nextcloud
Option to disable auto gain control / audio processing for practicing music
Is your feature request related to a problem? Please describe.
We're trying do practise music over nextcloud talk together so all participants have so send audio input at the same time. The default behaviour seems to be that only one (the first or the loudest) or whatever can talk and the other participants are muted or ther volume is set lower. For sure this is helpful in "normal" conferences while only one is talking and the others are listening but its impossible to practise music.
Describe the solution you'd like
It would be great to have an option for disabling the noise cancellation. I found some ideas doing this in the web like here: https://stackoverflow.com/questions/37326846/disabling-auto-gain-conctrol-with-webrtc-app
var constraints = { audio: { echoCancellation: false, noiseSuppression: false, autoGainControl: false, } };
Describe alternatives you've considered
We're trying to deactivate this with browser settings e.g. for chromium this way: URL: chrome://flags/ search for "webrtc" WebRTC multi-channel capture audio processing. -> Enabled
In zoom it seems to be possible to disable it but I don't want to use it for many reasons ;-) https://www.youtube.com/watch?v=wkG7yOTRIAk
Would be great if you offer this option.
The biggest problem we currently face is that when you want to change such a setting, you have to reconnect :(
It would be no problem for me to reconnect. Maybe the option could be global in the Talk settings?
OK, we tried to do the following with Chromium:
URL: chrome://flags/ search for "webrtc" WebRTC multi-channel capture audio processing. -> Enabled
And start Chromium with the following options chromium-browser --disable-audio-track-processing --audio-buffer-size=16 --enable-exclusive-audio
Both didn't work for us.
Does anyone of you know a way to force deactivating the auto gain control by the webbrowser? This would be a useable workaround for us.
Hi! I have the same problem as a musician. It's hard to practice with other band-members this way. I hope someone comes up with a solution. Thanx!
I'd love this feature as well. Talk has the habbit of auto adjusting the gain on my mic to 100 all the time, which can render Talk a nightmare for other participants of a call. I need to readjust the gain every couple of minutes, otherwise I blow the eardrums of the others. So yeah, "Disable auto gain" would be a godsend.
The microphone level auto reduction causes problems for me constantly and I'd really like to just disable it completely.
So can we please get this implemented? I have people just straight up not wanting to use Talk with me now because of this problem >:(
Sure, send a pull request and we can merge it. 🤷♂️ Otherwise please stop bumping issues with comments. It's just taking up developer time preventing the issues to be actually worked on
I'm not a developer, let alone know this codebase anywhere near well enough to properly send a merge request of quality. I don't know if you've noticed this but it's commonplace for github issues to get automatically closed when nobody expresses interest in such things getting addressed. Saying that it's still desirable to be addressed is a perfectly reasonable thing to do. The maintainers aren't mind-readers, by saying that it's still worth doing they can actually know. It's really frustrating to read the copy+paste response of "we accept merge requests". The majority of people are not software developers, but still want to see changes, that shouldn't make them magically ineligible to post here.
You said it 3 weeks ago already. The issue is more than 2 years old. It will remain open.
So can we please get this implemented? I have people just straight up not wanting to use Talk with me now because of this
Guys calm down. It's a pity when using Talk makes you uncomfortable. And sure is right to mention that, as a non-dev, the only thing one can do, is explain one's needs. But complaining won't help here.
Having said that, I too want to express my wishes, that the initially explained features come to Talk. Features I think are pretty essential and over due being implemented.
But we all know the devs are busy and have tons of (awesome) stuff on their roadmap. So let's wait and maybe spread the word about Talk and the few missing features ;) The day will come and lonesome dev will fulfill our wishes (regarding this issue).
Has there been any update on this? It's still looking like zoom is the only option I have and I'd really rather not give them money if at all possible.
Has anyone tried https://chrome.google.com/webstore/detail/disable-automatic-gain-co/clpapnmmlmecieknddelobgikompchkk?
@nickvergessen do you happen to know if there's something in Talk's code that explicitly enables echoCancellation/noiseSuppression/autoGainControl?
The Disable Automatic Gain Control Chrome plugin seems to have done the trick for me for a bunch of months. But this really still needs to get fixed, as a browser plugin should not be required for an actually usable eMeeting experience. It's still horrible without it!
Things to do:
- Add a small settings icon next to the device list in the first screen
- Add an "Advanced settings" option in the device menu during the call
- On click open a modal with 3 switches:
- Noise suppression
- Echo cancellation
- Auto gain
Looking at Jitsi Meet solution (which inspired previous virtual BG implementation):
- Library with WASM module is used: https://github.com/jitsi/rnnoise-wasm
- Need some testing, but looks like it's a simple plug-in for AudioContext, which we already have: https://github.com/jitsi/jitsi-meet/blob/9013881f764c02f3ee7d3792b6677ad3ef6d1dc5/react/features/stream-effects/noise-suppression/NoiseSuppressionEffect.ts#L260-L277
