AudioSegment.split_to_mono should support arbitrary number of channels

I have a setup that generates a 4-channel wav file. I would like to split this wav file into 4 separate mono tracks. The following will only generate two mono tracks:

myWav = AudioSegment.from_file('test.wav', format='wav')
print myWav.channels // 4
print len(myWav.split_to_mono()) // 2

It appears that this is due to left/right being hard-coded into split_to_mono and audioop.tomono. I took a look at adding the capability myself, but the values for fac1, fac2 passed to audioop.tomono are a bit cryptic and I’m not quite sure what needs to happen to iterate over each channel and return a variable length list of AudioSegment objects.

[Qdrew]

That’s excellent - thank you so much, thechriskelley!

I had tried frame_width: myWav.frame_width but wasn’t getting the right result. Not entirely sure why we use sample_width, but I’m new to all this. All working well with your suggestion anyway.

Hopefully jiaaro can update the split_to_mono routine to use this approach for >2 channel audio.

[jiaaro]

A fix has been merged to master. This will be in the next release of pydub