Project: chromium Issues People Development process History Sign in
New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.
Issue 507834 ☂ Project Spitzer: the desktop media pipeline on Android
Starred by 57 users Project Member Reported by w...@chromium.org, Jul 7 2015 Back to list
Status: Fixed
Owner:
Closed: Feb 14
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 2
Type: Bug


Sign in to add a comment
This is a tracking issue for enabling a new src= media playback path for Android. Using:

* FFmpeg for demuxing, and decoding audio
* WebMediaPlayer_Impl instead of WebMediaPlayerAndroid
* AndroidVideoDecoderAccelerator for decoding video with MediaCodec
 
Comment 1 by w...@chromium.org, Jul 7 2015
Blocking: chromium:401795
Comment 2 by w...@chromium.org, Jul 7 2015
Cc: liber...@chromium.org
Comment 3 by w...@chromium.org, Jul 17 2015
Blockedon: chromium:511375
Comment 4 by w...@chromium.org, Jul 17 2015
Blockedon: chromium:511376
Comment 5 by w...@chromium.org, Jul 17 2015
Blockedon: chromium:511430
Project Member Comment 6 by bugdroid1@chromium.org, Jul 20 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/45c3063debd193f953d5c4aa55902d185469f4bc

commit 45c3063debd193f953d5c4aa55902d185469f4bc
Author: watk <watk@chromium.org>
Date: Mon Jul 20 20:16:15 2015

Initial support for the desktop media pipeline on Android.

Add experimental support for the desktop src= media pipeline on Android
behind a compile-time variable. To compile with it
enabled, set the gyp variable media_use_ffmpeg=1.

Currently only H/W accelerated VP8 is supported because that is all that
the existing AndroidVideoDecodeAccelerator supports.

BUG= 507834 

Review URL: https://codereview.chromium.org/1230793009

Cr-Commit-Position: refs/heads/master@{#339498}

[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/content/renderer/render_frame_impl.cc
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/blink/media_blink.gyp
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/ffmpeg/ffmpeg_common.cc
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/filters/audio_decoder_unittest.cc
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/filters/ffmpeg_glue_unittest.cc
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/media.gyp
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/media_cdm.gypi
[add] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/media_variables.gypi
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/media/renderers/default_renderer_factory.cc
[modify] http://crrev.com/45c3063debd193f953d5c4aa55902d185469f4bc/third_party/opus/opus.gyp

Comment 7 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516540
Comment 8 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516765
Comment 9 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516766
Comment 10 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516845
Comment 11 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516850
Comment 12 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516857
Comment 13 by w...@chromium.org, Aug 4 2015
Blockedon: chromium:516872
Cc: klo...@chromium.org qin...@chromium.org ti...@chromium.org
cc: android folk as fyi for upcoming meeting.
Comment 15 by w...@chromium.org, Aug 6 2015
Blockedon: chromium:517673
Comment 16 by w...@chromium.org, Aug 7 2015
Cc: bol...@chromium.org tombergan@chromium.org
ccing flywheel people!
Comment 17 by w...@chromium.org, Aug 7 2015
Blockedon: chromium:517701
Project Member Comment 18 by bugdroid1@chromium.org, Aug 7 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/dce82facb9614af36897f3c076e1842773871f1e

commit dce82facb9614af36897f3c076e1842773871f1e
Author: watk <watk@chromium.org>
Date: Fri Aug 07 19:16:54 2015

Enable software decoding video with libvpx on Android

VpxVideoDecoder previously would reject VP8 without an alpha
channel, leaving it to be decoded by FFmpegVideoDecoder. Now,
when it's compiled without FFmpegVideoDecoder it will handle
all VP8.

