Frame rate produced by getDisplayMedia went up unexpectedly on Windows
Categories
(Core :: WebRTC, defect, P1)
Tracking
()
People
(Reporter: jib, Assigned: ng)
References
(Depends on 1 open bug, Blocks 1 open bug, Regression)
Details
(Keywords: regression)
Attachments
(2 files)
+++ This bug was initially created as a clone of Bug #1759734 +++
(note I cloned the above bug about Firefox no longer respecting the frameRate
constraint when passed by an application, to report a related symptom)
What appears to have "regressed" at the same time is that the unconstrained frame rate produced by getDisplayMedia went up, from around ~22 fps of a youtube window on Windows 10 to ~30 fps.
This is faster than Chrome (which is at around 17 fps). While this might sound good (whoo speed!) this discrepancy might have bad side effects, like use of more bandwidth, which could lead poorer degradation when available network bandwidth is low (causing symptoms like poor resolution compared to Chrome) since frame rate can't be curbed anymore apparently.
STRs are in bug 1760050 comment 3.
Assignee | ||
Comment 1•2 years ago
|
||
(In reply to Jan-Ivar Bruaroey [:jib] (needinfo? me) from comment #0)
+++ This bug was initially created as a clone of Bug #1759734 +++
(note I cloned the above bug about Firefox no longer respecting the
frameRate
constraint when passed by an application, to report a related symptom)What appears to have "regressed" at the same time is that the unconstrained frame rate produced by getDisplayMedia went up, from around ~22 fps of a youtube window on Windows 10 to ~30 fps.
This is faster than Chrome (which is at around 17 fps). While this might sound good (whoo speed!) this discrepancy might have bad side effects, like use of more bandwidth, which could lead poorer degradation when available network bandwidth is low (causing symptoms like poor resolution compared to Chrome) since frame rate can't be curbed anymore apparently.
STRs are in bug 1760050 comment 3.
This can cause a drop in perceived quality in low motion scenarios.
Comment 2•2 years ago
|
||
Set release status flags based on info from the regressing bug 1729367
Updated•2 years ago
|
Reporter | ||
Updated•2 years ago
|
Assignee | ||
Comment 3•2 years ago
|
||
Updated•2 years ago
|
Updated•2 years ago
|
Updated•2 years ago
|
Assignee | ||
Comment 4•2 years ago
|
||
Depends on D141937
Updated•2 years ago
|
![]() |
||
Updated•2 years ago
|
Comment 5•2 years ago
|
||
:ng is this something that can land or are we waiting on the dependent bug?
Assignee | ||
Comment 6•2 years ago
|
||
diannaS, this fix uncovers a threading deadlock issue. Further work is being done here.
Updated•2 years ago
|
Pushed by na-g@nostrum.com: https://hg.mozilla.org/integration/autoland/rev/d4b6bfda83c5 P0 - fix desktop capture frame timing;r=jib https://hg.mozilla.org/integration/autoland/rev/198b1e078f7e P1 - add WPT for gDM frameRate:max;r=jib
Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/34078 for changes under testing/web-platform/tests
Comment 9•2 years ago
|
||
Backed out 2 changesets (bug 1760843) for causing wpt failures in /screen-capture/getdisplaymedia.https.html
Backout link: https://hg.mozilla.org/integration/autoland/rev/9d4ff58359c27c213153f1cc9e9d523212d0d888
INFO - TEST-PASS | /screen-capture/getdisplaymedia.https.html | getDisplayMedia() must require user activation
[task 2022-05-16T21:09:59.776Z] 21:09:59 INFO - TEST-UNEXPECTED-FAIL | /screen-capture/getdisplaymedia.https.html | getDisplayMedia() must adhere to frameRate if set - assert_less_than_equal: expected a number less than or equal to 7.5 but got 8.359253499222396
[task 2022-05-16T21:09:59.776Z] 21:09:59 INFO - @https://web-platform.test:8443/screen-capture/getdisplaymedia.https.html:55:25
Upstream PR was closed without merging
Assignee | ||
Comment 11•2 years ago
|
||
I am investigating this. The fix in this bug was specifically for Windows. The failing test was added in this patch. It may have been that macOS will require its own fix, and this test may need to be disabled for macOS in the mean time.
Comment 12•2 years ago
|
||
Pushed by na-g@nostrum.com: https://hg.mozilla.org/integration/autoland/rev/1e5c9e81a56e P0 - fix desktop capture frame timing;r=jib https://hg.mozilla.org/integration/autoland/rev/285d75d53b7a P1 - add WPT for gDM frameRate:max;r=jib
Comment 13•2 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/1e5c9e81a56e
https://hg.mozilla.org/mozilla-central/rev/285d75d53b7a
Upstream PR merged by moz-wptsync-bot
Comment 15•2 years ago
|
||
Since the status are different for nightly and release, what's the status for beta?
For more information, please visit auto_nag documentation.
Updated•2 years ago
|
Comment 16•2 years ago
|
||
Is this something we should consider for uplift to 101 still this cycle (RC week is next week) or can this ride the 102 train?
Updated•2 years ago
|
Description
•