https://github.com/scrapjs/gl-fourier

A series of webgl experiments to explore fourier transform calculation complexity.
https://github.com/scrapjs/gl-fourier

Last synced: about 5 hours ago
JSON representation

A series of webgl experiments to explore fourier transform calculation complexity.

• Host: GitHub
• URL: https://github.com/scrapjs/gl-fourier
• Owner: scrapjs
• Created: 2016-05-07T06:59:49.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2016-05-08T01:30:35.000Z (about 8 years ago)
• Last Synced: 2024-05-02T00:03:57.006Z (2 months ago)
• Language: JavaScript
• Homepage:
• Size: 7.81 KB
• Stars: 48
• Watchers: 5
• Forks: 6
• Open Issues: 0
• Metadata Files:
  • Readme: readme.md

Lists

• awesome-regl - GPGPU Fourier Analysis

README

        
Experiments on various types of Fourier Transforms.

Later will be formed into a natural gl-fourier tool with various methods (hopefully).

* [DFT]() — due to high paralellism, it is possible to decrease complexity to O(N*N/p), where p can be equal to N. And indeed webgl calculates dft faster than fft. Tests show faster perf, but rendered result still takes time to render.

* [FFT]() — cooley tukey parallel implementation. Falls back to DFT on the small scale. The principle is avoiding overcalculation of already calculated values of sin/cos. That requires shader to run multiple times, which is worse than parallel overcalculation. So just omitted for now.

* [Stochastic Fourier Transform]() — stochastic spectrum estimation (unique method), based on formants. Requires implementing formant function, similar to phasor. (wip)