Hi Yokemuratakeshi,
We can provide a solution for your case in these 2 ways.
1- The actors uses WebRTC with sub-second latency( average 0,5 seconds ) within each other. Listeners have stream through HLS with 8-10 seconds latency. Scaling listeners with HLS is much more easy and cost effective.
2- The actors uses WebRTC with sub-second latency( average 0,5 seconds ) within each other. Listeners can
have about 0.5 second latency as well via HLS. In this model, scaling is not very easy and it costs more.
Ant Media Server WebRTC solution uses SFU(Selective Forwarding Unit) structure. So that each stream is delivered to listeners and actors as a different stream. It's not mixed in server side. In this solution, listeners and actors can mute any actor and listen others.
Ref:
I'm hoping that I was able to answer your queries. If you have any question, please feel free to ask us. Thanks.
Best Regards,
Selim