calculates rhythm complexity from the entropy of the sub audible spectrum. The basic Idea is that the Shannon entropy is a measure for peakyness of a distribution and a simple rhythm would have only one peak and more complex rhythms feature more peaks.
Use the notebook to analyze a YouTube video or better a playlist. Whole albums in one file tend to overestimate the entropy as song length also contributes to higher entropy. Also songs with varying tempos tend to have a very high score despite the rhythm and structure being fairly simple. The value is calculated via Shannon entropy but I wouldn't consider it to be information, as a Fourier spectrum does not satisfy the requirements of a probability distribution.
- Metronome : 100
- Techno: 400
- Indie Rock: 650
- Funk: 700
- Jazz: 800
- Classical music: >1000