May this comment help you. Software recommendation seems to be the only “ simple " solution. Or settings for Windows> = 7 , as described here .
Updated:
The default level should be set to “ 0 ” until the media stream starts to stream; use "setTimeout" to wait at least " 1 " and then set the volume to " 1 ".
In fact, noise arises from the "huge audio" bandwidth that occurs as soon as the onaddstream event fires.
If you check " audioInputLevel " through " chrome: // webrtc-internals ", you will see its value approximately like " 25000 " as soon as the onaddstream event fires ; however, after a few milliseconds, " audioInputLevel " falls between 1 and-150 and remains lower.
Updated January 12, 2016
This document may help: http://www.slideshare.net/MuazKhan/echo-in-webrtc-why
You can even set VoiceActivityDetection:false as SDP restrictions to minimize the echo level at the peer level.
source share