2 ms·

This sounds very promising. It'd be nice if audio verification was added and that the lip movement matches the audio. Maybe for v3.1?