Skip to content

anirudt/pipydenoiser

Repository files navigation

What?

Python audio filtering module.

TODO:

  • Analyse the noise and search for harmonics
  • Study about standard Audio Pre-processing techniques.
  • Read clip frame by frame, formatting from wave module
  • Dumping / Loading the filter coeffs, for faster computation
  • Artificial Combing + ButterWorth.
  • Time Computation Graphs

Capturing Audio from the Pi:

  • To record a clip, type in rec filename.wav.
  • To play the clip, type in aplay filename.wav.

Things Done

  • Test for more data, similar conditions.
  • Getter function for params
  • Normalise the x axis into frequency, and try to locate the voice-frequency component approximately.
  • Pass a normal LPF around it and try, get results
  • Get way for amplifying the audio, adding dB gain
  • Make a function for calculating SNR.
  • Adding OptParsing for convenience
  • Added Combing filter, for removing certain high power harmonics, using a naive thresholding method.
  • Adding ButterWorth filter, working

About

Denoiser for the Raspberry Pi in Python.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages