It used up to 3 threads for QVGA on Android before. This change disables Android-specific code path in NumberOfThreads() and uses the generic settings, which configure 1 thread for resolutions <=VGA, instead. The change is guarded by a killswitch.
For reference, frame encode time for VGA 512kbps using 1 thread on Pixel 2 (7 years old device; SD835) is ~5.5ms: https://chromeperf.appspot.com/report?sid=6e80c701ef6ff0d008a299fb122a16f0d2600ddfcd9981d3d75cd722c92b2869
Bug: webrtc:15828, b/316494683
Change-Id: I0e9571ede64c6cb77d529d21ccb0310ccb8bfdaf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/337601
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41770}
Example: "WebRTC-LibvpxVp9Encoder-SvcFrameDropConfig/Enabled,layer_drop_mode:1,max_consec_drop:7/"
It is only possible to enable LAYER_DROP (layer_drop_mode=1) for now. All other modes are ignored. Max consecutive frame drops (max_consec_drop) value from the field is always applied if the field trial is enabled.
LAYER_DROP requires flexible mode (is_flexible_mode_=true) which can be enabled by means of WebRTC-Vp9InterLayerPred: https://source.chromium.org/chromium/chromium/src/+/main:third_party/webrtc/media/engine/webrtc_video_engine.cc;l=976
Bug: webrtc:15827, b/320629637
Change-Id: I9c4d4838b11547e608d863198b109cb1485902d6
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/335041
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41755}
The division by 2 has been accidentally removed in https://webrtc-review.googlesource.com/c/src/+/76921
The code and comment are out of sync now.
Bug: None
Change-Id: If43a40461878ffe58dd9ed0ab8a6244ad79c4f6b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/336283
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Auto-Submit: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41627}
This allows using different encoder and decoder implementations in a test. For example, to encode with SW encoder and to decode with HW decoder or vice versa.
Bug: webrtc:14852
Change-Id: Ic100cba2158fb6311b84a54a0831f2a0dcff9270
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/335300
Auto-Submit: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Commit-Queue: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41571}
This is a reland of commit 63d03f586b
Original change's description:
> Unify access to SDP codec parameters
>
> which come from the a=fmtp:<pt> lines in the SDP and were used as either
> std::map<std::string, std:string>
> with three aliases,
> cricket::CodecParameterMap
> SdpAudioFormat::Parameters
> SdpVideoFormat::Parameters
>
> Use webrtc::CodecParameterMap in all places.
>
> BUG=None
>
> Change-Id: If47692bde7347834c349c6539b43309d8770e67b
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/330420
> Reviewed-by: Florent Castelli <orphis@webrtc.org>
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Commit-Queue: Philipp Hancke <phancke@microsoft.com>
> Cr-Commit-Position: refs/heads/main@{#41375}
Bug: None
Change-Id: I5f8f45688df232eb37b12fa3e56a893a1c754e17
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/331402
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#41467}
which come from the a=fmtp:<pt> lines in the SDP and were used as either
std::map<std::string, std:string>
with three aliases,
cricket::CodecParameterMap
SdpAudioFormat::Parameters
SdpVideoFormat::Parameters
Use webrtc::CodecParameterMap in all places.
BUG=None
Change-Id: If47692bde7347834c349c6539b43309d8770e67b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/330420
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#41375}
This is a reland of commit 496893e89e
Original change's description:
> Added an encode/decode test parameterizable via command line
>
> This enables testing different settings without updating code and rebuilding the test binary. Example of command:
>
> video_codec_perf_tests --gtest_also_run_disabled_tests --gtest_filter=*EncodeDecode --encoder=libaom-av1 --decoder=dav1d --scalability_mode=L1T3 --bitrate_kbps=100,200,300 --framerate_fps=30 --write_csv
>
> Also added writing per-frame stats to a CSV. It is more convenient to work with CSV than to parse metrics proto.
