Fill drops with last decoded frame to make them look like freeze at
playback and to keep decoded spatial layers in sync.
Bug: none
Change-Id: I65f7c21100985c22932a1edd441b6c724833c11e
Reviewed-on: https://webrtc-review.googlesource.com/73685
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23076}
If frame of current layer was dropped, pass base frame to decoder if
non_ref_for_inter_layer_pred is set to true.
Bug: none
Change-Id: If7bf5220b74f424106edf74867c9afa8cc2b1ec5
Reviewed-on: https://webrtc-review.googlesource.com/73440
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23074}
This allows to control inter-layer prediction at encoding VP9 SVC.
There are three options:
1. Disabled.
2. Enabled for all pictures.
3. Enabled for key pictures, disabled for others.
Inter-layer prediction is enabled for all pictures by default.
Bug: none
Change-Id: I49fe43d8744c92bec349d815100ba158519f0664
Reviewed-on: https://webrtc-review.googlesource.com/71500
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23049}
For consistency with the VP9 RTP spec which uses term "picture" for set
of frames which belong to the same time instance.
Bug: none
Change-Id: I30e92d5debb008feb58f770b63fe10c2e0029267
Reviewed-on: https://webrtc-review.googlesource.com/72180
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23040}
This deletes the resilienceOn flag in VideoCodecVP8 and VideoCodecVP9.
Instead, the implementations of VP8 and VP9 set resilience mode
internally, based on the configuration of temporal and spatial layers.
The nack_enabled argument to VideoCodecInitializer::SetupCodec becomes
unused with this cl. In a followup, it will be deleted, together with
the corresponding argument to VideoStreamEncoder methods.
An applications which really wants to configure resilience differently
can do that by injecting an EncoderFactory with encoders behaving
as desired.
Bug: webrtc:8830
Change-Id: I9990faf07d3e95c0fb4a56fcc9a56c2005b4a6fa
Reviewed-on: https://webrtc-review.googlesource.com/71380
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23025}
This converts the reserved bit in VP9 RTP payload descriptor into the
flag which indicates whether current frame can be used for prediction
of next spatial layer or not.
VP9 encoder wrapper sets non_ref_for_inter_layer_pred=false for all
frames for now.
Bug: none
Change-Id: I32f68868686475905fb09173cffd2b6e1bedcb7c
Reviewed-on: https://webrtc-review.googlesource.com/71080
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23010}
Since the webrtc_common build target does not have visibility set, we
cannot easily use BitrateAllocation in other parts of Chromium.
This is currently blocking parts of chromium:794608, and I know of other
usage outside webrtc already, so moving it to api/ should be warranted.
Also, since there's some naming confusion and this class is video
specific rename it VideoBitrateAllocation. This also fits with the
standard interface for producing these: VideoBitrateAllocator.
Bug: chromium:794608
Change-Id: I4c0fae40f9365e860c605a76a4f67ecc9b9cf9fe
Reviewed-on: https://webrtc-review.googlesource.com/70783
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22986}
We only support on (formely kResilientStream) and off (formely
kResilienceOff). The third mode, kResilientFrames, was not
implemented.
Bug: None
Change-Id: Ida82f6a33eda9d943ea70bc8ae4e6bddb720b0e8
Reviewed-on: https://webrtc-review.googlesource.com/71481
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22984}
This fixes misprint in the code which calculates target bitrate of a
VP9 spatial layer where "-" was used instead of "+".
Bug: none
Change-Id: I17d76a84d00e453c055c068968d7b276e9c23f51
Reviewed-on: https://webrtc-review.googlesource.com/71663
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22974}
Most are removed entirely, but RtcErrorType, RtpTransceiverDirection, IPAddress and
SocketAddress are kept behind gtest's #ifdef UNIT_TEST.
Bug: webrtc:8982
Change-Id: I36db19891e7d25aeacb08b9a08aa2b4004765e70
Reviewed-on: https://webrtc-review.googlesource.com/64143
Commit-Queue: Jonas Olsson <jonasolsson@webrtc.org>
Reviewed-by: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22916}
With rounding to the nearest the result can exceed the allocated
bitrate.
