/mcpJitter and Typical Voice Quality Symptoms
This section explains how to use the show call active voice command in order to identify jitter and typical
voice quality symptoms.
A general idea of jitter in the network can be determined by repeatedly issuing the show call active voice
command while a call is in progress. Ideally, these parameters should stay relatively steady. If they do, that is an
indication of smooth packet flow. However, if jitter is present, there are sharp, short term spikes such as those
shown in these two sample outputs:
GapFillWithSilence=950 ms
GapFillWithPrediction=1980 ms
GapFillWithInterpolation=0 ms
GapFillWithRedundancy=0 ms
HiWaterPlayoutDelay=350 ms
LoWaterPlayoutDelay=25 ms
ReceiveDelay=29 ms
LostPackets=0
EarlyPackets=0
LatePackets=83
.
.
GapFillWithSilence=1040 ms
GapFillWithPrediction=2350 ms
GapFillWithInterpolation=0 ms
GapFillWithRedundancy=0 ms
HiWaterPlayoutDelay=40 ms
LoWaterPlayoutDelay=28 ms
ReceiveDelay=35 ms
LostPackets=0
EarlyPackets=0
LatePackets=99
The incrementing number of late packets in these sample outputs reveal a degree of jitter. The silence insertion
indicated by an increase in the GapFillWithSilence value manifests itself as choppy voice. The predictive
insertion, indicated by an increase in the GapFillWithPrediction value, tends to manifest itself as
synthetic voice.
In order to alter the amount of voice signal that is buffered to avoid jitter buffer under-runs or over-runs, issue