Maybe I'm using a wrong sintax or something like that, but all the voices are just 100% flat sounding no matter what i do. Is this not yet implemented or I'm doing it wrong?