ffmpeg: How to merge multiple audio files into stereo

Question

I have an initially undefined number of input mono audio files (.wav) that I want to merge into a single stereo output audio file (.mp3), so that one of the input files goes into left channel, and all the remaining input files are joined into the right channel.

I have seen https://trac.ffmpeg.org/wiki/AudioChannelManipulation#a2monostereo of how to do it for 2 input files only, not sure how it should be changed when dealing with 2+ input files though?

At the end of the day I want to do it using the ffmpeg-python library. Right now I have this which is working fine for 2 ffmpeg_inputs:

            (
                ffmpeg.filter(
                    ffmpeg_inputs,
                    filter_name="join",
                    inputs=len(ffmpeg_inputs),
                    channel_layout="stereo",
                )
                .output(tmp_output_file.name)
                .run(overwrite_output=True)
            )

but as soon as more than 2 ffmpeg_inputs are passed, the extra ones do not end up in the output file. Guess it needs some fancy mapping, similar to the raw ffmpeg command above.

Answer 1

1. You can first create an input streams.
   inputs = [ffmpeg.input(f) for f in input_files]

2. Mix all but the first into one right-channel stream
   right_mix = ffmpeg.filter([inp.audio for inp in inputs[1:]], 'amix', inputs=len(inputs)-1, normalize=0)

3. Then, Join left (first file) and right (mixed)

   stereo = ffmpeg.filter([inputs[0].audio, right_mix], 'join', inputs=2, channel_layout='stereo')

4. output

   (ffmpeg.output(stereo, output_file).run(overwrite_output=True))

This combines the first input as the left channel and mixes all remaining inputs into the right channel. Hope this will help.