Tala classification approach?
#1
by
svetak
- opened
Hi,
Nice work! I really enjoyed trying this tool.
This reminds me of a class project I worked on around ~2018, where we tried to classify tala patterns. https://github.com/akan72/generativeMusic/blob/master/finalReport.pdf
The tl;dr is that we tried to convert sound frequencies to images, and apply methods to find "samam". It didn't perform as well as we'd hoped.
I wonder if
- a nice extension to this would be to characterize talas (or if thats a much harder problem to solve).
- There are better methods nowadays to characterize audio for tala (such as distilling percussive patterns).