Convert any stereo track back to it's original multi-track format by introducing machine learning and artificial intelligence by analyzing the stereo track and comparing the sounds and vocals to sounds and vocals in a database which contains a vast library of various sounds, instruments and vocals. Or let the user manually enter the types of instruments, sounds and vocals that are used on the track. Once the comparative analysis is done, it will convert the stereo track into it's original multi-track. The user can then remix, add or delete tracks from the multi-track and create something new.