W3C home > Mailing lists > Public > public-webrtc-logs@w3.org > April 2019

Re: [webrtc-stats] Request to add neteq waiting time to stats report (#427)

From: henbos via GitHub <sysbot+gh@w3.org>
Date: Fri, 12 Apr 2019 17:55:05 +0000
To: public-webrtc-logs@w3.org
Message-ID: <issue_comment.created-482665911-1555091704-sysbot+gh@w3.org>
I agree that you don't want to poll getStats() too often, I usually recommend once every second or every few seconds.

jitterBufferDelay is a sum of all delays of all samples, which means you can calculate the average for any interval you desire by dividing the difference between two snapshots and divide by the time difference. For example, if you poll getStats() once every second, you can calculate what the average delay was for all samples that went through the jitter buffer during that one second interval. This value should be stable enough that you can remember what the maximum value was.

I don't think the per-sample jitter buffer delay fluctuates to such extremes that you have to poll getStats() unreasonably often. Does that make sense?

I don't think jitterBufferDelay includes decoding delay, but I'm not sure.

GitHub Notification of comment by henbos
Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/427#issuecomment-482665911 using your GitHub account
Received on Friday, 12 April 2019 17:55:14 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 4 June 2019 15:32:56 UTC