On 7/5/17 10:18 PM, Dale R. Worley wrote:
In a perfect world, the caller would test each media stream for whether it is "silent", and then all of those that are not silent would be mixed for presentation to the user. (Humans can usually parse multiple audio sources.) If an early dialog receives a 18x response but there is no media stream that can be matched with that callee, it may be valuable for the calling UA to include an appropriate audio signal in the mix -- a ringing tone for a 180, a forwarding tone for a 181, etc.
In a perfect world the UAC would also try to distinguish in band ringback from other "more interesting" audio. Mixing multiple ringbacks is likely to just make a mess, and mixing ringback with voice also isn't ideal.
My impression is that in practice this doesn't happen enough to expend much effort on optimizing it.
Thanks, Paul _______________________________________________ Sip-implementors mailing list Sip-implementors@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/sip-implementors