Apprtc call send video FrameRateInput and FrameRateSent are very low on device Spring |
|||||||||||||||||||||
Issue descriptionChrome Version:52.0.2727.0 / 8302.0.0 dev Device : Spring What steps will reproduce the problem? 0. Browse to https://appr.tc/ from device Spring and desktop (Linux) 1. Open chrome://webrtc-internals in another tab 2. Listen for audio disruptions and watch for video stutter What is the expected output? Audio quality should be good without any hitches. Video framerate should be constant and smooth without too much jitter and freezes. What do you see instead? 1) Device outgoing video is not shown clearly on desktop and there is a video lag of 3-5 seconds. 2) Chrome://webrtc-internals -> Stats graphs for send(video) shows: bitsSentperSecond : as low as 500kbps sometimes FrameRateInput = 2-8fps FrameRateOutput= 2-8 fps
,
May 10 2016
Apprtc loopback calls also shows low FrameRateInput and FrameRateSent in send video stats
,
May 11 2016
srcv@ a) In the last known good test run on the Spring, what kind of googFrameRateInput were you seeing? b) Can you get a webrtc-internals dump of a loopback call? The screenshot, while helpful, doesn't include enough data for a deeper investigation. niklase@ - do you have any ideas? The screenshot of the loopback call looks especially strange not only because of the low googFrameRateInput, but also the drops in googFrame[Height|Width]Sent relative to googFrame[Height|Width]Input.
,
May 11 2016
The drop in sent resolution is nothing strange, it's CPU adaptation kicking in and dropping the resolution. You can see the corresponding changes in 'googCPUAdaptation' at the top right corner (title cropped out). The input frame rate is strange, since that indicates that the camera only delivers 8 fps.
,
May 13 2016
> it's CPU adaptation kicking in and dropping the resolution. Hmm, it's a bit worrisome that CPU adaptation would trigger in a loopback call. But for now, let's focus this bug on the low googFrameRateInput. Assigning to srcv@ to answer the questions in #3.
,
May 13 2016
> Hmm, it's a bit worrisome that CPU adaptation would trigger in a loopback call. Why? A loopback call uses more or less exactly the same CPU as a normal call.
,
Jun 1 2016
Moving this nonessential bug to the next milestone. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jun 10 2016
tnakamura@: WebRTC internal dumps for AppRTC peer2peer calls on device Spring were attached in comment#1 Tested apprtc calls and hangouts on device Spring with following builds: M52 52.0.2743.32 / 8350.21.0 beta M51 51.0.2704.79 / 8172.47.0 stable and M50 50.0.2661.104 / 7978.76.0 stable - Send(video) googFrameRateInput and googFrameRateSent are around 6-8 fps throughout the calls - Removing "Regression" label - WebRTC internal dumps from AppRTC loopback calls on device Spring with latest M52 beta (52.0.2743.32 / 8350.21.0): https://pantheon.corp.google.com/storage/browser/chromiumos-test-logs/bugfiles/cr/610858/
,
Jun 10 2016
jansson@, can you take a look?
,
Jun 13 2016
To me it looks like a CPU issue, have you tried H264 or VP8? VP9 uses quite a lot more CPU than the aforementioned codecs.
,
Jun 13 2016
,
Jun 13 2016
,
Jun 13 2016
Yes. This issue is also seen with H264 and VP8 on device Spring Added WebRTC-internal dumps from device Spring with vp8 and h264 to this link: https://pantheon.corp.google.com/storage/browser/chromiumos-test-logs/bugfiles/cr/610858/
,
Jun 14 2016
,
Jun 21 2016
,
Jun 21 2016
I wonder if this is related to http://crosbug.com/p/52823. jansson@ What's the next step? Spring is using sw encode for vp8 and h264.
,
Jun 22 2016
Yes it does sound like that could be the issue since it's affecting those devices. Next step I guess is to figure out the cause in http://crosbug.com/p/52823. Assigning to niklase@ to mirror the cros bug.
,
Jul 10 2016
This issue is Pri-1 but has already been moved once. Lowering the priority and moving to the next milestone. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Aug 31 2016
posciak@ will this make it to M54?
,
Sep 1 2016
,
Sep 1 2016
Removing milestone as there was no work for this done for 54, and this is not a regression per above reports. Were we able to identify if this is related to http://crosbug.com/p/52823? What is the resolution used in these tests?
,
Sep 1 2016
,
Jan 6 2017
Reduced camera resolution from 1280 x 720 (camera resolution used in intial bug test report) to 640 x 360 (device supported camera resoltuion) and tested apprtc loopback and peer2peer calls on chrome device Spring with M56 56.0.2924.53 / 9000.50.0 beta Camera resolution used for tests : 640 x 360 Network: Connected to Ethernet Send Video FrameRateInput and FrameRateSent for : 1.Apprtc loopback: a) VP9: FrameRateInput: First 1min 5 fps and later 20 fps FrameRateSent: First 1 min 3-5 fps and later 20 fps b) VP8: FrameRateInput: 15-20 fps FrameRateSent: 15-20 fps c) H264: FrameRateInput: 1min 5-15 fps and later 20 fps FrameRateSent: 1min 5-15 fps and later 20 fps 2. Apprtc peer2peer: VP9: FrameRateInput: 1min 1-10 fps later 10-20 fps FrameRateSent: 1min 2-6 fps later 10-15 fps VP8: FrameRateInput: 2-10 fps later 15-20 fps FrameRateSent:2- 5 fps later 10-20 fps H264: FrameRateInput: 10-25 fps FrameRateSent: 10-25 fps WebRTC internal dumps for apprtc loopback and peer2peer calls on device Spring: https://pantheon.corp.google.com/storage/browser/chromiumos-test-logs/bugfiles/cr/610858/Spring%20Apprtc%20calls%20Webrtc%20internal%20dumps%20M56%20beta/ P.S : Apprtc call send video FrameRateInput and FrameRateSent are not very low as mentioned in the initial bug report. niklase@ / posciak@ : Please suggest if these apprtc framerates are acceptable for device Spring.
,
Feb 17 2017
,
Mar 18 2017
Activating. Please assign to the right owner and the appropriate priority.
,
Apr 13 2018
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Apr 26 2018
Ping for triaging.
,
May 18 2018
Archiving. Please reopen if this is still an issue. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by srcv@chromium.org
, May 10 2016