Loving this! Thanks so much Nikolozi! Now that we have separate delay & feedback, it’s possible to do comb-filter physical modeling with additional modules inserted in the feedback loop for even more sophisticated sound sculpting. Here’s a video of something I’ve cooked up using this technique.

Apologies for my far-from-flawless performance here, but it does show off some of the range and unpredicatability of PM with this approach. This is all one Mela patch; the other visible apps are only there for monitoring/visualization. There are no oscillators in this patch! Just one noise generator and one delay set to pitch-scaled timing, feeding back through various filters, saturators and clippers being modulated in multiple dimensions by breath pressure, key tracking, and other expressive controls. The result is a tone that behaves very “acoustically”, shifting dynamically based on the subtlest changes in articulation, with each new note affected by the ones before it still resonating, so it’s nearly impossible to play exactly the same phrase exactly the same way more than once.

I’m definitely going to be digging deeper into this! The possibilities are endless.