The Polycom Viewstations and Via-Video units use unicast UDP (RTP) traffic for data streams and unicast TCP(RTCP) traffic for signaling and control.
Part of the initialization process is an agreement on what codec's are going to be used. This negotiation process is different depending on if there is a gatekeeper involved in the conversation. The important thing to remember about a 323 MCU is that it is essentially a h.323 terminal. Any I-frames or K-frames that happen between a terminal and the MCU are between the terminal and the MCU - not between participants in the conference. There is an initialization process between each endpoint and the MCU that would handle things like data rate and terminal capabilities. I would refer you to a handful of whitepapers available on polycom's web site, especially the ones from PictureTel. http://www.polycom.com/resource_center/0,1408,997,00.html The old pictureTel whitepapers are much better written and easier to use than anything else I've found on h.323 so far. There is another excellent resource on the web/mail-list; the h323 forum. I don't recall the web site right now, do a google search I'm sure you will hit. TTFN, Bill 'VTC over IPSec' Pearch, Anchorage AK -----Original Message----- From: Michael L. Williams [mailto:[EMAIL PROTECTED]] Sent: Saturday, July 27, 2002 11:39 AM To: [EMAIL PROTECTED] Subject: Re: polycom Video Unit [7:49882] "John Neiberger" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > I suppose it depends on the unit but ours mainly use unicast to the > Cisco MCU. As far as I know they use standard H.323. The downside if > you're using an MCU is that the PolyComm units have a lot of different > codecs available that might not be known by the MCU. For example, the > Cisco MCU can only do G.711 audio, but if you let two video units > speak directly to each other they use G.726 ( I think. Maybe it's > G.722?) and it sounds much > better. That brings up an interesting question tho.... unless the MCU is converting between codecs for end stations that might want to use different codecs, must the MCU "understand" the codec or would it simply act as a relay startion for that data..... (i.e. if two end-stations are using a codec that they understand but the MCU doesn't, would it be a problem since the MCU would merely forward the "unknown" (to it) audio data to the other end station). The Cisco MCU supports many more codecs than G.711 including the popular G.729 codec (which gives roughly G.711 quality with an 8:1 compression). The G.722 (you were right.. it's G.722, not G.726) that covers from 50-6900Hz instead of 50-3900Hz as most narrowband codecs do. So if you're trying to play more high fidelity sound, you may want to use that. I haven't seen many units that support this codec though (but I have by no means seen tons of units, just a few). However, if the audio you're trasmitting is human speech, the G.722 isn't going to gain you much in terms of sound quality since it would be preserving an additional frequency range that's not used alot by human speech. Does anyone have any input or experience with how and/or when the MCU codec support comes into play? I would think that if the endpoints are at the same datarate and using the same audio/video codecs, the MCU would just be a bounce point and the actual codecs in the MCU wouldn't be utilized.... Just a theory tho.. Mike W. [GroupStudy.com removed an attachment of type application/x-pkcs7-signature which had a name of smime.p7s] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=49917&t=49882 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
