James Cloos writes: > >> baresip and webrtc sip clients specify opus codec like this: > > >> a=rtpmap:96 opus/48000/2. > >> a=fmtp:96 stereo=1;sprop-stereo=1. > > >> whereas sems does it like this: > > >> a=rtpmap:105 opus/48000. > > SS> I noticed that, too, but I've had audio issues with e.g. using sems > SS> echo or conference with stereo clients in master. > > To be consistant with draft-ietf-payload-rtp-opus-03.txt, sems ought to > use: > > a=rtpmap:105 opus/48000/2 > > The two sides should be able to negotiate stereo=0 provided that the > rtpmap is opus/48000/2. (Note that on its own opus/48000/2 is mono, > even though it looks like stereo.)
so when sems receives a=rtpmap:96 opus/48000/2 a=fmtp:96 stereo=1;sprop-stereo=1 it should accept it and reply with a=rtpmap:96 opus/48000/2 stereo and sprop-stereo params are not needed in reply because they both default to 0. still, this reply does not guarantee that caller will actually sends mono signals. -- juha _______________________________________________ Semsdev mailing list [email protected] http://lists.iptel.org/mailman/listinfo/semsdev
