Comment by hurtigioll

people said "it's impossible to separate tracks (voice, bass, ...) after they are mixed". true in theory, but neural-networks can separate them in practice

same here, but there is no real market for somebody to bother yet