[Videoconferencing] Should we buy our own H323 MCU ?

[Isabel Trigger]

Andrew,

A couple of thoughts from the ATLAS side:

We switched to videoconferencing with EVO mainly because phone 
conferencing was expensive, so the 36$/hour/line will negate that 
argument for using videoconferencing in the first place.

I would think the refurbished RadVision for ~1000$ would come closest to 
fulfilling our immediate needs.  Something that would allow for up to 12 
connections is certainly desirable, and I don't think we are ready for 
the "future-proofed" system without more testing.  EVO is good in 
principle but by switching to H.323 we would eliminate the problems 
associated with people with really unsuitable equipment degrading the 
quality of the whole meeting.

thanks
.. Isabel

Andrew Daviel wrote:
> 
> (MCU - "multiple calling unit", aka. bridge)
> 
> Background:
> 
> Although 2 of our H323 video sets (Polycom Viewstations) can host up to 
> 3 connections, normally we route all our video conferences through the 
> ESnet MCUs at LBL. (http://www.ecs.es.net/ecs_reg/index.html)
> 
> This generally requires that endpoints be registered with the ESnet 
> gatekeeper (on our sets, a password-protected administrative option, not 
> a per-call setting). ESnet policy requires that the user be working on 
> DOE-funded research, which I interpret as "some user at the site must be 
> part of a collaboration that includes a DOE-funded user" to register, 
> and "the user must be part of a collaboration that includes a DOE-funded 
> user" to place a call. All I can find online is the general AUP
> at http://www.es.net/hypertext/esnet-aup.html
> which says:
> "Acceptable use of ESnet includes ... Video conferencing among Office of 
> Science investigators and their collaborators or facilities for use 
> between or among other organizations."
> 
> The infrastructure at ESnet includes a gatekeeper and multiple Codian 
> MCUs managed and maintained by ESnet, which are connected to GDS
> (http://commons.internet2.edu/gds.html)
> 
> There is also an MCU at BCnet which we can use. It is not part of GDS, 
> and requires booking. We have never used it significantly.
> 
> None of these MCUs are HD (High-Definition) capable, though an HD 
> endpoint can connect using legacy settings - just as an HD set can 
> connect to a legacy set. We currently have no HD sets at TRIUMF, 
> although we are considering a purchase.
> 
> (EVO is a totally separate system which does not use MCUs. There is a 
> path within the EVO client to allow EVO calls to connect to ESnet 
> conferences). (http://evo.caltech.edu)
> 
> I believe we should be able to have an ESnet conference with an 
> additional 2 non-registered sites, but only in ISAC2 room 223.
> 
> Any GDS site should also be able to take part in an ESnet conference, 
> with any of our endpoints.
> 
> 
> Discussion:
> 
> If we bought our own MCU, we could host multipoint conferences of more 
> than 4 participants with no policy restrictions. This has been requested 
> by one group.
> I am told that a bare MCU (no gatekeeper) does not require much 
> administration or setup.
> 
> New HD-capable (future proofed) MCUs are seriously expensive (> $100,000 
> according to one source). New non-HD Codians (like ESnet have now), 
> maybe $40k. A used/refurbished Radvision MCU may be had for about $1000. 
> Like ESnet used to have - it won't do transcoding so the conference runs 
> as the speed of the slowest endpoint. If I understand properly, that (a 
> used one) would let us have maybe a 4-site conference from any of our 
> endpoints at 768kbps, or an 8-site conference at 384k, or a 12-site 
> conference at 128k. And
> 
> As it happens, I have just been talking with Intercol (reseller trying 
> to sell us stuff) who offer an outsourced bridging solution. Either 
> $1800/month for 4 lines unlimited ad-hoc, or $36/hour/line for a booked 
> meeting, using HD-capable MCUs that can also bridge to ISDN or regular 
> phone.
> 
> Question:
> 
> Is there any interest in either of these (buying an MCU, or outsourcing) ?
> We might be able to get a Radvision on the maintenence budget.
> 
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: itrigger.vcf
Type: text/x-vcard
Size: 456 bytes
Desc: not available
Url : http://lists.triumf.ca/pipermail/videoconferencing/attachments/20081030/4ef73cb0/itrigger.vcf