Audio-only share option

implementors: please make this just a special case of a video share instead of an Entirely Distinct Thing Internally. i also share "just audio" sometimes. when i'm screensharing, i only use niri's "dynamic cast" functionality, which lets me easily switch which window I'm sharing using compositor hotkeys. but i can also set it to share nothing at all; at which point no more frames are sent to the recipient. different apps handle this differently, from freezing the last shown frame, to just displaying black, to "buffering" (even though there are no network issues), to just forcefully shutting off the stream because it assumes something broke. it would be really nice if Fluxer could instead just turn my screenshare tile into one that says "audio-only", and you could still interact with it (change volume / stop sharing / etc) just the same as a video screen share, and then when i select a window to share again it will just continue without interruptions. switching between audio/video share on the fly should be seamless for viewers/listeners. i should not have to restart my stream entirely to have "the other mode".

I’d love to see this feature too. It’s really useful, and in China, similar voice chat apps (such as Kook, Heybox and OOPZ) already offer it. They can detect supported music apps and let you pick one to share its audio directly into the voice channel. I’ve attached screenshots from HeyBox and OOPZ as examples. Most of the preset sources are music apps popular in mainland China (NetEase Cloud Music, QQ Music, Kugou, etc.), but Spotify is on the list as well. :) If the app you want to share isn’t in the presets, there’s also a “More sources” option that lets you pick any other running program and share its audio instead. There’s also a dedicated “Accompaniment Volume” slider, so you can adjust the shared music’s volume separately from your voice. On the listener side, you can also right-click a member in the voice channel and adjust the volume of the accompaniment they’re sharing, independently of their voice. Hopefully this gives the devs a bit of reference. :D