For them implemeted upscaling in libyuv metrics calculation.
Updated maximum number of SL in vp9 encoder to 3.
Refactored names of some fields in Video_quality_check analyzer.
BUG=webrtc:7095
Review-Url: https://codereview.webrtc.org/2681683003
Cr-Commit-Position: refs/heads/master@{#16625}
Reason for revert:
Perf test broke as it made assumptions that quality scaling was turned off. This turns out not to be the case. Fixed by turning quality scaling off for the tests.
Original issue's description:
> Revert of Drop frames until specified bitrate is achieved. (patchset #12 id:240001 of https://codereview.webrtc.org/2630333002/ )
>
> Reason for revert:
> due to failures on perf tests (not on perf stats, but fails running due to dcheck failures), see e.g., https://build.chromium.org/p/client.webrtc.perf/builders/Android32%20Tests%20(K%20Nexus5)
>
> Original issue's description:
> > Drop frames until specified bitrate is achieved.
> >
> > This CL fixes a regression introduced with the new quality scaler
> > where the video would no longer start in a scaled mode. This CL adds
> > code that compares incoming captured frames to the target bitrate,
> > and if they are found to be too large, they are dropped and sinkWants
> > set to a lower resolution. The number of dropped frames should be low
> > (0-4 in most cases) and should not introduce a noticeable delay, or
> > at least should be preferrable to having the first 2-4 seconds of video
> > have very low quality.
> >
> > BUG=webrtc:6953
> >
> > Review-Url: https://codereview.webrtc.org/2630333002
> > Cr-Commit-Position: refs/heads/master@{#16391}
> > Committed: 83399caec5
>
> TBR=perkj@webrtc.org,sprang@webrtc.org,stefan@webrtc.org,kthelgason@webrtc.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=webrtc:6953
>
> Review-Url: https://codereview.webrtc.org/2666303002
> Cr-Commit-Position: refs/heads/master@{#16395}
> Committed: 35fc2aa82fTBR=perkj@webrtc.org,sprang@webrtc.org,stefan@webrtc.org,minyue@webrtc.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=webrtc:6953
Review-Url: https://codereview.webrtc.org/2675223002
Cr-Commit-Position: refs/heads/master@{#16473}
Reason for revert:
Downstream project relied on changed struct.
Transition made possible by https://codereview.webrtc.org/2655243006/.
Original issue's description:
> Revert of Make RTX pt/apt reconfigurable by calling WebRtcVideoChannel2::SetRecvParameters. (patchset #7 id:160001 of https://codereview.webrtc.org/2646073004/ )
>
> Reason for revert:
> Breaks internal downstream project.
>
> Original issue's description:
> > Make RTX pt/apt reconfigurable by calling WebRtcVideoChannel2::SetRecvParameters.
> >
> > Prior to this CL, received RTX (associated) payload types were only configured
> > when WebRtcVideoChannel2::AddRecvStream was called. In the same method, the RTX
> > SSRC was set up.
> >
> > After this CL, the RTX (associated) payload types are set in
> > WebRtcVideoChannel2::SetRecvParameters, which is the appropriate place to set
> > them. The RTX SSRC is still set in WebRtcVideoChannel2::AddRecvStream, since
> > that is the code path that sets other SSRCs.
> >
> > As part of this fix, the VideoReceiveStream::Config::Rtp struct is changed.
> > We remove the possibility for each video payload type to have an associated
> > specific RTX SSRC. Although the config previously allowed for this, all payload
> > types always had the same RTX SSRC set, and the underlying RtpPayloadRegistry
> > did not support multiple SSRCs. This change to the config struct should thus not
> > have any functional impact. The change does however affect the RtcEventLog, since
> > that is used for storing the VideoReceiveStream::Configs. For simplicity,
> > this CL does not change the event log proto definitions, instead duplicating
> > the serialized RTX SSRCs such that they fit in the existing proto definition.
> >
> > BUG=webrtc:7011
> >
> > Review-Url: https://codereview.webrtc.org/2646073004
> > Cr-Commit-Position: refs/heads/master@{#16302}
> > Committed: fe2bef39cd
>
> TBR=stefan@webrtc.org,magjed@webrtc.org,terelius@webrtc.org,brandtr@webrtc.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=webrtc:7011
>
> Review-Url: https://codereview.webrtc.org/2649323010
> Cr-Commit-Position: refs/heads/master@{#16307}
> Committed: e4974953ceTBR=stefan@webrtc.org,magjed@webrtc.org,terelius@webrtc.org,kjellander@webrtc.org,kjellander@google.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
# NOTREECHECKS=true
# NOTRY=true
BUG=webrtc:7011
Review-Url: https://codereview.webrtc.org/2654163006
Cr-Commit-Position: refs/heads/master@{#16322}
Reason for revert:
Breaks internal downstream project.
