Instruo arbhar granular processor

[scragz]

Usually input voltages are enforced with hardware.

[Mr.Kus]

Hardware allows -5 to 5 V, I would assume how that voltage translates to buffer position (and other software parameters) is implemented in software, hence my suggestion, that maybe it could be made user configurable to ignore negative voltage and let 0-5 V to control the whole range.

[Sync]

But leaving the scan knob at 0 and using either an ascending or descending 0-5 ranged envelope would do what you suggest as well.

I guess I must use my Arbhar in a significantly different fashion than you do. One of the things I like to do is find a portion of a buffer to explore (manually, by setting scan), and then use CV to diverge from it within a specific range; unipolar modulation doesn’t lend itself to that. Similarly, the intensity knob really works best with bipolar input, allowing you to have the module “silent” at either extreme, and then moving towards higher activity on either the synchronous or asynchronous side. As I have very little interest in synchronous grain seeding, having the input respond to 0-5v would be a huge pain in the butt.

I’ve never encountered a module that changes its input behavior based on knob position; that sounds like a usability nightmare, especially mid-patch. Besides, I’m not sure whether the input ranges are determined by the hardware or software; my hunch is on on the former, but if it’s the latter I really hope Jason would make any change optional if they came to pass.

I don't see how what you're wanting to do is any problem with a -5 to 5V input but where a 5V swing was enough to cover the entire buffer, other than you'd need to attenuate your input by 1/2. With a 1/2 attenuation of the signal you're feeding it now, it'd behave exactly identical to how it behaves now. So in fact, I'm not saying to make it unipolar only, but in order to accomodate unipolar CVs, it be changed so that a 5V swing is enough to cover the entire length of the buffer.

There's no need to have it "ignore negative voltages." Just have it so that a 5V swing is enough to cover the entire buffer length in either direction-- so you could set the scan knob to the end of the buffer, give it a 0 to -5V signal and it'll scan the entire buffer backwards from the end, or put the scan in the minimum position and a 0-5V signal would scan the length of the entire buffer from the beginning. And if you set the scan knob to the middle position, you'd give it a -2.5 to 2.5 signal to scan from the beginning to end if that's what you wanted to do (any of these examples would need to be attenuated where you don't want to scan the entire buffer).

But whatever, I'd be fine with an optional setting for this, if that'd be possible-- that way to each his own. But right now it's a pain to use at all with 5V unipolar CVs to scan more than 1/2 the buffer, without using amplification which for CVs, isn't that common. The tradeoff is having to attenuate a CV if it has a 10V swing, but attenuation is a lot more common and a lot of modules that generate modulation have attenuators on the output already (Zadar, Wobbler, etc.)

[scragz]

All right you won me over. Somebody write to Instruo with the feature request.

[Mr.Kus]

All right you won me over. Somebody write to Instruo with the feature request.

I wrote them yesterday and got very thorough answer today:

Thanks for reaching out and making me aware of the Scan CV discussion on the forum.

I’ve looked through it, as well as your concise email, and I thought I'd you file you in about the reasons for the current implementation, hardware considerations and possible user customisation in a future firmware.

Essentially, most of the CV inputs on the arbhar sum with the parameter control knob, so the range of incoming CV is always in relation to the knob setting. The Scan CV is read by its own independent ADC, hence it was possible to allow different parameter control when the arbhar is in follow mode. There were several reasons we decided on a CV range of -5 to 5V for scan:

Bipolar CV is a fundamental aspect of synthesis and signal flow. In audio contexts signals are naturally bipolar, when brought down to LFO territories this typically remains true.
In a vast majority of CV implementations in classic analogue circuits, the CV signal will sum with the pot position giving a naturally bipolar control response over the parameter. This allows for bidirectional control over parameters through modulation.
Curation of control voltage is the foundation premise of synthesis and 10V peak-to-peak is fairly standard for the vast majority of signals.

Considering that one can scan through 10 seconds of audio, a CV range of -5 to 5V gives optimum resolution for precise playback control while working with a sufficient range to span over the entire buffer.
The CV range of -5 to 5V is implemented in the hardware, therefore -10 to 10V or 0 to 10V ranges will not be possible.

