Not my understanding of how it works. AFAIK, it uses intelligence to get to know certain sounds - a specific voice, a specific instrument, on a recording. It can then use digital technology to isolate out just that part of the recording and turn it into a separate track. It doesn't "create" the...