Bug: none
Change-Id: I0260a1640a1454951ca8e48fd447e047ef0271ee
Reviewed-on: https://webrtc-review.googlesource.com/69982
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22879}
This fixes inconsistency in names of variables and fields which
represent spatial/temporal index of layer:
simulcast_svc_idx -> spatial_idx
spatial_layer_idx -> spatial_idx
temporal_layer_idx -> temporal_idx
Also, this adds printing of spatial/temporal index and target bitrate
to RD report.
Bug: none
Change-Id: Ic4dfdadc57a1577bb3d35d1782a152a9dbef0280
Reviewed-on: https://webrtc-review.googlesource.com/69981
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22869}
Adding SVC rate allocator and layering configurator caused regression
for VP9 non-SVC senders. SVC bitrate limits, which were supposed to
be used only when spatial layering is enabled, are applied when
encoding single spatial layer. E.g. for VP9 360p sender maximum bitrate
is limited to 500kbps.
This fixes the regression. If sender is configured to send VP9 single
layer then codec's bitrate limits are applied to this layer.
Bug: webrtc:9151, chromium:831093
Change-Id: Ia1ae4087155ad7917a3443304a21532f1e68ea65
Reviewed-on: https://webrtc-review.googlesource.com/69813
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Michael Horowitz <mhoro@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22862}
This makes it easier to add new test cases without modifying the actual test class.
Bug: None
Change-Id: I48e4f14e26cd6610678ffb07ce9fd56e6bc1ac4e
Reviewed-on: https://webrtc-review.googlesource.com/69600
Commit-Queue: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22840}
OnRatesUpdated() is called every time the bitrate estimate, or once per
second. However, since we don't want to reconfigure libvpx too often,
just in case it interferes with the rate controller, so
ScreenshareLayers contains a boolean |bitrate_update_| which indicate
if the configuration should be updated on a call to
UpdateConfiguration().
However, it two rate updates happened between two frames, the first of
which changes the rates and second one does not, |bitrate_update_| will
be reset to false and the encoder won't get the desired config.
This CL makes sure we update the configuration iff the rate has changed
at any time since the last call to UpdateConfiguration().
Bug: webrtc:9012
Change-Id: I62af36cffe20ecb7d3f403b3eb11f23a9692d719
Reviewed-on: https://webrtc-review.googlesource.com/69040
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22826}
Set the flag based on coded length of buffered frame which is reset
after picture is encoded and, thus, is equal to zero when encoder
delivers first frame of next picture.
Before this change first_frame_in_picture was set based on index of
spatial layer of encoded frame. This is not right anymore since encoder
can drop base layer but deliver upper layers.
Bug: chromium:828350
Change-Id: I12c7534240de8bc4905f04ff368cc3704720a70b
Reviewed-on: https://webrtc-review.googlesource.com/68561
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22805}
It takes some time for rate controller to adapt to content. Quality of first
frames is usually worse than quality of following frames. It makes sense to
exclude first frames from analysis and, thus, avoid negative affect of this
interval on overall results.
Bug: none
Change-Id: Ib0a258889750cf794c7d6fdff26af958f7bbe48a
Reviewed-on: https://webrtc-review.googlesource.com/66100
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Michael Horowitz <mhoro@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22782}
In order to handle per-layer frame dropping both VP9 encoder wrapper
and RTP packetizer were modified.
- Encoder wrapper buffers last encoded frame and passes it to
packetizer after frame of next layer is encoded or encoding of
superframe is finished.
- Encoder wrapper sets end_of_superframe flag on last encoded frame of
superframe before passing it to packetizer.
- If end_of_superframe is True then packetizer sets marker bit on last
packet of frame.
Bug: webrtc:9066
Change-Id: I1d45319fbe6bc63d01721ea67bfb7440d4c29275
Reviewed-on: https://webrtc-review.googlesource.com/65540
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22722}
Measure time spent in frame encode callback, accumulate it for layers
and subtract it from measured encode time of next layer frame.
