Voices question
 

Hi there,

I have questions about the voices. I have a Virus TI 2 Polar (I love it :D), so I have 90 voices "20 and 90 voices can be achieved".

-Can I see how many voices i'm using instantly ?

-If I play one sound, with one key and my patch use two unison, I use 2 voices or 3 ?

-My patch use 4 voices with one key, if I play a chord with 3 keys, I use 12 voices, right ?

Thank you :)
Simon

Simon, it doesn't work like that. The Virus does not have a notion of a 'unit voice', that everything else can be measured in terms of. You may be able to get 90 voices of plain saw waves with no modulation, but for anything musically interesting or useful, the polyphony is going to be less. The 'patch complexity indicator' on the LCD can give you a rough idea of how demanding your patch is, but it doesn't take into account unison or FX. The only useful measure is CPU load, but unfortunately, it is not displayed anywhere.

If you start with an Init patch (simplest = 1 bar on complexity indicator) and fill all Mod matrix slots, you'll see the complexity rise.
That said, it is well known that the biggest CPU-eaters in Virus are:
- advanced oscillator models (grain/formant)
- 'analog' filter mode
- reverb

hope this helps 8-)

For the earlier Virus models (Virus A and B) it used to be almost as simple to calculate as you mentioned Simon, however as Boreg stated the TI and TI2 are different as it appears to be more a global CPU issue. The more complex the oscillator, signal flow and effects, the fewer the poly, and vice versa. You should get stacks of poly with the TI2 in normal use though.

PS> Congrats on the TI2 Polar. :)

Ok, thank you !

So, all parameters (osc, filters, fx, ...) use the cpu and reduce the number of voices (because voices use the cpu too), i'm right ?

Exactly, if by 'number of voices' you mean 'number of notes that can be played simultaneously'.

It looks more like a VSTi way of handling voices except that instead of crackling sounds you get muted voices when reaching the voices limits.

I wish one could have some more predictable alternate mode where you could assign fixed voices to each part: i.e. fixed voices allocation rather than dynamic voices allocation.

Well, there's 'part priority' setting in Multi mode, but I don't know how useful it is.