I have a hardware conference bridge which is used for ad-hoc conferences between on-site phones.
The device pool for this conference bridge (DP1) is such that one group of phones communicate with it via G.711 (DP2) and another via G.729a (DP3). This is due to region configurations on the CUCM.
What I find surprising is that when when DP2 phones add DP3 phones to an ad-hoc conversation, RTP traffic from the conference bridge communicates with DP2 phones in G.711 and DP3 phones with G.729 without any transcoding profile being configured on the conference bridge. The DSP only has a conferencing profile configured.
How does this work? Why aren't transcoding resources necessary? When would they be necessary in such a scenario?
Thanks in advance!
dspfarm profile 1 conference
maximum sessions 16
associate application sccp
The DSPs that are the conference bridge are doing transcoding in addition to conferencing, this is part of their function.
You can use the command "show dspfarm profile" to view the codecs supported by your dsp-based conference bridge. Any codecs listed will be supported to participate in a conference call with the DSPs providing transcoding in addition to conferencing.
You can use the command "show sccp connections" (among others) during an active conference call to view the codecs used when connecting to the DSPs.
But is the DSP supposed to be transcoding without explicit transcoding configuration? There's no transcode DSP profile, and I recall those are an integral part of transcoding between any two codecs. Is this not the case with PVDM3/ISR G2?
By listing different codecs in your conference dspfarm profile, you are telling the conference bridge to transcode among those codecs as part of the conference process. So your system is behaving as programmed.
Does this make sense? If not, let us know what questions you still have.
So the "dspfarm profile 1 transcode" syntax is deprecated/irrelevant for conferencing purposes? Is it only for transcoding between call legs for regular calls?