electron VS NoiseTorch

Still trash support on Linux too. We still cannot stream without frame rate issues and cannot stream audio with our video without piping through our mics. (the community fixed this for you: link)

SteamDeck and Linux users cannot stream properly (constant frame rate issues and absolutely no audio streaming). Its been FIXED by the community for years now (https://github.com/discord/electron/pull/22) but they just don't want to do it I guess?? If you are going to support the platform, SUPPORT THE PLATFORM. I'm not going to give you money for a product that doesn't work.

There is this https://github.com/discord/electron/tree/17-x-y from over half a year ago

But my golly gosh goodness me would I throw them a party in their honor if they would just update their 2 year out of date builtin shipped electron version that prevents all the progress we've made in the past few years from being at all relevant to their client.

You can see Discord's version of Electron on GitHub: This branch is 14 commits ahead, 2140 commits behind electron:main

The article from your quick google share says that you should use getUserMedia to capture the screen using Electron. Not only does Discord's client not use this, they even go as far as to disable it from their electron fork https://github.com/discord/electron/commit/e763956140aa6bf78f02202ab4df1e942ce672ae

They even forcefully disabled support for it in their version of electron, so there isn't even a way to use it via a betterdiscord plugin

The Discord Desktop app uses Electron and their Electron fork's source code is available on Github. If you take a look at their repo you'll see that one of the patches they apply is this one https://github.com/discord/electron/commit/ca8c921ed35c741acae57ea32c340f1eaf150044 that disables getUserMedia.

Noisetorch. https://github.com/noisetorch/NoiseTorch

Interesting! How does it compare with NoiseTorch/RNNoise?

Interesting! I'm currently using NoiseTorch-ng. Although NoiseTorch works well, I don't like that you need to reload NoiseTorch every time you change a setting.

I till now haven't figured out how to setup noise reduction in PipeWire. In Pulse, it was very easy. At the present, I'm using https://github.com/noisetorch/NoiseTorch.

(I do like PipeWire)

You could always try NoiseTorch. https://github.com/noisetorch/NoiseTorch

Maybe it will be possible to integrate this or something like this to SteamOS, given it's using PipeWire: https://github.com/noisetorch/NoiseTorch

If Apex allows you to select a mic, you can make a virtual (filtered) one with noisetorch.

Don't know if you are looking for the same - NoiseTorch - Works like a charm.