It would seem that, on paper, in theory, having stereo channels with three people speaking excitedly at similar times, it would help the voices have moee room to be clearly heard, compared to one channel.
I might open an episode in audacity and compare each channel independently.
If it is nearly identical for each channel, then mono would nearly double the audio quality, and you could drop to 112 kb/s vs 128, because it’s not splitting 128 to 64 kb/s per channel.
I think it’s un-necessary to use stereo for podcasts, and many other podcasts do this (stereo).
For instance, EFF’s podcast “How to Fix the Internet” is recorded in stereo at an excessive 320 kb/s quality. If instead, mono were used, the bitrate of 160 combined with mono, would in theory cut the file size by 75% for nearly identical sound quality.
I would ideally like to somehow reach out to all the podcasts I listen to, and suggest mono, as podcasts like hotboxin’ with Mike Tyson started out doing. Those early files were less than 20 MB for a 45+ minute show, using about 64 kb/s. Audio quality wasn’t stellar, and it is greatly improved with more recent episodes, so I haven’t checked what’s used now.
Switching topics to web rtc, what do you think about usong your Ethernet’s hardware clock, sybchronized to a stratum 1 time server like NIST, would affect the latency or timing accuracy?
One of the hosts could sync to that, then, using something like tailscale or some other way, could be the time server for the other hosts clocks. With all of that in the mix, it would take a long time, maybe even a few days, to stabilize the clocks together in such a complicated way. Maybe it would be better to just use the same time servers and you all sync at a set time of day just before the show. This way, your clocks are using the same source and it could help improve any hiccups in the transmission.
NTPd is the original software, but I use chrony, and if I understood how to get it working, I’d have it use my ethernet clock as a source but this seems to cause issues that I can’t understand, so I just use the normal system clock, and took that out of my configuration. I think it’s expecting to always have the Ethernet connection active, and when it is not, it can’t update the time. So I’d always be consistently a few minutes behind when offline.