Skip to content

Instantly share code, notes, and snippets.

Last active November 12, 2023 18:10
  • Star 1 You must be signed in to star a gist
  • Fork 1 You must be signed in to fork a gist
Star You must be signed in to star a gist
What would you like to do?
Accessing Sox Audio Effects from Python via Pysox
import pysox
import librosa
import numpy as np
def apply_sox_effect(signal, sr, fxstr):
# This writes signal to a .wav file, processes it sox to another file, loads that and returns it.
# signal: a numpy list of numbers; the audio signal
# sr: the sample rate in Hz, must be an integer
# fxstr: a semicolon-separated string starting with the effect name followed by parameter values in order
# e.g., "lowpass;500" or "vol;18dB" or 'compand;0.3,0.8;-50', or if you're feeling ambitious,..
# fxstr='ladspa;/usr/lib/ladspa/;mbeq;-2;-3;-3;-6;-9;-9;-10;-8;-6;-5;-4;-3;-1;0;0;0'
# See the sox docs for more on effects:
# (Why semicolon-separated? Because sox looks for both commas and colons!)
inpath, outpath = 'in.wav', 'out.wav'
librosa.output.write_wav(inpath, signal, sr) # write the input audio to a file
tmp = fxstr.split(';')
fxname = tmp[0]
fxvals = [str.encode(x) for x in tmp[1:]]
effectparams = [(fxname, fxvals),]
app = pysox.CSoxApp(inpath, outpath, effectparams=effectparams) # apply the sox effect & get new file
out_signal, sr = librosa.load(outpath, sr)
return out_signal
sr = 44100
noise = np.random.rand(4*sr) - 0.5 # 4 seconds of noise
new_noise = apply_sox_effect(noise, sr, 'lowpass;500')
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment