Polyphonic Voice Expander?

[DrAKS]

Agreed , thank you for your explanation.

 I think would be a wasteful project which in the analogue arena is dominated by Dave Smiths Rev2 Prophets etc.

This idea is for the deep, deep, deep freeze

[DrJustice]

Creating a voice expander would be a very complicated thing to do. It's not sufficient to double/triple/quadruple the oscillators, you would have to completely redesign the frontpanel. Take the Audio modulation section; now it's vco>vco2 etc. Having multiple osc would mean you have to create very different audio modulation routings.
Same with the Matrixboard; it's designed for routing all the modulators to two oscillators. If you want those modulators to modulate 6-8 oscillators  you would have to have a 64 by 16 matrix. Or if you want to stick to the matrix create oscillator  groups (osc1-4) and (osc5-8) and modulate the groups.
Same with the filters...you would have to create a completely different filter routing system or, similarly as above work with a grouping system (osc1-4) and (osc5-8) and route the groups through the filters.
...

I'm not talking about adding oscillators, or in any way changing the voice. I'm talking about about separate voices/voice-boards to be played polyphonically. There would be zero changes to the voice electronics, no additional matrixes needed. It's actually not complicated; the MatrixBrute voice has been developed. A poly application "only" needs replicating the voice electronics N times (probably on separate board for manufacturability), and making a polyphonic control system (a tiny main board with a  micro controller on it). When you have all the tech, as Arturia do, this is not huge project (we're talking a plain square box box with PCBs and minimal IO).

[DrJustice]

...
 I think would be a wasteful project which in the analogue arena is dominated by Dave Smiths Rev2 Prophets etc.
...

That's exactly the sad thing. I don't want a DSI. I don't want all VCO analogues to be DSIs. They seem to be well engineered, but the DSI VCO based voices are to simplistic for my tastes. It's time for something far more exciting. More in the spirit of the Oberheim Matrix 12.

[Reaktiv]

Brilliant idea, sample on the fly  to get a poly brute. Hope it will implemented!

[DrJustice]

Taking a sample snapshot can never even begin to come close to have live running analogue voices. It would also require development of hardware and software which isn't in the Arturia tech portfolio. Neither is there anything in the MatrixBrute UI to support operating a sampler.

[khidr]

Taking a sample snapshot can never even begin to come close to have live running analogue voices. It would also require development of hardware and software which isn't in the Arturia tech portfolio. Neither is there anything in the MatrixBrute UI to support operating a sampler.

I strongly disagree. A voice sampled on the fly would be undistinguishable from the real thing. Even more so if it's sampled at several points on the keyboard; there may be filters and VCO3 key tracking. You would not need to operate a sampler as a user. The sampling process is completely hidden. The moment you press "assign to voice x", the sampling process starts  and the current sound is assigned to one or more voices.

Also I suspect that Arturia has considerable knowledge in this field having developed so many brilliant software synths. I could mean upgrading and improving the Matrixbrute operating system. That will happen anyhow there's already a capacity problem: try changing sequencers steps (muting) when the sequencer is running. It's nearly impossible. Maybe that's a programming oversight , it could could also mean that the software is running at its limits.

[DrAKS]

I am not telling you to get any synth  Dr.Justice.

I do not mind what you buy. I have no idea  as to what you want or don't want so please stop assuming I am trying to tell you what to buy!

I am just countering your suggestion to develop a PolyBrute.

This would be very bad practice at a time when very few MBs have been delivered.

The delivery fiasco is causing a lot of anguish

Development of a PolyBrute is  ill judged and will lead to Arturia being severely at risk of failure as a company

[DrJustice]

Taking a sample snapshot can never even begin to come close to have live running analogue voices. It would also require development of hardware and software which isn't in the Arturia tech portfolio. Neither is there anything in the MatrixBrute UI to support operating a sampler.