A future firmware feature could offer an optional config file parameter to half the control resolution of the Scan CV, enabling to use a 5Vpp to span over the entire buffer, whether that is from 0 to 5V when the Scan knob is set fully CCW, -2.5 to 2.5 when the Scan Knob is centred or from -5 to 0V when the Scan knob is fully CW.

This feature request is now on the list, but at this time there is no real estimate on when it might be implemented. The current todo and want-todo list for new firmware versions I am working on is long, some very exciting new things are on the way.

All the best,

Sebastian

[corbetta]

Sebastian is great, as usual. Looking forward to what’s in the works... that last paragraph was a bit of a tease!

[closedLoop]

Considering that one can scan through 10 seconds of audio, a CV range of -5 to 5V gives optimum resolution for precise playback control while working with a sufficient range to span over the entire buffer.
The CV range of -5 to 5V is implemented in the hardware, therefore -10 to 10V or 0 to 10V ranges will not be possible.

His point about resolution seems a really important one, especially because of the analog to digital conversion happening. Because there's a fixed hardware restriction on the voltage range, I really don't see why 0-5v would be useful if it means losing the ability to address as many possible points within the buffer, unless the ADC is so fine grained that it wouldn't mean be a perceptible difference.

[Sync]

All right you won me over. Somebody write to Instruo with the feature request.

I wrote them yesterday and got very thorough answer today:

Thanks for reaching out and making me aware of the Scan CV discussion on the forum.

I’ve looked through it, as well as your concise email, and I thought I'd you file you in about the reasons for the current implementation, hardware considerations and possible user customisation in a future firmware.

Essentially, most of the CV inputs on the arbhar sum with the parameter control knob, so the range of incoming CV is always in relation to the knob setting. The Scan CV is read by its own independent ADC, hence it was possible to allow different parameter control when the arbhar is in follow mode. There were several reasons we decided on a CV range of -5 to 5V for scan:

Bipolar CV is a fundamental aspect of synthesis and signal flow. In audio contexts signals are naturally bipolar, when brought down to LFO territories this typically remains true.
In a vast majority of CV implementations in classic analogue circuits, the CV signal will sum with the pot position giving a naturally bipolar control response over the parameter. This allows for bidirectional control over parameters through modulation.
Curation of control voltage is the foundation premise of synthesis and 10V peak-to-peak is fairly standard for the vast majority of signals.

Considering that one can scan through 10 seconds of audio, a CV range of -5 to 5V gives optimum resolution for precise playback control while working with a sufficient range to span over the entire buffer.
The CV range of -5 to 5V is implemented in the hardware, therefore -10 to 10V or 0 to 10V ranges will not be possible.

A future firmware feature could offer an optional config file parameter to half the control resolution of the Scan CV, enabling to use a 5Vpp to span over the entire buffer, whether that is from 0 to 5V when the Scan knob is set fully CCW, -2.5 to 2.5 when the Scan Knob is centred or from -5 to 0V when the Scan knob is fully CW.

This feature request is now on the list, but at this time there is no real estimate on when it might be implemented. The current todo and want-todo list for new firmware versions I am working on is long, some very exciting new things are on the way.

All the best,

Sebastian

I'm good with this as a solution, seems like with this it would satisfy everyone (everyone here, anyway)?

I would think the ADC resolution may be sufficient, and in any case, I can't say I'd be looking to try to use a 0-5V input to select an exact byte position in the buffer. At this point we don't really know if a 10V swing -5 to 5 can address every bite in the buffer given the ADC resolution. No doubt "more accurate" but accurate down to an individual sample value? 10 seconds at a decent sample rate is a boatload of samples, likely exceeding a 16 bit ADC (sample rate 64000 = 10s = 640,000 samples). So yeah, dividing the CV halves the resolution but it may not be sample accurate anyway and the difference negligible enough.

In any case, it's a useful option even if it's less resolution-- if you don't like it, it's configurable so you don't have to use it.

[Mr.Kus]

I'm good with this as a solution, seems like with this it would satisfy everyone (everyone here, anyway)?

Yep.

[corbetta]

I have no problem with that either—and in the midst of the discussion I misconstrued Sync’s argument, confusing “5v for full swing” with “unipolar cv inputs”. Looking forward to this future firmware whenever it’s ready

[bobbyturban]

Eurorack-Noob here:
In wavetable mode, is it possible to send directly into a low pass gate (e.g. intellijel 1u lpg) inorder to create a Buchla-Bongo-like sound?
Or do I need something to control the gate length or an ADSR-envelope?

[jones]

Is there a particular reason the Scan CV input range is limited to -5/+5V? Since the scan knob offsets the CV value, it would be nice if the input accepted -10/+10V (but still required only 10V peak to peak for the full range). This way, it would not be necessary to offset unipolar CV before feeding it to Arbhar to control the full range of the parameter. Can this be changed somehow?

I have a small system and most of my modulation sources, including Pamela’s New Workout and 0-CTRL output 0-5V. I have to amplify these x2 and offset them to be bipolar to control the full range of the Scan parameter. It would be nice if the offset could be done within Arbhar since the functionality is already there. I imagine this is the case for others as well.

i totally agree on this, i would like to sweep the whole range of scan cv with pam and i can only use the half of it. Any ideas ?

Me too - came here to ask about this - seems like an oversight to me - I want to program scan position 'jumps' with my z8000 sequencer and also keyframes on my Mutable Frames both of these can push the position only halfway without offsets. Looking forward to the update.

[afaafa]

I think this just comes down to different eurorack manufacturers having their own voltage standards. Everything Instruo does points to bipolar control being their standard. They use bipolar v/oct. Their LFOs are bipolar. Harmonaig is bipolar, albeit with a really wide range.

I think bipolar cv is the way of the future for everything but envelopes. It makes summing sequences and transposition so much easier. Bipolar LFOs can also be mixed more easily to make complex shapes and they work better for FM and panning. Also I agree with Sebastian's comment about confining the voltage range for optimal resolution. When using Arbhar with a sequencer, I wouldn't want any less resolution.

Navigating different voltage standards is part of planning a rack. If you want to use Pam's and Arbhar, set aside 3hp for an Erica Pico Mscale, or some similar utility.

[antonriehl]

Eurorack-Noob here:
In wavetable mode, is it possible to send directly into a low pass gate (e.g. intellijel 1u lpg) inorder to create a Buchla-Bongo-like sound?
Or do I need something to control the gate length or an ADSR-envelope?

The wavetable mode won't change much in terms of how you could use this with a LPG (it's just a sound source), but the Intellijel 1U LPG has no shaping controls, so you'll need some kind of envelope to to get any kind of delay shape.

The big change in wavetable mode is that the trigger output is disabled, so you wouldn't be able to trigger any envelopes with the Arbhar.

Just testing it now, but in wavetable mode, a manual trigger of the Arbhar creates a pretty short burst on its own, and the shape creates some filter-like (or maybe PWM) changes to the sound.

[axm311]

Does anyone else notices noise from the leds of the Arbhar? I’ve been trying to minimize it and it’s definitely an eminently useable module, but I can hear some whining/buzzing with say intensity all the way down and the level controls pushed up. Just curious if it was “dead silent” for anyone in that config so I don’t go chasing noise gremlins...

input gain could have a factor, it does have the ability to soft clip nicely so if you've captured some noise floor that could be your issue?

nah, it exhibits the same behavior with empty buffers as well. I think it's a ground/common issue, since it's more apparent when connecting to external speakers or other gear. Iso transformer on the audio path helps a fair bit too.

My Arbhar is fairly noisy when it is fully dry (even with input gain fully CCW), but becomes silent at fully wet. However, I’m actually just about to return it partially because of this noise, but mostly because the in-built microphone is extremely noisy - unusable. I spent a week trying to figure out what could be causing the noise without success. I see there is a LED directly beside the microphone - maybe that is the source of my noise...

