PlanetZ forums - scopeusers.com

I'm working on it, look at this demo video:

http://youtu.be/Ds6eU1k9KUc

12 instances of solaris, each one on it's own dsp from dsp 7 to 18, running a patch that reaches around 50% per dsp.

Polyphony managed by midi(working on a device to do that).

Thank you

Sal

Yes! Thanks for getting a jump start on this way to work with polyphony!!!

And for the great work you've been doing.

Awesome can't wait! Thanks for working on this.

Great that you are working on it.......

It would be a dream ...For all of the zarg synthesizers!!!

djmicron wrote:I'm working on it, look at this demo video:

Awesome !!!

djmicron wrote: 12 instances of solaris, each one on it's own dsp from dsp 7 to 18, running a patch that reaches around 50% per dsp.

That looks impressing.
I have no idea how you manage to tweak the devices that way, but please allow one question:

Would it be possible, in theory, to run more voices on 1 chip and using less instances, so we´d have some more chips free for other devices in addition, or is it not advisable because there might be patches using more DSP tham 50% per chip and voice ?
The background is, the hardware Solaris is an expensive boutique machine and it´s a bit different than the software Solaris.
Now when Solaris runs w/ 10 voices (or more) on XITE-1 and we´d have the option to run a bit more stuff in addition (maybe modules in Modular ?), it would be ice on the cake,- now since hardware Solaris costs as much as XITE-1 costs.

djmicron wrote: Polyphony managed by midi(working on a device to do that).

You do unbelievable work,- thx a lot !

Bud

In my experiments, it does seem like at least 1 solaris will fit per dsp. Certainly some uses will allow for more than 1 per dsp, but since there are so many modules for oscillators and filters that can be loaded, each use may vary.

Placing a mixer for all the voices to minimize phase/latency is another task. A dynamic mixer could be cool, that changes with a max-voice setting via a script.

Would it be possible, in theory, to run more voices on 1 chip and using less instances, so we´d have some more chips free for other devices in addition, or is it not advisable because there might be patches using more DSP than 50% per chip and voice ?

It depends on the patch, but 2 voices per dsp is possible if we take care.

djmicron wrote:
It depends on the patch, but 2 voices per dsp is possible if we take care.

That would be cool !
Hardware Solaris works similar,- 5 chips do the 10 voices (2 per chip fixed for the time being) and the 6th chip does the rest (FX & routing etc.).

good luck and all the best

Bud

I like the way Sal shares this with us and then says thank you........

Good Work.

Djmicron, how do you do this ?
(sorry if I missed something...)

Do you have some kind of sdk ?

I already tried your dsp optimized files, really helpful with the Xite dsp gas factory... Are we going to take advantage of the Solaris ones ?

JoPo wrote:Djmicron, how do you do this ?
(sorry if I missed something...)

Do you have some kind of sdk ?

I already tried your dsp optimized files, really helpful with the Xite dsp gas factory... Are we going to take advantage of the Solaris ones ?

this is done with the sdk 5, i'm developing a midi voice allocation module to manage multiple solaris(and every other synth) as one poly synth.