BUG= 507834 

Review URL: https://codereview.chromium.org/1281433002

Cr-Commit-Position: refs/heads/master@{#342405}

[modify] http://crrev.com/dce82facb9614af36897f3c076e1842773871f1e/media/filters/vpx_video_decoder.cc

Project Member Comment 19 by bugdroid1@chromium.org, Aug 7 2015
The following revision refers to this bug:
  http://src.chromium.org/viewvc/blink?view=rev&rev=200193

------------------------------------------------------------------
r200193 | watk@chromium.org | 2015-08-07T20:26:36.686244Z

Changed paths:
   M http://src.chromium.org/viewvc/blink/trunk/LayoutTests/fullscreen/full-screen-iframe-allowed-video.html?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/Source/web/WebRuntimeFeatures.cpp?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/LayoutTests/fullscreen/video-controls-timeline.html?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/public/web/WebRuntimeFeatures.h?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/Source/web/FullscreenController.cpp?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/LayoutTests/fullscreen/compositor-touch-hit-rects-fullscreen-video-controls.html?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/LayoutTests/fullscreen/video-fail-to-enter-full-screen.html?r1=200193&r2=200192&pathrev=200193
   M http://src.chromium.org/viewvc/blink/trunk/LayoutTests/fullscreen/video-specified-size.html?r1=200193&r2=200192&pathrev=200193

Restore background transparency when leaving fullscreen video

This is a follow up to https://codereview.chromium.org/1254613003 which
removes a usage of OverlayFullscreenVideo that was missed. Previously
the overlay fullscreen runtime feature was used to decide
whether to restore background transparency when leaving fullscreen. Now
it's done whenever the fullscreen element is a video, because we
don't know whether the WebMediaPlayer belonging to the video element used
the overlay mode or not. So conservatively restore the state.

This also adds a setter for ForceOverlayFullscreenVideo to pave the
way for a chromium change to replace usages of OverlayFullscreenVideo
with it.

BUG= 511376 , 507834 
TEST=virtual/android/fullscreen layout tests

Review URL: https://codereview.chromium.org/1278583004
-----------------------------------------------------------------
Comment 20 by w...@chromium.org, Aug 7 2015
Blockedon: chromium:518037
Comment 21 by w...@chromium.org, Aug 10 2015
Blockedon: chromium:518933
Comment 22 by w...@chromium.org, Aug 11 2015
Blockedon: chromium:519669
Project Member Comment 23 by bugdroid1@chromium.org, Aug 12 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fdd36a8329fb60fc5c50612643e310cc40204f07

commit fdd36a8329fb60fc5c50612643e310cc40204f07
Author: watk <watk@chromium.org>
Date: Wed Aug 12 18:44:57 2015

Stop setting --enable-overlay-fullscreen-video on Android

Previously, --enable-overlay-fullscreen-video was used to enable the
blink runtime feature OverlayFullscreenVideo which enabled video to be
efficiently {over,under}-layed on Android. This was set by default on
Android and for layout tests on other platforms. Now the WebMediaPlayer
decides whether overlay fullscreen mode will be used. The blink runtime
feature is renamed to ForceOverlayFullscreenVideo to be used as an
override for running the overlay fullscreen layout tests on non-android
platforms.

This relies on blink changes:
  * https://codereview.chromium.org/1254613003
  * https://codereview.chromium.org/1278583004

A follow up change will remove the OverlayFullscreenVideo runtime
feature and corresponding command line flag.

BUG= 511376 , 507834 
TEST=virtual/android/fullscreen layout tests, manual

Review URL: https://codereview.chromium.org/1286583002

Cr-Commit-Position: refs/heads/master@{#343064}

[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/chromecast/browser/cast_browser_main_parts.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/browser/android/content_startup_flags.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/browser/renderer_host/render_process_host_impl.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/child/runtime_features.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/public/common/content_switches.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/public/common/content_switches.h
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/renderer/media/android/webmediaplayer_android.cc
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/renderer/media/android/webmediaplayer_android.h
[modify] http://crrev.com/fdd36a8329fb60fc5c50612643e310cc40204f07/content/shell/browser/blink_test_controller.cc

Project Member Comment 24 by bugdroid1@chromium.org, Aug 13 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9b16970b9f3b8b28054d32b15099bffe55ec1722

commit 9b16970b9f3b8b28054d32b15099bffe55ec1722
Author: watk <watk@chromium.org>
Date: Thu Aug 13 21:30:30 2015

Add support for H264 and VP9 to AndroidVDA

Previously the AndroidVideoDecodeAccelerator only supported
VP8. Now it has experimental support for VP9 and H264.