I just bought an Arbhar and am having a similar issue. It’s not really a noise floor hiss type of sound, but CPU/processor noise, I get a similar noise in my monitors from my audio interface through the computer USB which I’ve kinda given up on fixing. However, the computer noise doesn’t actually show up in recording while this does. It seems to be greatly reduced at full wet but is very noticeable full dry more than any other module and I have a few other digital modules, some with notorious “noise reputations”.

I was going to email Instruo with a recording but was curious if anyone is not getting this noise and also using in an Intellijel performance case, to help determine whether it is a PSU/module compatibility issue?

Here is what I tried to isolate the issue:

Installed in full Intellijel 7U performance case

Installed in half full Intellijel 4U 104 HP Pallette powered by outlet outside of studio (to fill out interference with other equipment)

Installed by itself in Intellijel 4U 104 HP Pallette powered by outlet outside of studio

Installed by itself in Intellijel 4U 104 HP Pallette powered by outlet outside of studio, with PCB USB stick removed (shot in the dark try since I associate this whine with USB audio interference)

Installed latest firmware and tried all of the above again.

The noise was about the same in all of these scenarios which leads me to believe it has to be either the module or it doesn’t jive with intellijel power?

[corbetta]

in my multiple attempts to get rid of the noise you describe, the two things that significantly helped are galvanic isolation and gain staging. Make sure your recorded signal is strong, and that your output level from the Arbhar is high (the noise doesn't seem to be influenced by the "output level" knob, so keeping the knob low gives you a worse s/r ratio.
I recently got a Joranalogue Transmit 2 and in terms of recording it definitely allows me to get healthy signals from the Arbhar without the interference noise becoming a problem. If I crank the Transmit 2 level or the interface level the noise becomes audible but the generated grains clip, so it's a moot point.


Things I've tried: different power supplies (not linear though); filtering doohickeys like the Trogrotronic silencer; ground lifts/transformers (help a fair bit, but the noise is still there)

[buyingitwontmakeucool]

I get similar CPU/processor type noise through the audio outputs. Pittsburgh Structure 360 case. Haven’t had noise with any other modules in this case. Tried moving it around the same case, but that was the extent of my tests. So not sure if noise is from Arbhar or some sensitivity of Arbhar to something else.

[corbetta]

The only instance in which my Arbhar is quiet is when it’s the ONLY module plugged into my tiptop studio bus. I think it’s probably a byproduct of the module design (especially all the LED feedback) in conjunction with subpar common/ground distribution.

[axm311]

Thank you both. I was quite surprised when I still had the noise in a palette case by itself. Corbetta, when you say ground lifts/transformers, do you mean a product like this?

https://www.sweetwater.com/store/detail ... terminator

[apfEID]

I had terrible noise problems with Arbhar in an Intellijel 7U case, but moving it from the bottom row directly over the power supply to the top row with nothing underneath it solved it entirely.

[corbetta]

No, I mean audio ones like the Lehle P-Split

[afaafa]

For what it's worth, I'm using mine in a small case with a variety of power hungry modules and a Konstand Lab BoardPWR and I only hear hiss when I turn it up, not processor noise. Layered granular playback will multiply any extraneous noise, so that can make the mic sound noisy if you have buzzy transformers in your case, if any other modules produce mechanical noise, or if there is any amount of background noise in the room. After all, it is an omni mic inside your case.

[buyingitwontmakeucool]

Not sure about others, but I get the same noise even if input knob full CCW so doesn’t seem like it’s related to the mic?

[corbetta]

I never use the mic and the noise I’m talking about is in the output, not the input.

[virtualpt]

I know this sounds weird, but I had noise from my Arhar when I first got it; I turned the internal switch to reverb & the noise went away. I then turned the reverb back off to see if I could reproduce it & the noise was gone with both settings. I don't know if maybe the switch wasn't fully switched one way or another or whether a cable wasn't seated properly & I knocked it when I put it back. I have no scientific explanation for this, but it might be worth trying it!