Bug: none
Change-Id: Ifc3baae2f9a49913a55a7de2de9507102edd0295
Reviewed-on: https://webrtc-review.googlesource.com/65981
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22720}
The configurator decides number of spatial layers, their resolution
and bitrate thresholds based on given input resolution and maximum
number of spatial layers.
The allocator distributes available bitrate across spatial and
temporal layers. If there is not enough bitrate to provide acceptable
quality for all spatial layers allocator disables enhancement layers
one by one until the condition is met or number of layers is reduced
to one.
VP9 SVC related unit tests have been updated. Input resolution and
bitrate in these tests have been increased to the level enough to
provide desirable number of spatial layers.
Bug: webrtc:8518
Change-Id: I9df790920227c7f7dd4d42a50a856c22f0f4389b
Reviewed-on: https://webrtc-review.googlesource.com/60340
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Michael Horowitz <mhoro@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22672}
There is no need to use real video as input for encoder in unit tests.
Using generator simplifies testing on mobile devices (no need to upload
files to device).
Bug: none
Change-Id: Ic48609cc6f8eecf90d9956edfdd33135be949398
Reviewed-on: https://webrtc-review.googlesource.com/64526
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22648}
Build superframe out of the nearest non-dropped base layer and current layer.
Bug: none
Change-Id: I26720ea6de44f27046208b220d03942cd2a3d6c7
Reviewed-on: https://webrtc-review.googlesource.com/64921
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22631}
It would be nice to also delete the fields from CodecSpecificInfo,
but those fields are used on the receive side.
Bug: webrtc:8830
Change-Id: I1a3f13ea2c024cbd73b33fd9dd58e531d3576a55
Reviewed-on: https://webrtc-review.googlesource.com/64780
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22625}
Have not figured out why this metric regressed, but submitting
this CL now to unblock Chromium roll into WebRTC.
Bug: webrtc:9057
Change-Id: I808ad194e1c9107d644a25502a55a7c6fddca7aa
Reviewed-on: https://webrtc-review.googlesource.com/64527
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Commit-Queue: Oleh Prypin <oprypin@webrtc.org>
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22600}
As the rate allocation has been moved into entirely into
SimulcastRateAllocator, and the listeners are thus no longer needed,
this class doesn't fill any other purpose than to determine if
ScreenshareLayers or TemporalLayers should be created for a given
simulcast stream. This can however be done just from looking at the
VideoCodec instance, so changing this into a static factory method.
Due to dependencies from upstream projects, keep the class name and
field in VideoCodec around for now.
Bug: webrtc:9012
Change-Id: I028fe6b2a19e0d16b35956cc2df01dcf5bfa7979
Reviewed-on: https://webrtc-review.googlesource.com/63264
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22529}
This CL moves all temporal layer rate allocation from
DefaultTemporalLayers and ScreenshareLayers into SimulcastRateAllocator.
This means we don't need an extra call-out to the TemporalLayers
interface to get the last allocation, which simplifies the code path a
lot.
It also paves the wave for removing the TemporalLayersFactory interface
(in a separate cl), which will further simplify the ownership model.
Bug: webrtc:9012
Change-Id: I6540b1848efa1a136dce449f13902ad479d5ee37
Reviewed-on: https://webrtc-review.googlesource.com/62420
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22502}
It previously owned only the picture id and only in the
WebRTC-VP8-Forced-Fallback-Encoder-v2 experiment.
Moving responsibility to PayloadRouter ensures that both
picture id and tl0 idx are continuous over codec changes,
as required by the specs for VP8 and VP9 over RTP.
Bug: webrtc:8830
Change-Id: Ie77356dfec6d1e372b6970189e4c3888451920e6
Reviewed-on: https://webrtc-review.googlesource.com/61640
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22448}
This work is in preparation for refactoring the TemporalLayers api.
Bug: webrtc:9012
Change-Id: I01908ee034fb79996e687ff72d10178acf102321
Reviewed-on: https://webrtc-review.googlesource.com/61781
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22445}
