Our answer to modular polyphony: Supercritical Synthesizers

Cwejman, Livewire, TipTop Audio, Doepfer etc... Get your euro on!

Moderators: Kent, Joe., luketeaford, lisa

Post Reply
Robrecht
Common Wiggler
Posts: 78
Joined: Sun Mar 04, 2018 6:08 am

Re:

Post by Robrecht » Wed Jun 17, 2020 5:07 am

Robrecht wrote:
Tue Aug 27, 2019 4:38 am
Even though multiple outputs would be ideal, the single output works well for me through any filter (Ripples, Wasp, Moog).

In my experience, the retrigger option is key. If you're holding a chord with your right hand and playing a bass line on the left, you don't want to retrigger the filter envelope (and re-articulate that higher sustained chord) with every bass note.

Conversely, a lead melody over a sustained chord works best if the filter envelope is retriggered with every lead note, and the fact that the chord goes through the same filter opening and closing again doesn't matter that much, as it's in a lower register.

By setting the retrigger behavior appropriately, I don't really miss having a separate filter envelope per voice, thanks in part to the individual VCA envelopes. Also, sending that mono output through a good stereo reverb makes it super lush and wide.
I still stand by this statement very much –– a bit of chorus by modulating a short stereo delay plus some stereo reverb makes the DCO sound as lush and wide as I could ever want –– but I was thinking about the following concept.

Since the sixteen oscillators are derived from a single analog core, they are produced as a single signal and can't be separated, right? But what if you were to add a second, identical oscillator core?

I know you can control the amplitude of each of the sixteen oscillators independently, since they all have their own ADSR envelope. So in theory, you could coordinate two oscillator cores so that they both play exactly the same notes, but core A grades the amplitude of its 16 oscillators from 100 to 0%, and core B does the same but from 0 to 100%. If the mono outputs of both cores are hard panned left and right, that should give the impression of 16 oscillators spread evenly across the stereo spectrum, no?

Like, oscillator 1 has an amplitude of 100% in the left channel and 0% in the right channel; oscillator 2 is at 93.4% left and 6.6% right; oscillator 3 at 13.3% and 86.6%, and so on until oscillator 16, which is at 0% left and 100% right.

OK, that's the theory. Now I'm imagining a small expander module -- let's say 4HP -- with the following components:
  • One oscillator core just like the one in the main module
  • A knob and corresponding CV input for stereo spread (from mono/centered to super wide)
  • Optionally, a bipolar knob and CV input for stereo movement, making all the oscillators continually shift to the left or to the right at a speed determined by the knob/CV
  • A single audio output, positioned so it would line up next to the main module's output (when the stereo expander is installed to the left of it), to form a stereo pair to be panned left and right
  • Two ribbon cable connectors on the back, so you can connect both the main module and the existing expander simultaneously
Since you need a pair of identical oscillators from core A and core B to give the impression of one panned oscillator, the net number of available oscillators would still be 16, so the 16 LED "star" interface of the main module could continue working in the same intuitive way.

Of course, this would require a big firmware update and there are probably lots of technical hurdles I haven't thought about. But it would be fun, wouldn't it? :mrgreen:

tisaalho
Learning to Wiggle
Posts: 25
Joined: Thu Feb 07, 2019 5:01 am
Location: Finland

Re: Re:

Post by tisaalho » Thu Jun 25, 2020 4:17 am

Robrecht wrote:
Wed Jun 17, 2020 5:07 am


OK, that's the theory. Now I'm imagining a small expander module -- let's say 4HP -- with the following components:
  • One oscillator core just like the one in the main module
  • A knob and corresponding CV input for stereo spread (from mono/centered to super wide)
  • Optionally, a bipolar knob and CV input for stereo movement, making all the oscillators continually shift to the left or to the right at a speed determined by the knob/CV
  • A single audio output, positioned so it would line up next to the main module's output (when the stereo expander is installed to the left of it), to form a stereo pair to be panned left and right
  • Two ribbon cable connectors on the back, so you can connect both the main module and the existing expander simultaneously
Since you need a pair of identical oscillators from core A and core B to give the impression of one panned oscillator, the net number of available oscillators would still be 16, so the 16 LED "star" interface of the main module could continue working in the same intuitive way.

Of course, this would require a big firmware update and there are probably lots of technical hurdles I haven't thought about. But it would be fun, wouldn't it? :mrgreen:
Great minds! This is something we've been thinking about for some time. However, we'd need to also generate the digital control signals for the output expander (there's no way to share them from the main unit), so it's not enough to duplicate the analog part, but we'd have to also throw in another microcontroller. It could be a smaller/cheaper unit though, since it wouldn't need to deal with the UI, tuning equations etc. Then there's the problem that there's really no digital bus suitable for the units to communicate in the expander cable... except if we press into service an existing bus which is used for factory testing and toggling the expander LEDs. Then comes the question whether this relatively slow bus could push all the necessary control data from the main unit in time. At the back-of-the-envelope -level, the answer appears to be maybe just barely. Which means that it would indeed be a very non-trivial effort to write the firmware, with a small but non-negligible chance that it won't work after all.

How much do you think such an expander should cost, when compared with the existing one? How about if, at the same time, it would actually produce 3 extra outs (since the analog section wouldn't be that expensive to triplicate once the communication would work) for total 4 voice true polyphony? :hmm:
Digital brain, analog heart

loydb
Super Deluxe Wiggler
Posts: 1203
Joined: Sat Feb 13, 2010 6:45 pm
Location: Austin, USA

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by loydb » Thu Jun 25, 2020 8:19 am

Whatever it cost, I'd buy one... but I would think $200-$250 would probably be the pain threshold for many people. The problem seems to be that the audience would be mostly limited to people who already own one, but I guess you already have experience with that.

