I'm currently working on webrtc project, and having a problem with audio echo when not using an earphone, or external mic, is there any work around or fix for this
问题:
回答1:
Echo cancellation is supposed to be on by default in WebRTC. You can turn it off to hear the difference:
navigator.mediaDevices.getUserMedia({ audio: { echoCancellation: false } })
.then(stream => audio.srcObject = stream)
.catch(e => log(e));
var log = msg => div.innerHTML += msg + "<br>";
<audio id="audio" controls autoplay></audio><br>
<div id="div"></div>
<script src="https://webrtc.github.io/adapter/adapter-latest.js"></script>
Echo cancellation technology can exist at many levels - software in the browser, in the OS, and sometimes even in your hardware mic - sometimes these technologies can interfere with each other.
Not knowing your system, I would recommend checking out your system's sounds settings to make sure you don't have helpful processing happening at too many levels.
For example, on OSX, there's a use ambient noise reduction feature under Sound Settings that unintuitively works better for me with WebRTC when I turn off. You only want one chef attempting echo cancellation at the same time.
If you are working on a web site for other clients, then there's not much you can do in software, and I would expect most systems to work, if not perfectly, decently well, though a headset will always be better.
回答2:
Currently there's no best solution for echo cancelation in webrtc, so the best solution i found is using AEC software, or built in echo cancelation software