Polyphonic/multi-voice modules (Edges...)

@BennelongBicyclist It is a broad issue with everyone thinking that certain components belong in certain areas. My original hope was that I could get sweet sonorous voice (like I do from tides) from edges through programming… But I guess not. What I crave is three or four identical osc only in one module so that I could route them through other chains of vca/filter etc. to get an end result that could be seen to be carved from the same stone , so to speak. More multitimbreal than pol I suppose…

@markmanster … what you describe is paraphonic.

Paraphony is when you have multiple oscillators able to do different notes but only one filter or VCA. Eg. Korg Poly 800.

I have an ongoing project that may satisfy your need for polyphony somehow. I agree, it’s an area that’s still missing in Eurorack. What I’m designing is an 8-voice “chord machine” (so, paraphonic), with meta-controls over all the voices at once: you’re not controlling the pitch of each voices atomically, but instead manipulating the “structure” of the chord at a higher abstraction level. The only thing is, it is not designed to put out your regular western scale chords. The first prototype sounds good but… weird. Not sure yet on what module it will be available, and probably not in the near future.

@mqtthiqs have you should design your own hardware to host this new firmware. It’s impressive how much stuff you can crowbar into an existing MI module, but you’re always working round the limitations of a UI designed for a different task. while this does present an interesting challenge, it also doesn’t always make for the most user-friendly module, for the end-user.


A propos of nothing, TipTop Audio did have a polyphonic modular in development at one stage-it used 8-pin mini-din cables for everything. I believe it foundered on the rocks for the same reasons Bennelong outlined before-that you are not realising economies of scale, and for analogue circuitry every panel control would now have to generate a control voltage feeding a VCA in each signal chain instead of being a simple pot in a circuit path, resulting in a very expensive patchcord version of the Matrix-6.

For those curious about what might have been

That TipTop poly VCO is interesting!

I’ve mentioned it before, but worth repeating: the Mungo State Zero provides a very interesting approach to modular, patchable polyphonic synthesis. Alas, it is all proprietary and its (Australian) developer is reputedly rather secretive, but as a paradigm it deserves study. I don’t know of anyone who actually has a State Zero, and they are vastly expensive, but I imagine a few were bought for reverse engineering by some of the larger synth manufacturers such as Roland. If not, they ought to have been. What we need now is an open-source version of the State Zero polyphonic patching paradigm.

Indeed! I assume the inspiration was the little-used multivoice/multipin connector meant to join up the old 12-bit Akai samplers and analogue synths, I guess the reality check was that Gur probably realised that he couldn’t make the P3000 oscillator for less than the price of an entire DSI polysynth…

I’ve heard figures like total sales of around 10 so far for the State Zero. It was demonstrated at Moduluxxx in Sydney in 2013 by the maker, and certainly sounds the goods, thanks to its FPGA architecture-IIRC @pichettes has stated that it generates waveforms with sample rates in the megahertz range…

I think Korg’s Legacy MS-20 controller is also an interesting approach to using patchcords to route a (virtual) polysynth

Ah yes, I’d forgotten about the Korgs. In fact, wasn’t the Korg PS-3100 supposed to have a complete monophonic synth voice for each key on its keyboard? So how did the patching work? Was it all done with a VCA matrix for each voice, and the patch cord just carried control voltages affecting every voice, rather than carrying audio signals? That sounds like a vastly expensive approach.

As mentioned before yarns provides 4 digital oscs in one module - and it provides several waveforms:
" The available waveforms are sawtooth, 25% rectangle, square, triangle and sine."

Yes, and not just any old sawtooth, square wave, triangle and sine waveforms, but nicely bandwidth-limited, high quality waveforms. And did I read somewhere that Yarns code will be getting a makeover and acquiring the new polyBLEP algorithm that has replaced the miniBLEP approach previously used in Braids? If so, the Yarns waveforms will be even nicer. Ideal as FM modulators or as raw material for wave folders.

I’ve played with all of the other yarns functions except the osc. The sequencer and arpeggiator are quite nice. Now I’ll need to explore this aspect. Though this is the first I’ve heard of multiple oscs in the yarns. I’ve only seen it termed as "an oscillator " Amazing module.

In 2M or 2P, you can repurpose CV outputs 3 and 4 as oscillators. In 4M or 4P, you can repurpose all 4 CV outputs as oscillators. So you can actually use Yarns as a MIDI controlled 4-voice, single-waveform oscillator, which is nice for sequencing chords :slight_smile:

eurorack needs more polyphonic options

If Yarns and Braids could be entwined into a module running an STM32F4 processor…show us the Ropes! Or Yarns and Tides.

Looking at the PS3300, it seems there’s a pretty limited set of external modulation possibilities. They used a divide-down oscillator scheme, so VCO modulation could only go to a master oscillator, and a simple buffering system would have distributed modulation to the VCFs…

Ah, OK. I thought that 48 analogue VCOs or 144 VCOs, was a bit OTT. I notice that it supports just intonation and other non-equal temperaments. Nice! I can see why they were held in awe.