This message was deleted.
# helpdesks
steep-balloon-41261
08/17/2023, 10:28 PMThis message was deleted.
d
dry-elephant-14928
08/18/2023, 6:20 PMHey @enough-holiday-1946, what is the context here?
Is this still related to Egress compositing? or are you building with our Components framework?
e
enough-holiday-1946
08/20/2023, 10:25 PMHi @dry-elephant-14928. First thanks for your hard work with livekit. It is amazing! I have not started yet, but yes, i am thinking of the Components framework (or using Flutter is also ok). I am just thinking out loud what event(s) and how to achive it.
So, basically, i would sometimes like to show a 100% width/height video track (covering the full room screen) and not share it with the placeholder of other participants in the room. For example if i with livekit-cli publish a local video to meet room with the "Botuser" identity, it would share half the screen while the other half is my camera/or placeholder. I tried to publish a video with my identify (the one that first joined Meet room) but then i got "disconnected" when published the video. I would still like the participant (non bot video track) to be able to speak (but not showing a grey placeholder or camera). More or less "hide" the participant from the viewarea programmatically.
enough-holiday-1946
08/20/2023, 10:27 PMI guess another way of saying it is that the video track published should be seen as the "Speaker" and the layout should be updated to speaker (from grid). Could i then still hear audio from other participant (will only be one) ? But is that even possible to promote the video track to be the speaker from a layout perspective ?
enough-holiday-1946
08/20/2023, 11:13 PMAnd just to add, yes, i would like this room viewport logic/layout to be exported in the room egress.
enough-holiday-1946
08/20/2023, 11:36 PMDid looke into StartTrackCompositeEgress with a audio+video trackID to export to a RTMP which seems what i would like to do, however only sometimes within the video stream. Most of the time i would like a plain StartRoomCompositeEgress with grid layout. Would it work to send to the same RTMP endpoint first using a StartRoomCompositeEgress then sometimes send StartTrackCompositeEgress. Or otherwise updating the StartTrackCompositeEgress would mabey have solved it, do dynamically change the VideoID and AudioID seamless.