There seems to be a lot of AI audio tools; anyone try them for audio isolation? My usecase is separating a song from a 5.1 movie mix, there's a great remix of Hot Blooded I wanna preserve
Are you referring to unmaking audio?
I use Steinberg's SpectrraLayers Pro as I am a Cubase user and whatever the last installation of it becomes resident as an extension but it does run standalone. I use it for things like audio repair and manipulation.
I think the other 'big name' in the field would be the RX10
None of them are perfect and it can be quite tricky to isolate to a forensic depth but I also know that SpectraLayers has better tool customisation and thresholds and also better layer management.
I suppose, like most audio things, people will tell you that the one they use os best so I wouldn't;t just take my word for it.
demucs is pretty good. It can be run locally, for free. I haven't tested it with 5.1 channel audio, but by default it splits the stems into bass, drums, vocals and other.