latency and jitter in voice calls(Audio Stream is not stable)

Unanswered Question
May 9th, 2010
User Badges:

Hi all

I have implemented CUCM 7 Publisher and Subscriber.

i had configured internal calling using four digits and also local callings.

The problem i am facing is that when users are making local calls the voice stream is not stable i have configured Codec G711.

When i tested it by playing the audio stream continuously on the other end and i was not able to listen it clearly.

Please any body can tell me why its like this ?????????????????

what should i do to resolve this issue.

I have attached the Region Configuration.

thanks and best regards

Jamil Hussain

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Paolo Bevilacqua Sun, 05/09/2010 - 03:44
User Badges:
  • Super Gold, 25000 points or more
  • Hall of Fame,

    Founding Member

What is there exactly between phones and voice gateway ?

Also when making a call press question mark on the phone twice, report results here.

Jamil Hussain Thu, 05/13/2010 - 07:10
User Badges:

My Phones are connected to POE Switch and My Gateway is connected to that switch also.

i made a local call where i am facing problem and after that i pressed ? Twice and it gave me the below output.

Rcvr Codec : G.729

Sender Codec : G.729

Rcvr Size : 20ms

Sender Size:20ms

Rcvr Packets : 1018

Sender Packets : 3489

Avg Jitter : 10

Max Jitter : 13

Rcvr Discarded : 19

Rcvr Lost Packets : 0

MOS LQK : 2.0000

Avg MOS LQK : 2.0000

Min MOS LQK : 2.0000

Max MOS LQK :2.0000

MOS LQK version : 0.95

Cumulative Conceal Ration : 0.7838

Interval Conceal Ration : 0.000

Max Conceal Ration : 1.0000

Conceal Secs : 116

Serverely Conceal Secs: 115

Latency : 0

Network Protocol :

this is the output that i received

William Bell Thu, 05/13/2010 - 07:17
User Badges:
  • Purple, 4500 points or more

Situations like this point to the voice quality provided by your network infrastructure not a CODEC compatility issue or even CODEC performance issue.  To troubleshoot this problem you will first need to determine the scope of impact.  Meaning, what areas of your network need to be involved for the quality to become poor.

1. Check a phone call between two phones on the same LAN network switch 2. Check a phone call between two phones on different LAN network switches (you may need to check other switches) 3. Check a phone call between two phones on different sites attached via a WAN 4. etc.
5. Check calls to/from the PSTN

You may have to try different scenarios to start getting an idea of scope.  In this exercise you aren't really focused on finding what exact scenario causes an issue as much as you are excluding scenarios that do not cause an issue.

Once you find the scenario(s) that demonstrate the issue, you will then need to check if there are other factors such as call duration, who places the call (e.g. PhoneA has a problem only when a PSTN caller calls PhoneA, outbound calls are OK), call features used during call, etc..  Using this additional pieces of data you can start honing in on root cause.

You may also ask yourself questions like:

1. Is it all phones or just a subset of phones that are affected?  For example, if it is always the same phone that has a problem, but all other phones are fine, then you have found your culprit.
2. Are phones of a particular model the only ones reporting the issue?  If so, then you have to look at firmware levels.
3. Are phones having the issue ONLY when calling (or receiving calls) the PSTN?  Then you need to look at the DSPs on your voice gateway.  You need to determine DSP firmware, IOS version, etc. and check for defects using the bug toolkit.
4. Are phones having an issue calling other IP phones when phones are on different switches?  You could like at design things: 
- QoS configurations
- Are voice and data on separate VLANs as they are supposed to be
- How large is the voice IP subnetwork?  Is it larger than 512 hosts.  If so, this is a violation of Cisco best practice. 
- Are inter-switch connections presenting a choke point due to not enough bandwidth.  Proper qos configs should mitigate the impact of this unless the problem is extreme and you would likely not need to ask what your issue is as it would be obvious.

You could look at operational things:
- QoS performance (see if packets are dropped)
- Link performance.  Particularly on uplinks, you will want to check for errors on the interfaces.  I have had many voice quality issues be resolved by swapping out a bad GBIC or compromised fiber connection.  If you see errors on your uplink interfaces.  Resolve them before checking/testing anything else.

There are plenty of things to check, consider this a jump start.  Summary: Isolate your issue to something that is better defined by excluding scenarios that are OK.  Drill into the various permutations of the remaining scenarios for further exclusion (and possible root cause identification).  Check design, check configurations, and check interface/DSP statistics.


Please remember to rate helpful posts.

Jamil Hussain Thu, 05/13/2010 - 07:23
User Badges:

Internal calling is working just fine the voice quality is ok but when I am calling outside the company I am facing this problem from any of the phone

I noticed one thing that when I made a call and when my call got connected and I pressed ? Twice in the result it showed Send and Receive codec g.729

Even when I am making internal calling It's showing me the same.

I think there is the codec issue.It should be G.711 because the voice quality is good with G.711 as compared with G.729

What is your opinion about this ????????

Best regards,


Jamil Hussain

jbburks Mon, 12/19/2011 - 13:46
User Badges:

This sounds similar to a problem we are having.

All internal calls are fine, as are incoming calls that go through one H.323 gateway.

Outbound PSTN calls get bad quality intermittently, and they go out through another H.323 gateway.

Did you ever figure out what the problem was with your system?

Athonia Cappelli Mon, 12/03/2012 - 14:44
User Badges:

Hi folks, did you determine what was causing this issue yet?

We have a similar problem but it's only with offices that don't have a PRI on-site. One issue we're seeing occurs at a satellite rental office we have for a few staff in the midwest. They replicated the problem on 7962 phones configured first as proxy then as VPN anyconnect phones. The VPN phones connect to ASAs in California. Bandwidth and ping times are decent at 2Mbps/2Mbps & 10ms respectively. However, we're seeing max jitter at over 2000ms. The MOS score is abhorent.  IF their bandwidth and latency is not at fault. what could be causing the audio anomalies? 

Bandwidth sample from the computer connected to trunk port on the 7962 phone:

Below is the output from Stream1 on the phone:

Avg Jitter
Rcvr Codec
Rcvr Reports Sent
Rcvr Report Time Sent
Rcvr Packets
Rcvr Octets
MOS LQK Version
Cumulative Conceal Ratio
Interval Conceal Ratio
Max Conceal Ratio
Conceal Secs
Severely Conceal Secs
Max Jitter
Sender Size
0 ms
Sender Reports Received
Sender Report Time Received
Rcvr Size
20 ms
Rcvr Discarded
Rcvr Reports Received
Rcvr Report Time Received
Gaurav Purohit Tue, 12/04/2012 - 04:22
User Badges:

Hello Jamil,

can you post the dial-peer configs for the GW? are you sure that these dial-peers are in fact being used for external calls?

Gaurav Purohit Tue, 12/04/2012 - 04:28
User Badges:

Hello Athonia,

you may try using "show call active voice brief" during active call can confirm if the late packets (jitter) is in reality increasing to unacceptable limit.

Athonia Cappelli Wed, 01/09/2013 - 15:16
User Badges:

Yes, jitter is very high with max jitter often above 1000ms. For some reason when I do an extended ping from Irvine to a phone in L A there's rarely a period where latency exceeds 10ms. Still, the continue to have problems when ever they make a phone call that goes out the PRI in the Irvine office. Here's a sample from a call today that had the "tron voice" (as one of our office staff calls it).

Conceal Secs
Severely Conceal Secs
Max Jitter
Sender Size
20 ms


This Discussion