I have been having echo problems w/ my system since installation. I have opened tac cases and have gone through everything that I have been told to try. Changing the output att and input gain helped a bit, but I am still having echos. I am currently at CM 3.1(1) and am planning to upgrade to 3.1(3). I am using a 3640 w/ a dual HDV mod for my h.323 gw. Is there a phone load that seems to help cut down on echo? I am running P00303010030 on my 7940 phones. Phone loads are the only thing I haven't tried. Any suggestions?
Do you experience echo on calls to the PSTN from an IP Phone?
There are some common problems there. First, there is a 10 second "convergence time". Not much you can do about that currently. That is the way the current firmware works.
You should be sure to set echo cancel coverage in the voice port configuration to the max value (32 on the 3640, 128 on the 5350,5400). Keep in mind that 32ms may not be enough depending on the echo tail length on the PSTN calls. The tail is the range of time the echo canceller can cancel echo (0-x)if it is beyond the max, it is not canceled.
Output attenuation should be 3 or 6 which will help. Input gain will have little or no effect unless the echo level is borderline.
It helps to understand echo on PSTN calls. There are two factors that make echo noticeable. The signal level of the echo and the delay. We need to understand the delay in a VoIP call. There is a ~25 ms delay in the IP phone, transport delay ~10ms and ~65ms in the 3640. There is usually very little delay in the PSTN, say ~5ms. Then you have the return trip, ~30 in the 3640, ~10 transport and ~25 in the phone. So from when you speak there is a ~170 ms delay to when you hear your own echo. The longer the delay, the more the echo has to be attenuated so you do not hear it.
If the outgoing voice is at level -6 and there is a echo return loss (ERL) of 15, then the echo level is around -21 db. This is where time comes in. If the time is short >25ms, you don't detect the echo. If the delay is 170ms you hear it loud and clear. So, we need echo cancellers. The echo canceller will attenuate the echo level to some target level. Considering the delay, the target has to be below -50db for this example so we don't hear it. So, the higher the ERL, the more the echo has to be attenuated before it is not detectable.
When echo is experienced between two IP phones, is normally acoustic or caused by the impendence mis-match in a headset. Normally, you can just lower the volume on the IP phone and make this kind of echo go away.
If the attenuated signal has not complete eliminated the echo (just made it very faint), input gain could possible reduce it even further. Thsi will be at some cost though. This will reduce all incoming levels, and will likely cause problems hearing folks from the PSTN. When this happens, IP phone users will turn up their volume which might cause PSTN users to start hearing some echo.
Input gain should not be used to attenuate the signal to reduce echo as the potential cost is too high. Input gain should be used to increase the signal level of incoming calls if the levels are too low.
I strongly suggest reducing the delay in the network as much as possible. This has a direct impact on
Are you getting this error “Installer User Interface Mode Not Supported. The installer cannot run in this UI mode. To specify the interface mode, use the -i command-line option, followed by the UI mode identifier. The value UI mode identifiers...
The below trick might come handy when you have to add a new node to a cluster but you don't have or is unsure of the security password for the publisher. This procedure has been around for ages.
1) Login into the CLI of the Publisher.