Only first channel of audio used?

[turian]

In utils.py:

waveform = waveform.numpy()[0, ...]

Why is only the first channel of audio used? Why can't every channel be processed?

I think this would be a great feature for a variety of reasons. The whole message of this project is: "Don't worry what audio you put in, it will come out at 48KHz", but dropping all channels is a bit user-hostile. I am trying to do batch processing with overlap add, to create a better replicate API, but this is really getting in my way. Thank you