94V
in 0V out 5V

any news from the removed Yarns Scales?
(30 scales from OC firmware 1.3.2)
chysn
Ebotronix wrote:
hi chysn!
is it possible to program an unipolar inverter app?
positive voltage only
examples
input 5V output 0C
in 4,85V out 0,15V
in 3V out 2V
in 2,5V out 2,5V
in 1,06V out 3,94V
in 0V out 5V


It would probably work best as a new Dual Calc function than a separate applet. Right now, you can put 5V into one of the inputs and use the Diff function of Dual Calc.

Quote:
any news from the removed Yarns Scales?
(30 scales from OC firmware 1.3.2)


They're committed to GitHub, if you can compile it yourself. Otherwise, they'll go out with the 1.8 release, which should be coming pretty soon, after the Buchla updates are done. I'm getting some help (edit: a ton of help) with that, so it might be sooner than I thought.
qbassa@gmail.com
chysn wrote:

The Hemisphere slope detector applet is called "Trending":


Heh, thanks! Couldn't find it....wink
qbassa@gmail.com
Chysn, is there any chance you can change or add additional mode of drawing scope, that would use lines not points. I also wondered if having 2 scopes per one applet would be possible (bpm setting could be displayed temporarily or not at all). Still being impressed by amount of work Hemisphere took. Best!
chysn
qbassa@gmail.com wrote:
Chysn, is there any chance you can change or add additional mode of drawing scope, that would use lines not points.


What do you mean by “lines not points” here?
qbassa@gmail.com
chysn wrote:
What do you mean by “lines not points” here?

I mean that there could be line drawn to every new position, so the graph would be more solid and that would add vertical lines when the signal is like square. Now chart has discontinuities because it is made point by point.
chysn
qbassa@gmail.com wrote:
chysn wrote:
What do you mean by “lines not points” here?

I mean that there could be line drawn to every new position, so the graph would be more solid and that would add vertical lines when the signal is like square. Now chart has discontinuities because it is made point by point.


I'll give it a try. If it's not part of a future release, it's because I didn't like the way it looked smile
Mdashdotdashn
I had a little down-time to spend on the O_C and because, well, I could, I did a little casio VL-1 preset machine for Hemisphere grin

https://www.instagram.com/p/BvLzpzEoJcI/

It's always nice to see how "much" one can do with this little box !
chysn
Mdashdotdashn wrote:
I had a little down-time to spend on the O_C and because, well, I could, I did a little casio VL-1 preset machine for Hemisphere grin

https://www.instagram.com/p/BvLzpzEoJcI/

It's always nice to see how "much" one can do with this little box !


That's amazing! Makes me want to break out my PT-1!
mcpepe
Do you think a noise applet possible?
Maybe with options for pink, white,...
chysn
mcpepe wrote:
Do you think a noise applet possible?
Maybe with options for pink, white,...


The Calculate applet's "Rnd" operator will output noise in the audible range, as long as it's not clocked. If you feed an oscillator's output into the corresponding digital input, the noise source becomes clocked, and you can control its frequency with your oscillator.
mcpepe
chysn wrote:
mcpepe wrote:
Do you think a noise applet possible?
Maybe with options for pink, white,...


The Calculate applet's "Rnd" operator will output noise in the audible range, as long as it's not clocked. If you feed an oscillator's output into the corresponding digital input, the noise source becomes clocked, and you can control its frequency with your oscillator.


Great. Thanks!
Shledge
Could boots 'n cats act as a noise source?
chysn
Shledge wrote:
Could boots 'n cats act as a noise source?


Sure, but not a continuous one.
rayultine
chysn: what kind of excitement should we be getting excited for in hemisphere 1.8?
chysn
rayultine wrote:
chysn: what kind of excitement should we be getting excited for in hemisphere 1.8?


Good question! Keep in mind that 1.7 took us to over 97% memory usage, so 1.8 is mostly fulfilling a grab-bag of small feature requests, hammering on bugs, and integrating pull request contributions from other developers.

However, there are two significant features in 1.8:

(1) Support for the Northern Lights 2OC (Buchla-format) module. I'm getting lots of help with this from an actual Buchla user, on both the development and testing sides.

(2) A more tightly-integrated internal clocking system. You'll be able to call up the clock setup mode at any time by long-pressing the DOWN button. This replaces the category filtering system. Category filtering was a solution in search of a problem, I think, and having instant access to the clock will be much more useful.

Also, a bunch of scales that were removed from OC 3.4 (because of memory constraints) have been replaced.

