Re: How to check whether audio bytes contain empty noise or actual voice/signal?

marc nicole <mk1853387@gmail.com> wrote or quoted:

I hope this question is not very far from the main topic of this list, but
I have a hard time finding a way to check whether audio data samples are
containing empty noise or actual significant voice/noise.

  The Spectral Flatness Measure (SFM), also called Wiener entropy, can
  separate the wheat from the chaff when it comes to how noise-like
  a signal is. This measure runs the gamut from 0 to 1, where:
  1 means you've hit pay dirt with perfect white noise (flat spectrum),
  0 is as pure as a Napa Valley Chardonnay (single frequency).
  (Everything in between is just different shades of gnarly.)

import numpy as np
from scipy.signal import welch

def noiseness(signal, fs):
    # Compute the power spectral density
    f, psd = welch(signal, fs, nperseg=min(len(signal), 256))
   
    # Compute geometric mean of PSD
    geometric_mean = np.exp(np.mean(np.log(psd + 1e-10)))
   
    # Compute arithmetic mean of PSD
    arithmetic_mean = np.mean(psd)
   
    # Calculate Spectral Flatness Measure
    sfm = geometric_mean / arithmetic_mean
   
    return sfm