Robrecht
Common Wiggler
Posts: 78
Joined: Sun Mar 04, 2018 6:08 am

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by Robrecht » Thu Jun 25, 2020 2:25 pm

tisaalho wrote:
Thu Jun 25, 2020 4:17 am
Great minds! This is something we've been thinking about for some time.
Thanks for your extensive reply! Really cool that you guys were thinking along the same lines. And of course, as I expected, it is a lot more complicated than I imagined. :)
tisaalho wrote:
Thu Jun 25, 2020 4:17 am
How much do you think such an expander should cost, when compared with the existing one? How about if, at the same time, it would actually produce 3 extra outs (since the analog section wouldn't be that expensive to triplicate once the communication would work) for total 4 voice true polyphony? :hmm:
My initial thinking of adding just one extra output was that the panning of all the voices could then take place inside the module (using the complementary amplitudes thing between the cores), including some CV-able modulation of stereo spread and placement, so it would deliver a ready-to-use stereo signal with all the voices already panned from left to right. That would simplify its use further along in the chain: just one stereo filter needed etc. (something like, I don't know, the upcoming Neutron Flux filter? ;) ). Also, fewer outputs could possibly help keep the module small.

But now, I can imagine that the amplitude coordination between two cores required for such granular stereo panning would add another layer of difficulty for the firmware, especially given the limited digital bandwidth of the expander cable. Would separate outputs for four (unmixed) voices actually be more feasible?

Having four separate outputs would of course be amazing in itself. And more in the spirit of modular, I guess: panning and mixing would then occur downstream, in dedicated modules, and each voice could go through its own filter for true polyphony (although when using MIDI, that would require additional gate outs as well, I guess?).

So: convenience or flexibility? Hard to say what most users would prefer. In my case, and given what I have available in my rack, I think I would slightly lean towards the convenience of a ready-to-use stereo output. But I'd love it either way.

Edit: I've been reading it as an either-or situation, but if you meant having both the stereo option and the separate outs per voice selectable in the same module, that would be amazing! In my mind, these options would roughly correspond to the two existing options for playing polyphonically: simple stereo if you're playing over MIDI, four separate outs if you're going full modular and using the four CV ins (in that case you'd likely have gates to go with those CVs already, from the same external source, for triggering separate filter envelopes).

Price is a difficult question, too. Something in the neighborhood of the existing expander's price, I guess, for most people? Or a little more in my case. Which may still not be proportional to the amount of work it would require, if it is at all feasible.

I hope more people will pitch in to give an idea of the market!

yaongi~
Learning to Wiggle
Posts: 23
Joined: Thu Feb 07, 2019 7:15 pm

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by yaongi~ » Mon Jul 13, 2020 5:05 am

tisaalho wrote:
Thu Jun 25, 2020 4:17 am
How much do you think such an expander should cost, when compared with the existing one? How about if, at the same time, it would actually produce 3 extra outs (since the analog section wouldn't be that expensive to triplicate once the communication would work) for total 4 voice true polyphony? :hmm:
The opportunity to have true polyphony really makes a difference, I think - at least when evaluating investing in the module (as I am). Having the choice between a mono out, stereo spread, or 4 voice polyphony - sounds pretty cool to me.

Can't say what the cost should be, but even having that as an option makes the basic setup more desirable, since it gives the knowledge that it can be expanded, if need be.

dip_registered
Common Wiggler
Posts: 75
Joined: Sat Feb 01, 2020 3:18 pm

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by dip_registered » Tue Aug 11, 2020 10:39 am

Has anyone found a good MIDI type B (from keystep) to TRS 1/8 cable to use with the expander? I’m having no luck finding one!

tisaalho
Learning to Wiggle
Posts: 25
Joined: Thu Feb 07, 2019 5:01 am
Location: Finland

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by tisaalho » Wed Aug 19, 2020 2:27 am

dip_registered wrote:
Tue Aug 11, 2020 10:39 am
Has anyone found a good MIDI type B (from keystep) to TRS 1/8 cable to use with the expander? I’m having no luck finding one!
Quick note: the expander input is bipolar, i.e. type A and B will both work.
Digital brain, analog heart

dip_registered
Common Wiggler
Posts: 75
Joined: Sat Feb 01, 2020 3:18 pm

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by dip_registered » Tue Sep 22, 2020 11:47 pm

Finally got my MIDI cable working! Excited but can't figure out how to get polyphony? With the midi plugged into TRS and playing fine, except no polyphony. Am I missing something?

Edit: Quickly got it. Keystep was in chord mode :foul:

hzzzu
Learning to Wiggle
Posts: 41
Joined: Fri Feb 05, 2016 8:09 am

Re: Our answer to modular polyphony: Supercritical Synthesizers

Post by hzzzu » Wed Sep 23, 2020 2:59 am

dip_registered wrote:
Tue Sep 22, 2020 11:47 pm
Finally got my MIDI cable working! Excited but can't figure out how to get polyphony? With the midi plugged into TRS and playing fine, except no polyphony. Am I missing something?

Edit: Quickly got it. Keystep was in chord mode :foul:
Cool! Didn't know the Keystep had a chord mode though I have one. Should I do a basic MIDI polyphony video of this. Do you feel it's intuitive enough to get the polyphony going on the DCO?
Supercritical Synthesizers - Email us for all inclusive time and space travel packages. Go now before Tom Cruise does!

Post Reply

Return to “Eurorack Modules”