[FFmpeg-user] amix vs amerge / how to

[Media Mouth]

I'm looking to take Avid MXF outputs with up to 20 audio tracks, and mix them down to various combinations, e.g. Stereo Mix, Stereo music, mono dialogue only, Mono DM&E splits, etc etc. all kinds of permutations and combinations.

So the basic pipeline: Export MXF from Avid with "Direct Out" (i.e. all timeline audio tracks sent to discreet channels in the MXF) and then use FFmpeg to mix down the audio tracks in any combo.
If we can figure this out, it will make our post production far more efficient.

I perused FFmpeg's docs and various forums, but have yet to find clear, comprehensive way to use -map, -filter_complex / amerge and/or amix to achieve this.

Thanks

C. Munque.