I strongly disagree. A voice sampled on the fly would be undistinguishable from the real thing. Even more so if it's sampled at several points on the keyboard; there may be filters and VCO3 key tracking. You would not need to operate a sampler as a user. The sampling process is completely hidden. The moment you press "assign to voice x", the sampling process starts  and the current sound is assigned to one or more voices.

Also I suspect that Arturia has considerable knowledge in this field having developed so many brilliant software synths. I could mean upgrading and improving the Matrixbrute operating system. That will happen anyhow there's already a capacity problem: try changing sequencers steps (muting) when the sequencer is running. It's nearly impossible. Maybe that's a programming oversight , it could could also mean that the software is running at its limits.

Even if you would be happy with static samples, it remains a fact that there's nothing in the hardware to support sampling. No UI, no CODEC and in all probability no huge amounts of leftover CPU/DSP power.

With regard to the viability of a sampling solution, to not be super cheesy they'd have to be multi sampled across the keyboard, and doing that for 6-8 voices (as you suggest elsewhere) would need a sizeable chunk of hardware: CPU/DSP, CODECs, memory, storage, UI. I.e. an external box would be needed. For this you can use any already existing sampler to much better effect.

[DrJustice]

...
This would be very bad practice at a time when very few MBs have been delivered.

As said, the MatrixBrute has been developed. The R&D investments have been done. Arturia has a staff of engineers which is steadily growing. They are hiring 4 more as we speak. For the engineers to develop new products will have zero impact on the delivery of MatrixBrutes. Please stop assuming that the engineers are on the factory floor assembling MatrixBrutes. In my many years as an embedded systems engineer, I have first hand knowledge of both the technology and R&D and manufacturing processes of electronic products.

Leveraging the technology you have to create and new products is anything but bad practice; it's standard fare in engineering companies. As to whether Arturia would consider it risky or not, we don't know. It's only a point of discussion from my side, not even likely to happen. One thing is certain, their engineers are already working on other products - possibly a PolyBrute, it would be a natural progression especially with analogue polys being in high demand again.

So once more: do not fear, me speculating about a poly expander will not delay your MatrixBrute.

[DrAKS]

Hey DrJustice ,

I don't  care any more!!!

My MatrixBrute is being delivered tomorrow

[F5D]

I would be interested in a 5 voice rack expander for sure, but maybe Arturia will develop a dedicated polyphonic synth next with panel and hardware optimized for a poly. With MatrixBrute, they went to the top of mono synths already. The downside of sampling is that it will never catch the modulations properly. The sampling time could be increased, but all samples would still become the same, if not multi-sampling each key and that would make the feature too slow to use. The best thing of the MatrixBrute is exactly the modulations, because that is what moves the sound. A polyphonic version could be achieved only by duplicating the voices and giving each voice the same set of modulators, but different key inputs. Then, each voice could live its own life with modulations and be truly polyphonic. The DSI VCO synths are too simple, in practice giving you only 1 LFO + 1 OSC for modulation with fixed mod wheel routing, whereas the MatrixBrute gives multiple freely assignable controllers with dozens of parallel modulation routings at once, all with their own individual mod amounts, like 100x more powerful.

I have not even received my synth yet, but I am already thinking about the various mod routings that I will try... I will probably end up with nearly fully lit matrix for modulations, a little bit of this and that too...

[mikehuckaby]

i was thinking this the first day they announced it would be a mono synth.
And yes, i hope this will become an option !!

[ruuud]

here is something really simple that could be realized in the OS alone.

voices 1,2 and 3 go to the matrixbrute like they do now, for instance in paraphonic mode. voices (say) 4,5 and 6 when you hit these notes on the keyboard, are routed to 3 of the cv outs, for instance 3 of the 4 assignable routings in the matrix.
instad of arturia developing an expander, you develop your own in the form of a small eurorack. all you need is oscillators and maybe a small mixer. the oscillators are controlled via the overspilled cv from the matrix, and their audio can come back via the external input where they can get processed through the filter and vca just like the internal vco's. big plus is that as a user you can select your own sounds (a matrixbrute with wavetables) and that all Arturia has to do is create some extra code in the OS.