Original issue's description:
> Make RTX pt/apt reconfigurable by calling WebRtcVideoChannel2::SetRecvParameters.
>
> Prior to this CL, received RTX (associated) payload types were only configured
> when WebRtcVideoChannel2::AddRecvStream was called. In the same method, the RTX
> SSRC was set up.
>
> After this CL, the RTX (associated) payload types are set in
> WebRtcVideoChannel2::SetRecvParameters, which is the appropriate place to set
> them. The RTX SSRC is still set in WebRtcVideoChannel2::AddRecvStream, since
> that is the code path that sets other SSRCs.
>
> As part of this fix, the VideoReceiveStream::Config::Rtp struct is changed.
> We remove the possibility for each video payload type to have an associated
> specific RTX SSRC. Although the config previously allowed for this, all payload
> types always had the same RTX SSRC set, and the underlying RtpPayloadRegistry
> did not support multiple SSRCs. This change to the config struct should thus not
> have any functional impact. The change does however affect the RtcEventLog, since
> that is used for storing the VideoReceiveStream::Configs. For simplicity,
> this CL does not change the event log proto definitions, instead duplicating
> the serialized RTX SSRCs such that they fit in the existing proto definition.
>
> BUG=webrtc:7011
>
> Review-Url: https://codereview.webrtc.org/2646073004
> Cr-Commit-Position: refs/heads/master@{#16302}
> Committed: fe2bef39cdTBR=stefan@webrtc.org,magjed@webrtc.org,terelius@webrtc.org,brandtr@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:7011
Review-Url: https://codereview.webrtc.org/2649323010
Cr-Commit-Position: refs/heads/master@{#16307}
Prior to this CL, received RTX (associated) payload types were only configured
when WebRtcVideoChannel2::AddRecvStream was called. In the same method, the RTX
SSRC was set up.
After this CL, the RTX (associated) payload types are set in
WebRtcVideoChannel2::SetRecvParameters, which is the appropriate place to set
them. The RTX SSRC is still set in WebRtcVideoChannel2::AddRecvStream, since
that is the code path that sets other SSRCs.
As part of this fix, the VideoReceiveStream::Config::Rtp struct is changed.
We remove the possibility for each video payload type to have an associated
specific RTX SSRC. Although the config previously allowed for this, all payload
types always had the same RTX SSRC set, and the underlying RtpPayloadRegistry
did not support multiple SSRCs. This change to the config struct should thus not
have any functional impact. The change does however affect the RtcEventLog, since
that is used for storing the VideoReceiveStream::Configs. For simplicity,
this CL does not change the event log proto definitions, instead duplicating
the serialized RTX SSRCs such that they fit in the existing proto definition.
BUG=webrtc:7011
Review-Url: https://codereview.webrtc.org/2646073004
Cr-Commit-Position: refs/heads/master@{#16302}
This CL adds an RTP module to FlexfecReceiveStreamImpl, and wires it up
to send RTCP RRs. It further makes some methods take const refs instead
of values, to make it more clear where packet copies are made. This
change reduces the number of copies by one, for the case when media
packets are added to the FlexFEC receiver.
The end-to-end test is modified to check for RTCP RRs being sent.
Part of this modification involves some indentation changes, and the
diff thus looks bigger than it logically is.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2625633003
Cr-Commit-Position: refs/heads/master@{#16106}
The existence of FlexfecConfig is due to a naive design. Now when it
is not used on the receiving side (see https://codereview.webrtc.org/2542413002),
it is time to remove it from the sending side as well.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2621573002
Cr-Commit-Position: refs/heads/master@{#16097}
That object will be used when we enable RTCP reporting from FlexfecReceiveStream.
Other related changes:
- Stop using FlexfecConfig (from config.h) at receive side in WebRtcVideoEngine2.
- Add a IsCompleteAndEnabled() method to FlexfecReceiveStream::Config, to be
used in WebRtcVideoEngine2.
- Centralize the construction of the FlexfecReceiveStream::Config in unit tests.
This will make future additions to the unit tests cleaner.
- Simplify setup for receiving FlexFEC in VideoQualityTest.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2589713003
Cr-Commit-Position: refs/heads/master@{#16059}
Problem fixed: RTP header extensions were not properly set in tests.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2593963003
Cr-Commit-Position: refs/heads/master@{#15741}
Prior to this change, the receiver_time metric had huge outliers
whenever FlexFEC was enabled. This was due to a measurement problem,
where the time of the incoming packet was incorrectly set to zero.
