- From: henbos via GitHub <sysbot+gh@w3.org>
- Date: Mon, 20 Jun 2022 08:36:00 +0000
- To: public-webrtc-logs@w3.org
+1 to what fippo said. To elaborate a bit: For general CPU metrics I think that is a question outside of the WebRTC statistics API, since that has applications that reach far outside of WebRTC. The internal "CPU overuse detection" in WebRTC, which is used to automatically decrease resolution (or frame rate depending on degradationPreference), is actually based on the ratio of time it takes to encode versus time in-between frames. So you should be able to achieve something very similar for the decode side based on totalDecodeTime and framesDecoded. That's probably your best bet for interop too and considering asking the SFU to send smaller streams is an application level decision I think letting the application handle all "decode overuse" logic makes the most sense and AFAIK the relevant metrics already exist. Or is something else missing that is specific to WebRTC? -- GitHub Notification of comment by henbos Please view or discuss this issue at https://github.com/w3c/webrtc-stats/issues/645#issuecomment-1160144354 using your GitHub account -- Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Monday, 20 June 2022 08:36:02 UTC