Is MaxMSP capable of this level of Analysis / Resynthesis?
I came across this video recently and I have been obsessed with achieving similar results:
The video creator used Spaark to analyze speech and resynthesize his voice with 516 sinusoidal oscillators. Has anyone explored analysis and resynthesis to this level in MaxMSP?
I would love to analyze audio files into a buffer and have them play back with such high quality but manipulate them in interesting ways with additive synthesis.
I am not an expert in audio DSP or Max for that matter, but any ideas, patches, or tutorials are welcome. I don't mind studying!
I would suggest studying:
- the traditional vocoder (Examples -> Effects -> Classic Vocoder Folder -> Classic Vocoder).
- the phase vocoder (Examples -> FFT-Fun -> Phase Vocoder Sampler Folder -> Phase Vocoder Sampler).
For reading connecting both ideas, I recommend "The Phase Vocoder: a Tutorial" by Mark Dolson, Computer Music Journal 1986.
Further study includes:
- very well done tutorial for a renewed version of the first patch: https://cycling74.com/tutorials/a-basic-vocoder-tutorial-part-1
- help files for [fffb~] and [oscbank~]
Thank you Jean-Francois Charles! I appreciate the information, I am excited to learn as much about FFT's as I can. :)