- From: Bernard Aboba via GitHub <sysbot+gh@w3.org>
- Date: Mon, 11 Jun 2018 17:51:34 +0000
- To: public-webrtc-logs@w3.org
@ibc Perhaps we should provide examples for how this (and other encoding attributes) can be used with simulcast. But there are some practical questions to answer first. When simulcast is used, we see multiple video streams (originating from different SSRCs) outbound on the same IP address and port. Marking these streams differently typically requires resetting the socket options on each packet. I am curious whether the performance of this has been measured - it is not something that is encouraged on Windows, where it would cause multiple user mode/kernel mode transitions on each packet. So yes, the API allows this to be done, but does it actually work?? Another area of confusion relates to degradationPreference. Is this expected to influence the encoder behavior when sending simulcast and/or SVC? For example, if degradationPreference is set to "maintain-resolution" will temporal layers be dropped first, then higher resolution simulcast layers? Also, does degradationPreference affect encoding optimizations? Overall, degradationPreference seems considerably more fuzzy to me than the Priority API. BTW, RtpSender does not have a send() method in WebRTC 1.0. :) -- GitHub Notification of comment by aboba Please view or discuss this issue at https://github.com/w3c/webrtc-pc/issues/1888#issuecomment-396328806 using your GitHub account
Received on Monday, 11 June 2018 17:51:37 UTC