This is how we get rich
Main problem: Take in a sound file and in real time:
- Split it into short intervals
- Compress signal (make all notes equally loud)
- Do clustering (this way we handle legato phrases as easily as staccato)
- For each cluster, find earliest and latest observation
- Split signal
- Apply FFT
- Machine Learning is pathway to many abilities, some consider to be... unnatural.
- ????
- PROFIT