Vuk Radovanovic

---

For this project, I implemented an older, less commonly used technique. I used a bank of 80 heterodyne frequency analyzers, paired with 80 sine oscillator voices of corresponding frequencies. This leads to a pretty interesting-sounding approximation of the original signal.

I also included some controllable filtering of those 80 voices, such that only voices of selected semitones would be enabled. That way, you could input a chord, and then only the synths corresponding to that frequency would respond to the input signal - almost like a coarse tuned resynthesis of the original sound.

You can watch a short demo video here:

Or, if you’d like to try it out for yourself, you can check out the source code at my Pd projects GitHub repo.