BUG= 511375 , 507834 
TEST=Manually play h264 src= video in chrome shell

Review URL: https://codereview.chromium.org/1282083003

Cr-Commit-Position: refs/heads/master@{#343278}

[modify] http://crrev.com/9b16970b9f3b8b28054d32b15099bffe55ec1722/content/common/gpu/media/android_video_decode_accelerator.cc
[modify] http://crrev.com/9b16970b9f3b8b28054d32b15099bffe55ec1722/media/base/video_decoder_config.cc
[modify] http://crrev.com/9b16970b9f3b8b28054d32b15099bffe55ec1722/media/base/video_decoder_config.h
[modify] http://crrev.com/9b16970b9f3b8b28054d32b15099bffe55ec1722/media/ffmpeg/ffmpeg_common.cc

Cc: xhw...@chromium.org
Blockedon: chromium:521731
The EME support for AVDA in the GPU process is tracked by  issue 521731 .
Comment 28 by w...@chromium.org, Aug 20 2015
Blockedon: chromium:523053
Comment 29 by w...@chromium.org, Sep 4 2015
Blockedon: chromium:528417
Comment 30 by w...@chromium.org, Sep 9 2015
Blockedon: chromium:529887
Blockedon: chromium:530681
Blockedon: chromium:531588
Blockedon: chromium:531606
Comment 34 by w...@chromium.org, Sep 14 2015
Blockedon: chromium:531620
Project Member Comment 35 by bugdroid1@chromium.org, Sep 14 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c765882aa08a8dd38847400b05312c0b76aa2334

commit c765882aa08a8dd38847400b05312c0b76aa2334
Author: liberato <liberato@chromium.org>
Date: Mon Sep 14 23:54:19 2015

Begin refactor of AVDA to support zero-copy.

Refactor AndroidVideoDecodeAccelerator into a class that manages the MediaCodec buffers, but delegates handling the SurfaceTexture to BackingStrategy implementations.  The AndroidCopyingBackingStrategy class implements the texture copy into the picture buffer, as it did in previous versions.

BUG= 507834 

Review URL: https://codereview.chromium.org/1313913003

Cr-Commit-Position: refs/heads/master@{#348774}

[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/BUILD.gn
[add] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_copying_backing_strategy.cc
[add] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_copying_backing_strategy.h
[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_video_decode_accelerator.cc
[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_video_decode_accelerator.h
[add] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_video_decode_accelerator_state_provider.h
[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/android_video_decode_accelerator_unittest.cc
[add] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/avda_return_on_failure.h
[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/common/gpu/media/gpu_video_decode_accelerator.cc
[modify] http://crrev.com/c765882aa08a8dd38847400b05312c0b76aa2334/content/content_common.gypi

Comment 36 by bengr@chromium.org, Sep 15 2015
Cc: bengr@chromium.org
Comment 37 by bengr@chromium.org, Sep 15 2015
Blocking: chromium:400789
Cc: siev...@chromium.org
Comment 39 by w...@chromium.org, Sep 18 2015
Blockedon: chromium:533630
Project Member Comment 40 by bugdroid1@chromium.org, Sep 23 2015
Labels: merge-merged-2490
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2775eed637df401a27b84c2beade757908464d80

commit 2775eed637df401a27b84c2beade757908464d80
Author: watk@chromium.org <watk@chromium.org>
Date: Fri Aug 07 20:26:36 2015

Restore background transparency when leaving fullscreen video

This is a follow up to https://codereview.chromium.org/1254613003 which
removes a usage of OverlayFullscreenVideo that was missed. Previously
the overlay fullscreen runtime feature was used to decide
whether to restore background transparency when leaving fullscreen. Now
it's done whenever the fullscreen element is a video, because we
don't know whether the WebMediaPlayer belonging to the video element used
the overlay mode or not. So conservatively restore the state.

This also adds a setter for ForceOverlayFullscreenVideo to pave the
way for a chromium change to replace usages of OverlayFullscreenVideo
with it.

