Hey Fizy, I’ve been working on a neural source‑separation patch that can pull individual tracks out live—got any thoughts on how we could use that for a performance?