>
> Bug: webrtc:14852
> Change-Id: I1b3970f7ffa88c016133197aff585de5bc4e35c6
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327600
> Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
> Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
> Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#41179}
Bug: webrtc:14852
Change-Id: Iccb9af8bf6a6c37704bc58b6e57238b55761b079
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327781
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41194}
This reverts commit 496893e89e.
Reason for revert: Breaks https://ci.chromium.org/ui/p/chromium/builders/webrtc.fyi/WebRTC%20Chromium%20FYI%20ios-device/16103/overview
Original change's description:
> Added an encode/decode test parameterizable via command line
>
> This enables testing different settings without updating code and rebuilding the test binary. Example of command:
>
> video_codec_perf_tests --gtest_also_run_disabled_tests --gtest_filter=*EncodeDecode --encoder=libaom-av1 --decoder=dav1d --scalability_mode=L1T3 --bitrate_kbps=100,200,300 --framerate_fps=30 --write_csv
>
> Also added writing per-frame stats to a CSV. It is more convenient to work with CSV than to parse metrics proto.
>
> Bug: webrtc:14852
> Change-Id: I1b3970f7ffa88c016133197aff585de5bc4e35c6
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327600
> Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
> Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
> Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#41179}
Bug: webrtc:14852
Change-Id: Ifdce738058c3e3fc7c76886939add2813405cae7
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327722
Owners-Override: Christoffer Jansson <jansson@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Christoffer Jansson <jansson@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#41183}
This enables testing different settings without updating code and rebuilding the test binary. Example of command:
video_codec_perf_tests --gtest_also_run_disabled_tests --gtest_filter=*EncodeDecode --encoder=libaom-av1 --decoder=dav1d --scalability_mode=L1T3 --bitrate_kbps=100,200,300 --framerate_fps=30 --write_csv
Also added writing per-frame stats to a CSV. It is more convenient to work with CSV than to parse metrics proto.
Bug: webrtc:14852
Change-Id: I1b3970f7ffa88c016133197aff585de5bc4e35c6
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327600
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41179}
* Pass codec factories to the video codec tester instead of creating and wrapping codecs into a tester-specific wrappers in video_codec_test.cc. The motivation for this change is to simplify the tests by moving complexity to the tester.
* Merge codec stats and analysis into the tester and move the tester. The merge fixes circular deps issues. Modularization is not strictly needed for testing framework like the video codec tester. It is still possible to unit test underlaying modules with rather small overhead.
* Move the video codec tester from api/ to test/. test/ is accessible from outside of WebRTC which enables reusing the tester in downstream projects.
Test output ~matches before and after this refactoring. There is a small difference that is caused by changes in qpMax: 63 -> 56 (kDefaultVideoMaxQpVpx). 56 is what WebRTC uses by default for VPx/AV1 encoders.
Bug: webrtc:14852
Change-Id: I762707b7144fcff870119ad741ebe7091ea109ba
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/327260
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41144}
Before this change bitrate limits for VP9 single spatial layer case were set in VideoCodecInitializer. Move this logic to GetVp9SvcConfig. This simplifies replication of WebRTC behaviour in codec level tests. The similar AV1 logic sits in SetAv1SvcConfig, not VideoCodecInitializer.
Bug: webrtc:14852
Change-Id: Ie7202ec880d0e4b903e7265721eeef9b3920f21a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/324286
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40992}
Sometimes OpenH264 returns a key frame even though we have not
requested one. However, SVC controller does not know about this
and will not reset its state. Since we are comparing expected tid
from SVC controller with actual tid from OpenH264, and drop frames
if they do not match, that causes a missing frame.
This CL resets the SVC controller state on key frames, ensuring
that it accurately maintains its state and does not drop frames.
Also, changes the message of the error log to be more descriptive.
Now, it will print the expected tid and actual tid.
Bug: webrtc:14877
Change-Id: I6c9e7532b2478773f03e5707bf7a1ca56e4f7b99
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/324001
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40972}
This field trial is configured via command line flag, so may use flag system directly, reducing dependency on global field trial string.
