Software Physical Modeling


Creating a midi sequencer has all these ‘stuck note’ problems which require the sequencer to get pretty involved in counting ringing tones…

Standardising different ‘families’ of note gestures would be very much necessary to avoid all those problems and be able to sequence osc.

My day gig is with computer bus protocols, I have little faith in humanity’s ability to work out sane standards for this sort of thing…


Thank you so much for clarifying! Makes perfect sense to me now. Looking forward to reading the paper.


You have a tough gig!


Just ran across this:


Feel like I might have killed this thread with the derailment into the black hole of OSC sequencing…

However I finally got stuck into my Faust karplus-strong implementation, huzzah! So far I think the ‘tricks’ are:

  1. Bandpassed white noise excitation source - seems to ‘just work’ compared to other schemes I tried
  2. Nonlinearity in feedback path - I really like this one:

half_tanh(X) = 0.5 *( tanh(X) + X);
nonlinearity = +(0.1) : half_tanh : -(0.1);


Applied Acoustics’ Chromaphone and String Studio are great PM apps. They can run as AU/VST and as standalone.