Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 
New Member

Designing Transcoding After the Fact

My company consists of three sites connected by metroethernet. Site A could be considered the main site--CCM Pub and 2 Subs, Unity, ICM Side A, and IPIVR live at this site, along with over 1000 IP phones. Site B has a couple of CCM Subs, a Unity server, ICM Side B and IPIVR. Site C is a smaller branch office with approximately 80 IP phones and a local CCM Sub.

Codec g.711 is currently used throughout the enterprise. However, I need to look into the possibility of changing the links to use g729 to reduce bandwidth consumption.

My big concern with the implimentation revolves around transcoding--with IPIVR (which to my knowledge is g711 aware only)involved, I feel that transcoding will become a big necessity between sites--especially in a scenario where calls need to get routed over the WAN to an IPIVR. I can configure the change of codec in CCM and I'm pretty familiar with configuring the voice gateways for transcoding support. But I have several questions, mostly design related:

1) How can I begin to tell how many transcoding sessions I would need at each site--I would assume it has something to do with the number of IVR ports at each site?

2) How do I know if I have enough DSP resources available to configure the necessary transcoding resources? The current voice gateways are pretty packed with DSPs, but they are also supporting several T1 PRIs each, which obviously use DSPs as well.

My biggest fear is that I switch over to the g729 codec and have to wait for call volume to increase to a certain level before problems arise (transcoding sessions failing due to lack of resources). Obviously, I'd need to be reasonably confident that things are provisioned correctly before cutting from g711 to g729.

As always, I really appreciate the help of all the folks who are in the field and may have experienced this before.


Re: Designing Transcoding After the Fact

My CRS experience is pretty limited, but a couple thoughts come to mind. CRS 3.1 or later can be configured to use G711 or G729 (but not both). Changing the codec requires a compete reinstall, so it wouldn't be fun, but would eliminate the need for transcoding.

Are you using speech recognition? This is not supported with the G729 codec configured. I suspect it may not work with G729 calls transcoded to G711 as well. You would want to test this.

Please rate helpful posts.

New Member

Re: Designing Transcoding After the Fact

The issue with that is that I will still be using g711 intrasite---g729 will only be used between sites. Example: Phone/call in site A needs to connect to IVR in site A-should be no problem (all g711). But if Phone/call on Site A needs to go over the g729 WAN to hit the IVR on site B-for whatever reason, IVR side A failure being one that comes to mind- I'm assuming that transcoding will need to be invoked on site B to transcode from g729 to g711 for the IVR to work. We have IP-IVRs on both sites for IPCC redundancy purposes, so we need to be able to hit them from either site.

CreatePlease to create content