Quoteth my friend Smash from Fairlight:
Would this be feasible?
R
Which synths do need such a feature, or for what exactly does he need this?
maybe smash means a way to use multi-timbral synths?
Renoise supports Aliassing for quite a while now, so currently a multitimbral VSTI synth is supported by adding an alias and assigning a different midichannel to the alias. Just the separate audiostreams are not individually controllable and affectable by different chains of effects.
Since I hardly use (multitimbral) vsti, I’ve never tested this, but I wonder if adding an alias does increase the cpu like it would when adding a new instance of the vsti?
Well, logically thinking you could consider what would happen if you would ask a CPU to calculate a certain algorithm twice. For the routines that have to calculate graphical items, that CPU usage would be saved. Any algorithm producing audio being used twice (in original and alias) would probably consume twice the CPU resource.
Considering the amount of CPU graphics routine would consume, i think the resource savings would probably be near to nothing when using an alias.
The Aliases itself do not create any noticeable overhead, but of course they indirectly do, as soon as you trigger a new sound layer with them on a VSTi.