This happened for packets that were lost in transit, but recovered
through FEC.
This CL fixes this problem by simply not recording samples where the
incoming packet time is undefined. The CL also removes the possibility
of timestamp collisions in the data structures.
TESTED=Ran './webrtc_perf_tests --gtest_filter="*ForemanCifPlr5H264Flexfec*" | grep receiver_time' locally 10 times, without experiencing any outliers.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2596793002
Cr-Commit-Position: refs/heads/master@{#15735}
Reason for revert:
Crashes perf tests, e.g.,
./out/Debug/webrtc_perf_tests --gtest_filter='FullStackTest.ScreenshareSlidesVP8_2TL_VeryLossyNet'
dies with an assert related to rtc::Optional.
Original issue's description:
> Delete VideoFrame default constructor, and IsZeroSize method.
>
> This ensures that the video_frame_buffer method never can return a
> null pointer.
>
> BUG=webrtc:6591
>
> Committed: https://crrev.com/bfcf561923a42005e4c7d66d8e72e5932155f997
> Cr-Commit-Position: refs/heads/master@{#15574}
TBR=magjed@webrtc.org,stefan@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6591
Review-Url: https://codereview.webrtc.org/2574123002
Cr-Commit-Position: refs/heads/master@{#15597}
This ensures that the video_frame_buffer method never can return a
null pointer.
BUG=webrtc:6591
Review-Url: https://codereview.webrtc.org/2541863002
Cr-Commit-Position: refs/heads/master@{#15574}
There's no longer any need to make the two arguments have the same
signedness, so we can remove a bunch of superfluous (and sometimes
dangerous) casts.
It turned out I also had to fix the safe_cmp functions to properly handle
enums that are implicitly convertible to integers.
NOPRESUBMIT=true
BUG=webrtc:6645
Review-Url: https://codereview.webrtc.org/2534683002
Cr-Commit-Position: refs/heads/master@{#15281}
There's no longer any need to make the two arguments have the same
signedness, so we can drop the "u" suffix on literal integer
arguments.
NOPRESUBMIT=true
BUG=webrtc:6645
Review-Url: https://codereview.webrtc.org/2535593002
Cr-Commit-Position: refs/heads/master@{#15280}
Pass the selected cricket::VideoCodec to H264EncoderImpl::H264EncoderImpl. The cricket::VideoCodec contains relevant information for H264 about selected profile and packetization mode.
BUG=chromium:600254,webrtc:6402, webrtc:6337
Review-Url: https://codereview.webrtc.org/2474993002
Cr-Commit-Position: refs/heads/master@{#15270}
Reason for revert:
The WebRtcBrowserTest.NegotiateUnsupportedVideoCodec test has been fixed in Chromium with the following change:
function removeVideoCodec(offerSdp) {
- offerSdp = offerSdp.replace('a=rtpmap:100 VP8/90000\r\n',
- 'a=rtpmap:100 XVP8/90000\r\n');
+ offerSdp = offerSdp.replace(/a=rtpmap:(\d+)\ VP8\/90000\r\n/,
+ 'a=rtpmap:$1 XVP8/90000\r\n');
return offerSdp;
}
Original issue's description:
> Revert of Stop using hardcoded payload types for video codecs (patchset #6 id:210001 of https://codereview.webrtc.org/2493133002/ )
>
> Reason for revert:
> Breaks chromium.fyi test:
> WebRtcBrowserTest.NegotiateUnsupportedVideoCodec
>
> Original issue's description:
> > Stop using hardcoded payload types for video codecs
> >
> > This CL stops using hardcoded payload types for different video codecs
> > and will dynamically assign them payload types incrementally from 96 to
> > 127 instead.
> >
> > This CL:
> > * Replaces 'std::vector<VideoCodec> DefaultVideoCodecList()' in
> > webrtcvideoengine2.cc with an explicit WebRtcVideoEncoderFactory for
> > internally supported software codecs instead. The purpose is to
> > streamline the payload type assignment in webrtcvideoengine2.cc which
> > will now have two encoder factories of the same
> > WebRtcVideoEncoderFactory type; one internal and one external.
> > * Removes webrtc::VideoEncoder::EncoderType and use cricket::VideoCodec
> > instead.
> > * Removes 'static VideoEncoder* Create(EncoderType codec_type)' and
> > moves the create function to the internal encoder factory instead.
