Fish Questions

home of post-70s electronics

Moderators: Kent, bartlebooth

User avatar
bartlebooth
Super Deluxe Wiggler
Posts: 1059
Joined: Sat Dec 13, 2008 2:26 pm
Location: DC

Post by bartlebooth » Wed Jan 15, 2014 10:45 am

VortexRanger wrote:K here's a question: what is the default srate? 127? 0? Somewhere in between?
default srate is 16

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Wed Jan 15, 2014 4:10 pm

Awesome thanx! One more question, this about short since I'm not near the computer and I'm impatient to know: is the bigg value the same as a regular (non-short) value and the small adds 256 steps between? E.g. If I had a bigg of 14 and small of 20 (in dirac) it would be slightly over a (non-short) value of 14?
:hobbes:

User avatar
windspirit
Super Deluxe Wiggler
Posts: 3014
Joined: Thu Sep 20, 2012 11:43 pm
Location: krellifornia

Post by windspirit » Wed Jan 15, 2014 4:39 pm

stellvia wrote:Question of my own:

How would you go about mixing or changing the volume levels of sound sources? Seems like if you have 2 or more (or sometimes even 1) sound sources playing, it seems like it is clipping or distorting a little, and it would be nice to be able to lower the level going into the left and right DAC's. Tried using modo to do that (in dirac) but doesn't seem to work for turning levels down, only turning them up.
Theres also a really cool compressor opcode called [press]. Found it useful in a few situations.

Ok here's the real question: when will there be a tutorial on how to write our own opcodes? :D. I realize that the fish system has had its surface barely scratched but I already have a few ideas for some simple stuff that I would find useful.

Also perhaps some preset user-defined "widgets" would be nice. Like a mini-patch in PD where you have a user defined chunk of fish code that only shows up as a single poisson unless you double click it. User defined inputs and outputs etc.

I am already super stocked on this system, but Im pretty sure that further modification is part of the plan since it's open source yes?

User avatar
windspirit
Super Deluxe Wiggler
Posts: 3014
Joined: Thu Sep 20, 2012 11:43 pm
Location: krellifornia

Post by windspirit » Wed Jan 15, 2014 4:54 pm

@SHLISP "WANILLA" ASSEMBLAGE
@COPYLEFT SHBOBO PLZ MUTATE

ORIGINATE

soseDEEP .req workTRI

.macro SCANETTO
mov scanDEEP, 0 @depth
1:
.rept 4
ldrb lispRET, [lispCOD], 1
cmp lispRET, 0
it eq
subeq scanDEEP, 1 @un-deeper
cmp lispRET, 0xFF
it eq
addeq scanDEEP, 1 @deeper
cmp scanDEEP, 1
blt 2f
.endr
b 1b
.endm

.macro SAUCE
VAKLORDEIGHT sauceVALE
POPSEX
RECTARET
SYNTHLOADER workONE, saucePOSZ
ldrb workTWO, [workONE, lispMEX]
add workTWO, workTWO, 1
cmp workTWO, lispRET, asr 8
ittte gt @executo
movgt workTWO, 0
strbgt workTWO, [workONE, lispMEX]
movgt lispACC, 0
ble 2f
POPSEX
mov lispACC, lispRET
strh lispACC, [lispWOR, lispMEX, lsl 1]
LISPMULADD
2: @execute sexpressions
strb workTWO, [workONE, lispMEX]
SCANETTO
2:
LISPMULADD
.endm
INSTORG SAUCE 0
INSTORG SAUCE 4

.macro SALSA @sauce per s1 ... sn
VAKLORDEIGHT salsaVALE
POPSEX

TRIG_IDEE_TOO lispRET, workTWO, salsatrig, workTRI
ite ne
movne lispACC, 0
beq 2f
POPSEX
mov lispACC, lispRET
strh lispACC, [lispWOR, lispMEX, lsl 1]
LISPMULADD
2: @execute sexpressions
SCANETTO
2:
LISPMULADD
.endm
INSTORG SALSA 8
INSTORG SALSA 0xC
As I expected, the arm assembly code for the opcodes is undocumented and uses lots of random variables and macros like "SEXY" and "VAKLORDEIGHT". :goo:

