Transcoding showing up more on the log file

[voipguy]

Hi,

 

Since moving from pbxnsip to snom one I have noticed the media CPU usage graph a little higher then normal - even without people playing recordings via the web interface.

 

 

Can you look at this and tell me if this is normal?

 

[6] 2012/05/24 10:13:01:

Call-leg 134: Sending RTP for 34dfca20e4@(my voip proxy ip) to (ip removed):6020, codec not set yet

 

[6] 2012/05/24 10:13:02:

Call-leg 134: Codec g729/8000 is chosen for call id 34dfca20e4@(my voip proxy ip)

 

[6] 2012/05/24 10:13:07:

Call-leg 135: Codec g729/8000 is chosen for call id b06fd3c8@pbx

 

[6] 2012/05/24 10:13:07:

Call-leg 135: Sending RTP for b06fd3c8@pbx to (ip removed):49990, codec g729/8000

 

 

[6] 2012/05/24 10:19:09:

Call port 134: Different Codecs (local nse/8000, remote g729/8000), callid b06fd3c8@pbx, falling back to transcoding

 

 

In the actual call log this call was an incoming call to a DID in the PBX. The call started at 10:13am lasted 6 minutes and ended at 10:19am - it looks like when the call was ending at 10:19am that's when the PBX wanted to do transcoding.

 

We have each trunk set to G729 first then G711 second.

 

Thanks

[pbx support]

PBX has the capability to detect the codec change during the call. Probably it is detecting one side has changed the codec in the middle of the call(or some NSE event). You can wireshark to verify if that is really the case.