BUG= 511376 , 507834 
TEST=virtual/android/fullscreen layout tests

Review URL: https://codereview.chromium.org/1278583004

git-svn-id: svn://svn.chromium.org/blink/trunk@200193 bbb929c8-8fbe-4397-9dbb-9b2b20218538

[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/LayoutTests/fullscreen/compositor-touch-hit-rects-fullscreen-video-controls.html
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/LayoutTests/fullscreen/full-screen-iframe-allowed-video.html
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/LayoutTests/fullscreen/video-controls-timeline.html
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/LayoutTests/fullscreen/video-fail-to-enter-full-screen.html
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/LayoutTests/fullscreen/video-specified-size.html
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/Source/web/FullscreenController.cpp
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/Source/web/WebRuntimeFeatures.cpp
[modify] http://crrev.com/2775eed637df401a27b84c2beade757908464d80/third_party/WebKit/public/web/WebRuntimeFeatures.h

Blockedon: chromium:535659
Blockedon: chromium:536097
Comment 43 by w...@chromium.org, Oct 5 2015
Blocking: chromium:263654
Blockedon: chromium:540865
Blockedon: chromium:541229
Blockedon: chromium:541659
Comment 47 by w...@chromium.org, Oct 12 2015
Blockedon: chromium:542425
Blocking: chromium:544268
Comment 49 by w...@chromium.org, Oct 19 2015
Labels: Proj-Spitzer
Summary: ☂ Project Spitzer: the desktop media pipeline on Android (was: Support FFmpeg and the desktop media pipeline on Android.)
Project Member Comment 50 by bugdroid1@chromium.org, Oct 21 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/f3510926e755ff862910aa54dcf71921a1380f78

commit f3510926e755ff862910aa54dcf71921a1380f78
Author: xhwang <xhwang@chromium.org>
Date: Wed Oct 21 22:29:34 2015

media: Fix gn build for AVDA related files on Android.

BUG= 507834 
TEST=Compiles with "media_use_ffmpeg = true".

Review URL: https://codereview.chromium.org/1421673002

Cr-Commit-Position: refs/heads/master@{#355412}

[modify] http://crrev.com/f3510926e755ff862910aa54dcf71921a1380f78/content/common/BUILD.gn

Blockedon: chromium:549660
Cc: william....@intel.com
cc:william.xie@intel.com since liberato@ just came across a presentation from him on May about optimizing video rendering :)

https://docs.google.com/presentation/d/1U4cXdqF2uEciCCjQv0cktrNVtNq9EDm_dWVLjKxtJtE/edit#slide=id.p28


Project Member Comment 53 by bugdroid1@chromium.org, Nov 9 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2935ca1f9d285939c1ac7a27dbb52975375c4f7c

commit 2935ca1f9d285939c1ac7a27dbb52975375c4f7c
Author: dalecurtis <dalecurtis@chromium.org>
Date: Mon Nov 09 23:06:15 2015

Add runtime flag for choosing WMPI or WMPA.

Adds a new switch, --enable-unified-media-pipeline to
control usage of WMPI vs WMPA.  Will be needed for finch
experiments as well as local testing.

BUG= 507834 
TEST=builds, command line flag works

Review URL: https://codereview.chromium.org/1431193003

Cr-Commit-Position: refs/heads/master@{#358695}

[modify] http://crrev.com/2935ca1f9d285939c1ac7a27dbb52975375c4f7c/content/browser/renderer_host/render_process_host_impl.cc
[modify] http://crrev.com/2935ca1f9d285939c1ac7a27dbb52975375c4f7c/content/renderer/render_frame_impl.cc
[modify] http://crrev.com/2935ca1f9d285939c1ac7a27dbb52975375c4f7c/media/base/media_switches.cc
[modify] http://crrev.com/2935ca1f9d285939c1ac7a27dbb52975375c4f7c/media/base/media_switches.h

Comment 54 by w...@chromium.org, Nov 13 2015
Blockedon: chromium:555703
Blockedon: chromium:560607
Blocking: chromium:563279
Comment 57 by alokp@chromium.org, Dec 10 2015
Cc: alokp@chromium.org
Blocking: chromium:302729
Blocking: chromium:246757
Blockedon: chromium:570497
Blocking: chromium:419446
Blockedon: chromium:570711
Blockedon: chromium:570754
Blockedon: chromium:570762
Blocking: chromium:455905
Blocking: chromium:570788
Project Member Comment 67 by bugdroid1@chromium.org, Dec 18 2015
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fa2d9d4704677627af5c703fc4c04717f5425e4a

commit fa2d9d4704677627af5c703fc4c04717f5425e4a
Author: xhwang <xhwang@chromium.org>
Date: Fri Dec 18 23:31:18 2015

Switch from build-time to run-time flags for Project Spitzer.