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Wed Jan 15, 2014 6:48 pm

VortexRanger wrote:Awesome thanx! One more question, this about short since I'm not near the computer and I'm impatient to know: is the bigg value the same as a regular (non-short) value and the small adds 256 steps between? E.g. If I had a bigg of 14 and small of 20 (in dirac) it would be slightly over a (non-short) value of 14?
Quick experiment and it seems this is indeed the behavior.
:hobbes:

User avatar
bartlebooth
Super Deluxe Wiggler
Posts: 1059
Joined: Sat Dec 13, 2008 2:26 pm
Location: DC

Post by bartlebooth » Wed Jan 15, 2014 8:45 pm

@vortexranger: yeah that sounds right, bigg is like the normal number and then up to 256 (or 128 in dirac) for small, to get more precision

@ windspirit: its definitely intended for people to be able to write their own opcodes or even their own matrices so everything needed should be in shnth483. if you have questions about getting started with this the best approach would be to post them in the shbobo technical forum because peter watches that or else just email him directly.

User avatar
stellvia
Wibbly-wobbly timey-wimey
Posts: 852
Joined: Mon Jun 04, 2012 12:26 am
Location: Perth, Aus

Post by stellvia » Sun Jan 19, 2014 2:43 am

Another quick fish question.

Playing around with nume and deno ratios to get a chromatic scale based on whatever the srate happens to be at the time. Put this short patch together to test out cycling through a chromatic scale using two togo's and a minor
{(left (right (hornh (togo )(togob )25 )))
(togo (minor )0 80 80 90 90 90 92 90 90 120 120 126 120 120 );chromatic nume
(togob (minor )0 80 75 80 75 72 69 64 60 75 72 70 64 60 );chromatic deno
}
Unfortunately, this doesn't seem to work. I've checked all the ratios individually, and they are all correct multiples of ratios to achieve the desired frequency. However, when stepping through the togos, I get a repeatable pattern, but it is not a chromatic scale as desired. From examining the problem, it appears that having the same value repeated immediately in the list causes problems with the togo's stepping.

Changing the signs (of both the nume and the deno) so that no value is repeated (i.e. sequentially all values are a different discrete value) fixes the problem:
{(left (right (hornh (togo )(togob )25 )))
(togo (minor )0 80 -80 90 -90 90 92 90 -90 120 -120 126 -120 120 );chromatic nume
(togob (minor )0 80 -75 80 -75 72 69 64 -60 75 -72 70 -64 60 );chromatic deno
}
Just wanted to check if this is the expected behaviour of the togos.

User avatar
krz
Ultra Wiggler
Posts: 873
Joined: Tue May 18, 2010 11:16 pm
Location: Canada
Contact:

Post by krz » Sun Jan 19, 2014 4:41 am

Sidestepping togo's? Sometimes, in a bowl with multiple sitches, I'll get a dead end at the final situation, and instead of being able to change sitch totally, a togo makes a hybrid situation, somehow. If this happens now I just power off, and start again at the start. This happens less often with AC power - which I think is the strangest part. Most situations I've put in a, (or simultaneous a & g) button(s) for scrolling thru.

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Sun Jan 19, 2014 10:40 pm

krz wrote:Sidestepping togo's? Sometimes, in a bowl with multiple sitches, I'll get a dead end at the final situation, and instead of being able to change sitch totally, a togo makes a hybrid situation, somehow. If this happens now I just power off, and start again at the start. This happens less often with AC power - which I think is the strangest part. Most situations I've put in a, (or simultaneous a & g) button(s) for scrolling thru.
Are you talking about using togo to bend? This is something I've been curious about, I would think each situation would start the togo over as it was initiated, thus making it impossible to reach the end of the togo, creating some sort of feedback loop. Thus I've pretty much only bent with total unpredictability (dusty salsa) or values that will stay consistent across situations (bars and corps).
:hobbes:

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Sun Jan 19, 2014 10:43 pm

stellvia wrote:Another quick fish question.