I'm expecting the release to be near the end of April.
bemushroomed
Great to hear, really looking forward to 1.8!
Southfork
1.8 Yes! I find myself having my u_oc module mainly in hemisphere these days thumbs up
Sync
Sounds like it would be worth setting up a build config file that you can use to pick and choose what modules you want to include-- in order to make room for additional modules at the expense of modules you may not need. Maybe that already exists?
chysn
Sync wrote:
Sounds like it would be worth setting up a build config file that you can use to pick and choose what modules you want to include-- in order to make room for additional modules at the expense of modules you may not need. Maybe that already exists?


Anybody who compiles it themselves can use the hemisphere_config.h file to choose which applets to leave out. But that's not the solution for musicians who just want to use it without fussing with the IDE.

Another option is a web-based interface that knows the size of each applet, allows you to pick enough to fill memory, and then compiles a customized hexfile for you to download. This is technically really easy, but the mere thought of actually doing such a project bores me to the soul, and isn't worth the no money involved.

No, the only real solution is more memory, which will come from Teensy 3.5 or 3.6, which will come from using Spoken Into The Void. I hope to get a good start on that over the summer.
Maximars
As long there’s no more memory left, maybe you could rework the firmware of „Temps Utile“. spinning spinning spinning spinning

There wouldn't be a lack of good ideas. hihi

I’m just kidding.

Thank you for everything!
qbassa@gmail.com
Is VectorEG envelope speed cv controlled or the speed is only set with the waveform editor? If not - Is it possible to add speed modification control? It could be -5V is 5 times slower EG, +5V is 5 times faster. Best!
nectarios


After procrastinating, I finally got round to doing it. I have a uOC that I left with the "normal" firmware.

Time to search and read up on the site instructions.
chysn
qbassa@gmail.com wrote:
Is VectorEG envelope speed cv controlled or the speed is only set with the waveform editor? If not - Is it possible to add speed modification control? It could be -5V is 5 times slower EG, +5V is 5 times faster. Best!


The speed is controlled by the frequency (in Hz) from the panel, not from the waveform editor. That its speed is in Hz as opposed to seconds (or ms) is a by-product of its kinship with the other Vector applets.

The speed is not currently CV-controllable. The CV inputs are unused, and under normal circumstances, back in 2018, I'd have said, "Yeah, that's a good idea, I'm going to do that." But now, memory is really tight, with literally hundreds of bytes remaining. So I can't promise that this can happen. It is near the top of the list, though.

nectarios wrote:
After procrastinating, I finally got round to doing it.


Thanks for trying it out!
Dibek
I really liked the possibility to automate recording and playback on the cv recorder but it seems to be gone in the latest version (haven't updated for a while)

Am I missing something and if not, might this feature be reintroduced?

Thanks!
chysn
Dibek wrote:
I really liked the possibility to automate recording and playback on the cv recorder but it seems to be gone in the latest version (haven't updated for a while)

Am I missing something and if not, might this feature be reintroduced?


Not sure what you mean. All four inputs of CV Recorder have always been the same. Digital 1 advances to the next step and Digital 2 resets, and the CV inputs are the inputs for each track.
Dibek
chysn wrote:
Dibek wrote:
I really liked the possibility to automate recording and playback on the cv recorder but it seems to be gone in the latest version (haven't updated for a while)

Am I missing something and if not, might this feature be reintroduced?


Not sure what you mean. All four inputs of CV Recorder have always been the same. Digital 1 advances to the next step and Digital 2 resets, and the CV inputs are the inputs for each track.


Sorry, I managed to confuse CV Recorder with LoFi Tape, it´s been a long day yesterday smile
chysn
Dibek wrote:
Sorry, I managed to confuse CV Recorder with LoFi Tape, it´s been a long day yesterday smile


Oh! Cool.
kall
chysn wrote:
Another option is a web-based interface that knows the size of each applet, allows you to pick enough to fill memory, and then compiles a customized hexfile for you to download. This is technically really easy, but the mere thought of actually doing such a project bores me to the soul, and isn't worth the no money involved.


If that's the direction you want to go, and have more applets to add, doing the web interface / compiler service part of that sounds like fun to me.

I'm not great with embedded stuff though. Do you already have a way to build a hex file from the command line?
chysn
kall wrote:
Do you already have a way to build a hex file from the command line?


No, and I'm only speculating wildly that it's even possible with the TeensyDuino stuff...

https://lifehacker.com/arduino-ide-now-supports-building-software-in-t he-comma-1740786363