It’s now possible to do two things:
1] Take and audio file and get MIDI notation in Cubase
2] Take an audio track of a whole band and dissolve it in to parts - very effectively in most cases - example Vocal Remover & Instrumental AI Splitter | LALAL.AI - at a cost of money and time
Yes, I know we have spectral layers, again at a cost, but what I am imagining is:
An in house ability to parse an audio track into parts and provide both MIDI and Audio parts for the speerate band members, with one click on an audio file - this would be a huge game changer and selling point for Cubase 13.
Considering that you describe pretty much what Spectralayers does, and, that I hope Steinberg continues to make money so they can continue to grow and bring into their fold amazing products like Spectralayers, while keeping the developers employed, as they had done with Wavelab, Dorico (there they actually started from scratch, but with devs who had already proven their talent and skill) and others, I’ll speak frankly.
I wouldn’t expect it to be implemented, nor would I support such a request.
I speak for myself of course.
Why not? There was a time when only Melodyne supported wave to MIDI, but now we pretty much expect it in a top end DAW. Progress it’s called. What I describe could be a key feature used a lot if it were integrated into the Cubase biosphere and easy to use.
Those separation algos constantly improving, and also they are mostly open source so it’s matter of time when all daws will support audio separation. It’s just who will hit the trigger first
I think Cubase needs more frequent updates to keep up with latest models for audio separation