Audio: Multichannel mixdown

[Joachim Gossmann]

Hi,


I am developing an app that receives stream containing video plus an array of independent stereo tracks (stems) to be dynamically mixed down to stereo on an iPad.
The stream is e.g. matroska with opus audio channel layout 255 (unspecified).

I have a pipeline using gst-launch-1.0 that kind of does the trick, but of course I need to set the volumes from the app.

The pipeline looks e.g. something like this:


gst-launch-1.0 playbin uri=file:/media/test.mkv ! matroskademux name=md ! decodebin ! autovideosink sync=false md. ! queue ! vorbisdec ! audioconvert ! deinterleave name=d d.src_0 ! volume volume=0.5 ! audiomixer name=am ! autoaudiosink sync=false d.src_11 ! volume volume=0.1 ! am. d.src_13 ! am.



Has anyone done something like this and would be willing to chat, or is there even an example online?



Cheers,

J