Oh, it’s is absolutely, 100%, no question using a lossy format. There’s never been a case where tools like Zoom have used anything but lossy formats.
It’s a fundamental part of how they work and, more importantly, it’s adaptive. So as the quality of the connection changes (due to congestion, packet loss, latency, etc.) so will the quality of the audio and video. You can’t do that with a lossless format.
Hence, if you have congestion, it can cause lower quality, and the more data you are trying to use, the more congestion is likely to occur.
So, yes, I would suggest turning off all cameras except for the performer.
And then, on top of that, echo cancellation can further mess things up because, fundamentally, if you aren’t using a headset or monitoring on headphones, your computer is trying to prevent feedback between your speakers and microphone. That is, effectively, what echo cancellation is (it’s call “echo cahcellation” rather than “feedback suppression” because there’s some latency involved, but it’s basically feedback control). Modern echo cancellation is pretty good, but it’s a long way from perfect.
Multiply the effect of this by however many people are in the conference who aren’t using headsets and aren’t muted, and you can end up with a big old mess.
I would say the mute policy is the number one thing you can do to prevent problems as, regardless of Internet bandwidth or quality, echo cancellation will cause degredation.
Cheers,
Keith