I am hoping that
@boatbod will spot your post and provide a correct response to your first question regarding the "number of channels"
selection when configuring the audio section of a multi_rx.py json file. I almost exclusively utilize Liquidsoap for my audio processing and
streaming applications and haven't used the sockaudio module for a very long time. However, I want to recall that it was necessary to specify
two channels when setting up a Voice Channel to receive DMR transmissions.
With regard to your second question (setting up stereo broadcast streams), there are no internal provisions to stream directly to an
Icecast type server from within op25. Instead, for example, you'd configure two Voice Channels (VC's) in your json having different UDP
destination ports. Then, using Liquidsoap, you build two input sources to receive these two independent channels from op25 and process
them into left and right audio channels going to the desired Icecast server. IE: BCFY.
For additional reference, see this
example of setting up an OP25 stereo stream using Liquidsoap.