Contributed by dalecurtis@chromium.org.

This puts the desktop media playback pipeline behind a runtime flag
for Android; previously this was a combination of build-time and
run-time flags to avoid introducing a binary size increase before
we were ready for experiments.

This change will result in a ~480kb apk size increase on Android that has been
approved via Chrome Eng Review and klobag@ from the Chrome for Android team.
http://crbug.com/570711 tracks this size increase and plans to reduce it in
the future.

It paves the way for unifying our playback stacks across platforms and brings
previously-missing features and improved security to Android playback. More
technical details can be found in the linked bug below and design doc:
https://goo.gl/qC3OuL

BUG= 507834 , 570711, 570762
TEST=builds with gn/gyp work with unified path.
TBR=grt@chromium.org,dpranke@chromium.org,liberato@chromium.org,creis@chromium.org

Review URL: https://codereview.chromium.org/1541513002

Cr-Commit-Position: refs/heads/master@{#366208}

[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/BUILD.gn
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/DEPS
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/build/common.gypi
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/build/gn_migration.gypi
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/chrome/app/generated_resources.grd
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/chrome/browser/about_flags.cc
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/content/common/gpu/media/android_video_decode_accelerator.cc
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/content/renderer/render_frame_impl.cc
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/BUILD.gn
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/base/BUILD.gn
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/media.gyp
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/media_options.gni
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/media_variables.gypi
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/renderers/default_renderer_factory.cc
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/media/test/BUILD.gn
[modify] http://crrev.com/fa2d9d4704677627af5c703fc4c04717f5425e4a/tools/metrics/histograms/histograms.xml

Blockedon: chromium:574948
Blockedon: chromium:575243
Blockedon: chromium:559236
Blocking: chromium:578296
Blocking: chromium:578311
Blocking: chromium:371500
Blockedon: chromium:580626
Comment 75 by horo@chromium.org, Jan 25 2016
Blocking: chromium:580930
Blockedon: chromium:582170
Blocking: -400789
Blocking: -544268
Blockedon: 584562
Blockedon: -584562
Cc: mlamouri@chromium.org
Blockedon: 584562
Cc: sanfin@chromium.org
Will "brings previously-missing features" apply in both directions? I.e., with a unified pipeline, will all the same media containers/protocols be supported across desktop and Android? I am thinking here of M2TS demuxing support on desktop to enable HLS (or, even better, native HLS support on desktop).

Thanks--
JL

John, we have no plans to support M2TS demuxing on desktop at this point in time, nor native HLS - though this topic did get a lot of discussion, which we can discuss offline (or in another bug - just doesn't belong in this one).
No, desktop will no receive any of the Android codecs/containers and Android will be clamped to desktop codec/container support with the exception of HLS which has very small (<5%) but still non-trivial usage. HLS will continue to use the OS level MediaPlayer so it will not support any of the Spitzer features.
Thanks, guys. Still very cool that Android will use the ffmpeg pipeline for most things.

markdavidscott@, see b/602985, FWIW.

JL

As I understand, this feature is implemented in Chrome mobile 52+? Is the only way to detect this via User-Agent sniffing? As it has quite a significant impact on service workers.

https://github.com/slightlyoff/ServiceWorker/issues/703#issuecomment-223801628
If all goes well it'll be default on for M52, unfortunately there are a few blacklisted devices and it can't be used for HLS. You could test for this by serving a service worker URL to the media player and see if it throws a network error.
Comment 89 by kra...@amazon.com, Jun 30 2016
Are the plans to have the Unified Media Pipeline still on for v52?
If so, are there any plans to address existing video playback regressions?
(See e.g.  crbug.com/624832 )
Comment 90 by w...@chromium.org, Jun 30 2016
We still plan to turn it on for 52. Please file bugs if you know of playback regressions. The one you mentioned is fixed for 52.
Cc: tombergan@google.com mdw@chromium.org kbr@chromium.org ddorwin@chromium.org
 Issue 401795  has been merged into this issue.
Blocking: 240415
Cc: jin.a.y...@intel.com
Status: Fixed
We've been shipping the unified media pipeline since 52, and there's no way to disable it anymore, so this is done!

Deletion of the old media path is tracked by Issue 570711.
Sign in to add a comment