Ok now this suggestion could quite possibly make Me look like the stupidest person who has ever lived but I think it's a pretty cool idea anyway.
I was reading about the Nvidia Cuda processing not being very good for audio because it is serial in nature where as to be good for audio it would have to be more inclined towards parallel processing.
I was thinking about how cuda's potential could be tapped to use for Nebula given that it's such a CPU hog.
The thought occurred to Me that since programs are being written for rendering graphics utilizing Cuda couldn't the same thing be done for audio?
Couldn't a program be written that utilizes cuda so that tracks could be rendered while other tasks like mixing could be carried out whist Cuda was busy rendering and thus speeding up work flow?
Further more, wouldn't it be possible to have a the latency that is present in most DAW's at mix down work in our favour so the track could be streamed back to the host as it is being rendered?
OK, so the finished product would be something like this.
You are mixing using Nebula as per normal in Your daw of choice and You decide that you would like to keep the processing that You have applied as perminent and also to ease the burden on You CPU. You hit a button and Viola, Cuda jumps to attention and starts rendering Your chosen track while You keep on working.
Dumb idea? Maybe.
At the very least it could render Your tracks while You keep on working. At best it could be like one of those tape devices that someone invented to impart the sound of tape to digital audio workstations. The audio is recorded onto tape and then into the daw all with only a few ms latency.
What do You think?
It seems a shame to write cuda off just because it is in it's infancy.