Playing around with nume and deno ratios to get a chromatic scale based on whatever the srate happens to be at the time. Put this short patch together to test out cycling through a chromatic scale using two togo's and a minor
{(left (right (hornh (togo )(togob )25 )))
(togo (minor )0 80 80 90 90 90 92 90 90 120 120 126 120 120 );chromatic nume
(togob (minor )0 80 75 80 75 72 69 64 60 75 72 70 64 60 );chromatic deno
}
Unfortunately, this doesn't seem to work. I've checked all the ratios individually, and they are all correct multiples of ratios to achieve the desired frequency. However, when stepping through the togos, I get a repeatable pattern, but it is not a chromatic scale as desired. From examining the problem, it appears that having the same value repeated immediately in the list causes problems with the togo's stepping.

Changing the signs (of both the nume and the deno) so that no value is repeated (i.e. sequentially all values are a different discrete value) fixes the problem:
{(left (right (hornh (togo )(togob )25 )))
(togo (minor )0 80 -80 90 -90 90 92 90 -90 120 -120 126 -120 120 );chromatic nume
(togob (minor )0 80 -75 80 -75 72 69 64 -60 75 -72 70 -64 60 );chromatic deno
}
Just wanted to check if this is the expected behaviour of the togos.
I don't have the shnth booted up with me right now. So are you saying it skips over repeat values? Like in your first example goes straight to 90 after the first 80? That's kind of a bummer, though it's nice that at least in this case you can work around by just flipping polarity.
:hobbes:

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Sun Jan 19, 2014 10:44 pm

bartlebooth are you back home with that matrix list yet? :hyper:
:hobbes:

User avatar
bartlebooth
Super Deluxe Wiggler
Posts: 1059
Joined: Sat Dec 13, 2008 2:26 pm
Location: DC

Post by bartlebooth » Sun Jan 19, 2014 10:52 pm

VortexRanger wrote:bartlebooth are you back home with that matrix list yet? :hyper:
just got back.. thanks for the reminder though

User avatar
stellvia
Wibbly-wobbly timey-wimey
Posts: 852
Joined: Mon Jun 04, 2012 12:26 am
Location: Perth, Aus

Post by stellvia » Mon Jan 20, 2014 3:01 am

@VrotexRanger: with regards to the togo skipping, that's what it seemed like to me. I tried a version of the above bouillabaisse, but had only 2 or 3 values in the togo list, with 1 set of repeated values, and it definitely seemed like it was skipping to the next 'different' value rather than repeating the same value. Not a huge problem in this case, but might more of a problem in arab mode (haven't tried it yet) if you want to, say, have repeated notes in a sequence. Haven't verified this though.

User avatar
krz
Ultra Wiggler
Posts: 873
Joined: Tue May 18, 2010 11:16 pm
Location: Canada
Contact:

Post by krz » Mon Jan 20, 2014 6:55 am

[quote="VortexRanger"]
Are you talking about using togo to bend? This is something I've been curious about, I would think each situation would start the togo over as it was initiated, thus making it impossible to reach the end of the togo, creating some sort of feedback loop. Thus I've pretty much only bent with total unpredictability (dusty salsa) or values that will stay consistent across situations (bars and corps).[/quote]

I was being a bit of a comedian with the word "sidestep" as a descriptor.

Somehow, in some of the situations I have combined in a bowl, about 30, the previous situation's opcode will continue to play/be played/be heard simultaneously, when I've changed (stepped back/forth) to another in the bowl, and the previous situation will be affected by the next selected situation's modifiers - whereupon both opcodes can then be heard. I would liken it to a mother and a father - situation - 'combining' with the resultant child situation then becoming a mix of opcodes modified with the newly selected bars, corps, buttons (or 'whatever' has been coded/programmed) from that next selected situation. Normally, most times I'd just roll thru the situations - A thru Z. When I get a block of time, hopefully I can narrow it down - things like this irk me. To add, I combined a few of the shbobo forum 'user submitted' situation, together with vancouver, and wanilla. I asked for assistance with a dead situation amongst the bowl, awhile back on the shbobo forum. At the time, I needed sounds for gigs/recording, and didn't have the means (a computer) to test/build. Justints was all I could use with a text editor and iphone, for quite awhile, borrowing a computer (briefly) 3-4 times a month to serve up bouilibasse's. I 'rebuilt' it all last week, and there was one (oddly...very basic) situation that would work fine on it's own, but it wouldn't sound/react at all- in the larger bowl. As to the omniscient view of what processes were taking place with the modifiers and opcodes above? Sorry, I can't answer that. When I find out what the exact steps are - I'll report it.

