Core Issue
The audio power originating from the IP network may be too high, which causes a reflection to return from the Public Switched Telephone Network (PSTN). Delay in processing in the packet network causes this reflection to be perceived as an echo.
Note: The sound of an echo can vary depending on a number of factors. If the secondary signal is delayed considerably and has very little loss, then it would likely be perceived as a second discrete signal or echo. At the other end of the spectrum, a secondary signal could be minimally delayed with more loss. In this case it might be perceived more like Tunnel voice or it may not be perceived at all.
Resolution
The key issues are identifying which end hears the echo and whose voice is heard. When a talker hears his own voice echoed, it is known as talker echo. This is the usual case. When a listener hears the speaker's voice twice, it is known as listener echo, which is uncommon. If the talker hears his own voice echoed, the echo is generated at the far end.
In this case, the IP phone user hears his own voice echoed. Therefore, the echo is sourced from the PSTN.
To resolve this issue, choose from one of these options:
- Reduce the level of the signal on output from the gateway to the PSTN by using a positive output attenuation value.
For specific procedures, refer to these examples:
For more information on echo problems, refer to these documents:
Voice Quality
Echo (Listen to short or long echo samples.)