Hi
for me it makes no difference. I did a test and want separate 2 guest to separate audio inputs. I used 2 new NDI / Audio only
and separate by names. In my test both guest comes in both NDI.
So for semi use i only use the NDI surces with Video and the sound comes from one guest (all guests in this channel)
and i mute the other one. Now i´m looking for a better IP Audio solution with multitrack.
Perhaps the studiolink solution. Perhaps i find another one.
mk
I used to do that, muting all ndi/skype sources except one (I even created shortcuts to open mic for the guest who is speaking while muting all the others, to improve sync since there's a slight shift between)
Then I discovered the amazing obs.ninja p2p solution, which in theory pulls a/v separately for each guest via browser source, but that hit a wall: it appears that obs still can't handle browser source audios very well, they simply stop being recorded at a random point during the session, giving no signs of error (the bars still move and nothing odd appears in the log file. It just stops recording, period). I'm still digging into that, perhaps some hardware related issue in my end? But anyway, the video quality I get from ninja is WAY superior than skype or any other conference app I ever tried, so for now I'm just turning off the "control audio via obs" ticker for all browser sources and going with desktop audio. Other audio inputs can still be mixed in using virtual audio devices such as VB Cable or voicemeeter. So, yeah, I'm still getting all guests audios mixed together, same as skype, but with virtually no lag and with much better video, you can even pull 4k streams depending on the guest's connection and camera setup. Hope OBS developers look into this browser source audio odd behavior soon, it would be wonderful to gain control over individual audios. You can do some filtering on desktop audio, such as normalization and noise cancelling.