It’d be great if Vuo could simplify the process of converting FFT data to MIDI notes. Suggested node would have a monophonic (melodies) and polyphonic (chords) detection modes. There might be some challenges with requiring a high sample buffer size…
Some folks attempting it here in Processing: https://processing.org/discourse/beta/num_1208016366.html
This resource might be helpful for coding methodologies such as: 2015:Audio Melody Extraction - MIREX Wiki