It would be cool if a model could tell you if you are singing or playing a piece of music with the right intonation and other ways.