/haar-transform

Haar transform of a sound file.

Primary LanguagePython

haar-transform

Haar transform of a sound file based on the technique described in http://www.cis.upenn.edu/~jean/gbooks/linalg.html .

The original sound file (handel_truncated.wav) requires 131 KB of storage. After haar transform with compression (handel_haar.npz), storage requires 89 KB. The reconstructed sound, although less crisp, resembles the original file.

A demo can be found here:

haar transform demo