Decomposing sounds from (fully produced?) tracks into underlying components, and then giving the user the option to synthesize them with different parameter settings. I think.

▲

pennomi a day ago | parent [-]

I was trying (and failing) to do this the other day. It’s a really interesting problem space and I love to see someone with a more solid foundation give it a try.

	▲	tgv 10 hours ago \| parent [-]
		There is better software, although it might not meet OP's goal. There's e.g. Melodyine, which can identify (and change) notes from individual instruments in polyphonic passages, and there are also tools to estimate and remove reverb and other forms of processing. Those are based on classical DSP. OP just wanted to use "AI".