Egregoros

Signal feed

Timeline

Post

Remote status

dear friends

is there a way on Linux/PipeWire to do screen capture with all audio sources on my system being recorded as separate tracks? i.e. rather than monitoring my output device, i'd like OBS (or similar) to create capture devices for each of my audio sources. so, e.g. game audio in one track, browser playing music on a separate track, and voice chat in a third track. and this is crucial: i don't want to manually configure one or two tracks here, i want it to automatically capture everything separately!!!! also, is there a video container format that even supports such a file? dynamically created audio tracks midway through? surely that exists?

additionally, do you know any chat apps/modifications (especially discord extensions, but others welcome too) that make it easy to not just capture the chat app individually, but each participant individually?

---

the end goal here is to make my system audio mixing at capture time irrelevant. no matter if i can hear a particular source or particular user very well, i want it to be easy as fuck to fix that in editing to make sure that source
is comprehensible after all.

Replies

4

@sodiboo@gaysex.cloud Go to Output settings, set it to Advanced and enable more than the default Audio Track 1 under the Recording tab.

Then go in Advanced Audio Properties (from the mixer) you can select which audio source goes to which track.

And yes practically all common video container formats support multiple tracks (MP4, MKV, etc.)

@sodiboo@gaysex.cloud This is up to 6 tracks, not 1 or 2. You set it up once and everything goes right automatically after that. I guess there's a problem if you need more than 6 tracks?

I mean I guess you could make a simple Bash script that detects all available audio sources and starts a separate
pw-record for all of them using GNU Parallel?