Bug: webrtc:7101, webrtc:10335
Change-Id: I1e48e0e3fdc251b73a375c6d7f1a46fa4f8a179b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/322624
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40897}
ARM-specific settings were intended to be used on mobile ARM devices which may not be powerful enough. But the settings were also applied to ARM-based Macs. This changes restricts ARM-specific settings to Android and iOS platforms.
Bug: none
Change-Id: I68764b4c0679db07399bba5923f4a6be89c5ad80
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/321861
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Jerome Jiang <jianj@google.com>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40884}
This is a reland of commit 0d4b350006
Patchset 1 is the original CL. Patchset 2 contains a small tweak of the target bitrate in the unit test, in order to make in less susceptible to flakiness on runtime environments running a slightly different libvpx.
Original change's description:
> Add mitigation for very long frame drop gaps with vp8
>
> Bug: webrtc:15530
> Change-Id: I11f5e3f31f71301700dbff3fc9285236160bee45
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/322320
> Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
> Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
> Auto-Submit: Erik Språng <sprang@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#40866}
Bug: webrtc:15530
Change-Id: I096b7d952286f7f53852d1ca70aea398b2747784
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/322540
Auto-Submit: Erik Språng <sprang@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40874}
To avoid name collision with Timestamp type,
To avoid confusion with capture time represented as Timestamp
Bug: webrtc:9378
Change-Id: I8438a9cf4316e5f81d98c2af9dc9454c21c78e70
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/320601
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40796}
The main goal of this change is to disable the quality scaler when multiple spatial layers are used.
Bug: b/295129711
Change-Id: I25e0b7440a8c2adee3e97720a1e0ee5e0a914334
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/319181
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40709}
Remove EncodedFrame::MissingFrame, as it was always false in actual
in-use code anyway, and remove usages of the Decode missing_frames param
within WebRTC. Uses/overrides in other projects will be cleaned up
shortly, allowing that variant to be removed from the interface.
Bug: webrtc:15444
Change-Id: Id299d82e441a351deff81c0f2812707a985d23d8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/317802
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Auto-Submit: Tony Herre <herre@google.com>
Commit-Queue: Tony Herre <herre@google.com>
Cr-Commit-Position: refs/heads/main@{#40662}
Use 8 threads for > 720p
Use 4 tile columns and 2 tile rows for 8 threads
Use 2 tile columns and 2 tile rows for 4 threads
For VGA, 2 tile_col x 2 tile_row configuration has
- ~9% speedup over 4 tile_col x 1 tile_row
- ~5% speedup over 1 tile_col x 4 tile_row
Bug: None
Change-Id: I3c1ea948437aece7c6734ce25351158cbdf0a15b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/307880
Commit-Queue: Jerome Jiang <jianj@google.com>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40237}
The preferred method to create codecs is to use the function
cricket::CreateAudioCodec or cricketCreateVideoCodec.
Empty codec objects are deprecated and should be replaced
with alternatives such as methods returning an
absl::optional object instead.
Bug: webrtc:15214
Change-Id: I7fe40f64673cd407830dbbb0e541b85a3aee93aa
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/307521
Commit-Queue: Florent Castelli <orphis@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40226}
Empty codec objects do not make sense. Instead of creating an empty
object to be used as a placeholder in the API, at least create a
video codec with the right name.
Bug: webrtc:15214
Change-Id: I705d9d1361f353fe5dc538a6fe972c8a346f1247
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/307221
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Florent Castelli <orphis@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40218}
Most of the usage of the H264Encoder::Create(codec) method passes a
simple codec with just the H264 codec name. This simplified the call
sites in many places and removes references to the codec types.
Bug: webrtc:15214
Change-Id: I4039c0be4ce6e3147c14c7853df4635f344b7d70
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/307222
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Florent Castelli <orphis@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#40214}
