Help Configuring Connection for High Quality Audio

[David Ramsay]

Hi Everyone,

I'm working on an application in which I can tolerate a bit of extra latency in order to preserve high-fidelity audio (the audio in my tests has been coming through very compressed and narrowband with music tests).  I also want to turn all audio processing off (like echo cancellation, AGC, filtering), and make sure the CELT part of OPUS is what's handling my audio. I've been googling all night and I'm still a bit unsure how to do this and test this, though I'm convinced it is possible.

Right now I've set my getUserMedia with  {video: false, audio: { mandatory: { echoCancellation: false }}} and I can control the echoCancellation feature.  I found this site which seems to suggest options like asc=ISAC/16000 and arc=opus/48000 are possible, as well as setting audio bit rates (arbr and asbr) with webRTC, though they appear to be doing it through some SDP settings.  I've also seen some hints that "B=AS" in the SDP settings might make a difference.

I have a few questions: 

1) is there a way to use PeerJS to have this level of control, or do I need to use another tool?  

2)  How do I do it?  I'm extremely new to all of this and I don't really understand how the SDP stuff relates and/or how to tie it in with node/peerjs.  I've found a few examples which seem cumbersome that incorporate the SDP session stuff and use offer/answers, but that seems to take it out of the node world.  

3) is there an easy way to test that these features are working once I've attempted to install them?  I looked into the chrome://webrtc-internals tab and I see a json file and an audio dump option.  If I see configurations in that json, does that mean they 'worked'?  How do you trace these settings through the pipeline?

Any help would be wonderful, thanks for your time!  Any links to resources that you think would help me wrap my head around this would also be welcome.

David