Continuing the discussion from Meetup: Our regular Friday meetup will be moved to Monday:
On local echo, I hear myself with no delay and no dropouts. On server echo I hear myself with some delay, 150ms, but no dropouts. Same results with jitter buffers enabled or disabled. With or without audio noise reduction, no dropouts. With audio noise reduction it mutes (as expected) when low threshold level is reached. but my problem wsn;t sending, it was receiving the stream.
My backhaul is sprint-net San Jose. tracert to the stun server is <28ms, to the HF WEB site <30ms.
I tried this on InMotion (server on my local network), Spiritmoving (server in the same computer), then the HF sandbox, and HQ places. It all worked the same - very well. Then I realized in all those cases my local audio mixer was probably chosen, so I took down my two domains. And then in sandbox and hq, I saw the ping times rise to 115+ms, and also experienced those long dropouts.
@chris, in the video you can hear me commenting to you on the setup. My audio was working mostly OK at that point, but as more people arrived audio fell apart for me quickly. Not sure what on my end could have caused that since the audio mixer at/near HQ is supposed to deliver a steady audio stream, hence mix together the audio streams of all sound sources and avatars present. At the last two meetups I attended I did not have this problem.
Running the task manager to view network activity, I saw the incoming bandwidth at 1.5 to 2.0Mb/sec, which given the number of people there was nominal. So, where do you think the breakups were happening?
I don’t want this problem to just be forgotten. If HF cant deliver decent audio to everyone, then it’s missed an important goal.