Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Examples: Monday, today, last week, Mar 26, 3/26/04

First successful test and latency question

Was finally able to make a successful connection between remote sites after a few weeks of setting up and troubleshooting the Jamblasters. The audio for the most part sounded good and the latency was low enough to be as effective as the musicians being in the same room. I miss not having gain controls on the hardware now that levels are software based and the software still feels a bit sluggish so hopefully will keep improving.

My question is around the statistics provided within Jam Kazam. A snaphot of the stats during the session lists the Total latency as 14ms. This appears to be made up by totalling the audio interface latency, input jitter, output jitter, Frame size, Internet Latency, Jitter Queue and Jitter.

Where is the 5ms amount from the Opus codec encoding / decoding within these stats? I've tested this though the Jamblaster box by comparing the time difference of the sound going into the box and out of the headphones compared to the sound in real air. If there 2.5ms required at minimum to encode the audio before sending it over the internet, why doesn't this amount appear in device latency? If it's then 2.5ms to decode at each end, wouldn't the effective total latency of each way be 19ms in this case?

Maybe I'm missing something but any insight into this would be much appreciated, thanks.

Comments

  • The stats are incorrect. There is a bug open to get it fixed.
  • ah ok thanks for the info. I've yet to do a loopback test to measure the actual roundtrip latency so when I get a chance to do that hopefully that will provide some more accurate stats in the meantime.
Sign In or Register to comment.