ValueError during feature extraction about pyaudioanalysis

[tyiannak]

Yes, stereo2mono() needs to be called after file reading to ensure that the signal is converted to mono. Feature extraction on stereo signals is not supported. However, if the information from two different channels actually have a conceptual meaning for your problem, you could try extract features on the two channels separately.

[zarubaf]

I've had a similar problem. Looks as if stFeatureExtraction does not support stereo audio. I called
audioBasicIO.stereo2mono(x)
beforehand, although using numpy this:
x = np.mean(x, axis=1)
maybe faster (I actually have no idea if this is true...).