User avatar
bartlebooth
Super Deluxe Wiggler
Posts: 1059
Joined: Sat Dec 13, 2008 2:26 pm
Location: DC

Post by bartlebooth » Tue Jan 21, 2014 2:09 pm

VortexRanger wrote:bartlebooth are you back home with that matrix list yet? :hyper:

ok here's what i've got..there may be some errors, it was all by ear and i haven't tried to confirm it w/ peter.

also, just to reiterate, this is provided purely for informational purposes for help with matrix hacking and other non-conventional applications, it is not at all necessary for normal shnth use.

wanilla.bin matrix IDs
1 wind
2-3 corp
4-7 bar
8 -15 butts
16-23 horn
24-31 saw
32-39 togo
40-47 toggle
48-55 swoop
56-63 mount
64-71 smoke
72-79 dust
80-83 fog
84-87 swamp
88-91 haze
92-95 ??empty??
96-99 string
100-103 comb
104-111(?) zither
112-119 wave
120-127 water/salt
128-143 unknown
144-151 slew
152-159 wheel
160-167 slave
168-175 pulse
176-183 sauce
184-191 salsa
192-195 weirdo osc
196-215 squarewave osc?
.
.
.

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Tue Jan 21, 2014 10:48 pm

Hey, awesome thank you! I'll be trying these out and posting examples soon!
:hobbes:

User avatar
bartlebooth
Super Deluxe Wiggler
Posts: 1059
Joined: Sat Dec 13, 2008 2:26 pm
Location: DC

Post by bartlebooth » Wed Jan 22, 2014 2:30 am

krz wrote:
VortexRanger wrote: Are you talking about using togo to bend? This is something I've been curious about, I would think each situation would start the togo over as it was initiated, thus making it impossible to reach the end of the togo, creating some sort of feedback loop. Thus I've pretty much only bent with total unpredictability (dusty salsa) or values that will stay consistent across situations (bars and corps).
I was being a bit of a comedian with the word "sidestep" as a descriptor.

Somehow, in some of the situations I have combined in a bowl, about 30, the previous situation's opcode will continue to play/be played/be heard simultaneously, when I've changed (stepped back/forth) to another in the bowl, and the previous situation will be affected by the next selected situation's modifiers - whereupon both opcodes can then be heard. I would liken it to a mother and a father - situation - 'combining' with the resultant child situation then becoming a mix of opcodes modified with the newly selected bars, corps, buttons (or 'whatever' has been coded/programmed) from that next selected situation. Normally, most times I'd just roll thru the situations - A thru Z. When I get a block of time, hopefully I can narrow it down - things like this irk me. To add, I combined a few of the shbobo forum 'user submitted' situation, together with vancouver, and wanilla. I asked for assistance with a dead situation amongst the bowl, awhile back on the shbobo forum. At the time, I needed sounds for gigs/recording, and didn't have the means (a computer) to test/build. Justints was all I could use with a text editor and iphone, for quite awhile, borrowing a computer (briefly) 3-4 times a month to serve up bouilibasse's. I 'rebuilt' it all last week, and there was one (oddly...very basic) situation that would work fine on it's own, but it wouldn't sound/react at all- in the larger bowl. As to the omniscient view of what processes were taking place with the modifiers and opcodes above? Sorry, I can't answer that. When I find out what the exact steps are - I'll report it.
i don't think that's what's happening as the way the code works its not possible to have two situations playing simultaneously. if you'd like to send me the .txt file you're referring too i'll take a look at it

User avatar
krz
Ultra Wiggler
Posts: 873
Joined: Tue May 18, 2010 11:16 pm
Location: Canada
Contact:

Post by krz » Wed Jan 22, 2014 4:13 am

[quote="bartlebooth"]i don’t think that's what's happening as the way the code works its not possible to have two situations playing simultaneously. if you'd like to send me the .txt file you're referring too i'll take a look at it[/quote]

Thanks for the offer. Whatever’s happening…I hadn’t heard before, and since I (had) used the same exact bowl for 6 months, without alteration…it was surprising. That said- it’s probably me, and I think it’s best that I spend some time pairing it down. You don’t want it with the size it is now. =)

User avatar
pannikel
Wiggling with Experience
Posts: 376
Joined: Sun Nov 28, 2010 4:02 pm

Post by pannikel » Sun Aug 21, 2016 9:20 am

Apologies if this is a dumb question– I take it Fish wouldn't allow a kind of switchable patch, i.e to be able to use one of the buttons to flick between two distinct patches - it could only be one at a time only right?

[I saw there was some kind of crazy random matrix modulation possible inside fish but I mean more like clear switching between 2 distinct 'stable' instruments]

So, how fast can you usb-load a new instrument to shnth + start playing it? Could you risk doing it during a performance perhaps, at least between tracks?

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Sun Aug 21, 2016 10:03 am

pannikel wrote:Apologies if this is a dumb question– I take it Fish wouldn't allow a kind of switchable patch, i.e to be able to use one of the buttons to flick between two distinct patches - it could only be one at a time only right?

[I saw there was some kind of crazy random matrix modulation possible inside fish but I mean more like clear switching between 2 distinct 'stable' instruments]

So, how fast can you usb-load a new instrument to shnth + start playing it? Could you risk doing it during a performance perhaps, at least between tracks?
Uploading is nearly instantaneous. But also, multiple situations do not have to be related to each other. In performance I use one button to step through about 20 distinct situations, and I don't need a computer for this.
:hobbes:

User avatar
wednesdayayay
Super Deluxe Wiggler
Posts: 1576
Joined: Wed Nov 16, 2011 8:17 am

Post by wednesdayayay » Sun Aug 21, 2016 1:52 pm

switching back and forth between patches is where the shnth sheens

no need for USB you can switch between a lot of different patches (not sure what the limit is) I think I"ve topped out at 40 or so different patches
sh is the sound of data
bobo is the gesture

PB

User avatar
VortexRanger
reticulating splines
Posts: 2453
Joined: Fri Mar 04, 2011 12:13 pm
Location: North Carolina

Post by VortexRanger » Sun Aug 21, 2016 3:54 pm

So I was saying "situation" when I meant "bouillabaisse". You can have either 128 or 256, I forget which, of the latter, in a single one of the former.

I like to do this: make a bouillabaisse I like, then in the bouillabaisse menu, hit "randuplicate..." with the dot dot dot ellipsis, and it lets you specify how many random copies you want to make. It's helpful to put a jump in the first one and freeze it (double-click to turn it "milky") so that you can jump in the same manner regardless of which one you are in. If you name your bouillabaisse it will give each one the same name but with the letters scrambled, which I love.

Then upload all and you've got nuff fish.
:hobbes:

User avatar
smonkey
Wiggling with Experience
Posts: 258
Joined: Mon Nov 26, 2012 11:39 pm
Location: lost angles (mercury pond)

Re: Fish Questions

Post by smonkey » Thu Feb 20, 2020 8:49 pm

and this whole thread should be higher up and easier for me to find...whoa...boom...just like that.
"Flingo bungie, whingie FLIP!"

jimmie
Veteran Wiggler
Posts: 545
Joined: Thu Jun 03, 2010 2:27 pm
Location: Japan

Re: Fish Questions

Post by jimmie » Sat Feb 22, 2020 12:38 am

Yeah,

By the way, is there Windows binary of Fish2.0 anywhere?

User avatar
smonkey
Wiggling with Experience
Posts: 258
Joined: Mon Nov 26, 2012 11:39 pm
Location: lost angles (mercury pond)

Re: Fish Questions

Post by smonkey » Sat Feb 22, 2020 5:14 pm

its on shbobo.net as shbobozip download
"Flingo bungie, whingie FLIP!"

Post Reply

Return to “Ciat-Lonbarde”