> > * Removes video_encoder.cc. webrtc::VideoEncoder is now just an
> > interface without any static functions.
> > * The function GetSupportedCodecs in webrtcvideoengine2.cc unifies
> > the internal and external codecs and assigns them payload types
> > incrementally from 96 to 127.
> > * Updates webrtcvideoengine2_unittest.cc and removes assumptions about
> > what payload types will be used.
> >
> > BUG=webrtc:6677,webrtc:6705
> > R=hta@webrtc.org, ossu@webrtc.org, stefan@webrtc.org
> >
> > Committed: https://crrev.com/42043b95872b51321f508bf255d804ce3dff366b
> > Cr-Commit-Position: refs/heads/master@{#15135}
>
> TBR=hta@webrtc.org,stefan@webrtc.org,ossu@webrtc.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=webrtc:6677,webrtc:6705
>
> Committed: https://crrev.com/eacbaea920797ff751ca83050d140821f5055591
> Cr-Commit-Position: refs/heads/master@{#15140}
TBR=hta@webrtc.org,stefan@webrtc.org,ossu@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6677,webrtc:6705
Review-Url: https://codereview.webrtc.org/2511933002
Cr-Commit-Position: refs/heads/master@{#15148}
Reason for revert:
Breaks chromium.fyi test:
WebRtcBrowserTest.NegotiateUnsupportedVideoCodec
Original issue's description:
> Stop using hardcoded payload types for video codecs
>
> This CL stops using hardcoded payload types for different video codecs
> and will dynamically assign them payload types incrementally from 96 to
> 127 instead.
>
> This CL:
> * Replaces 'std::vector<VideoCodec> DefaultVideoCodecList()' in
> webrtcvideoengine2.cc with an explicit WebRtcVideoEncoderFactory for
> internally supported software codecs instead. The purpose is to
> streamline the payload type assignment in webrtcvideoengine2.cc which
> will now have two encoder factories of the same
> WebRtcVideoEncoderFactory type; one internal and one external.
> * Removes webrtc::VideoEncoder::EncoderType and use cricket::VideoCodec
> instead.
> * Removes 'static VideoEncoder* Create(EncoderType codec_type)' and
> moves the create function to the internal encoder factory instead.
> * Removes video_encoder.cc. webrtc::VideoEncoder is now just an
> interface without any static functions.
> * The function GetSupportedCodecs in webrtcvideoengine2.cc unifies
> the internal and external codecs and assigns them payload types
> incrementally from 96 to 127.
> * Updates webrtcvideoengine2_unittest.cc and removes assumptions about
> what payload types will be used.
>
> BUG=webrtc:6677,webrtc:6705
> R=hta@webrtc.org, ossu@webrtc.org, stefan@webrtc.org
>
> Committed: https://crrev.com/42043b95872b51321f508bf255d804ce3dff366b
> Cr-Commit-Position: refs/heads/master@{#15135}
TBR=hta@webrtc.org,stefan@webrtc.org,ossu@webrtc.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=webrtc:6677,webrtc:6705
Review-Url: https://codereview.webrtc.org/2513633002
Cr-Commit-Position: refs/heads/master@{#15140}
This CL stops using hardcoded payload types for different video codecs
and will dynamically assign them payload types incrementally from 96 to
127 instead.
This CL:
* Replaces 'std::vector<VideoCodec> DefaultVideoCodecList()' in
webrtcvideoengine2.cc with an explicit WebRtcVideoEncoderFactory for
internally supported software codecs instead. The purpose is to
streamline the payload type assignment in webrtcvideoengine2.cc which
will now have two encoder factories of the same
WebRtcVideoEncoderFactory type; one internal and one external.
* Removes webrtc::VideoEncoder::EncoderType and use cricket::VideoCodec
instead.
* Removes 'static VideoEncoder* Create(EncoderType codec_type)' and
moves the create function to the internal encoder factory instead.
* Removes video_encoder.cc. webrtc::VideoEncoder is now just an
interface without any static functions.
* The function GetSupportedCodecs in webrtcvideoengine2.cc unifies
the internal and external codecs and assigns them payload types
incrementally from 96 to 127.
* Updates webrtcvideoengine2_unittest.cc and removes assumptions about
what payload types will be used.
BUG=webrtc:6677,webrtc:6705
R=hta@webrtc.org, ossu@webrtc.org, stefan@webrtc.org
Review URL: https://codereview.webrtc.org/2493133002 .
Cr-Commit-Position: refs/heads/master@{#15135}
This is a refactoring change in preparation for enabling AudioMixer
with the goal to have a small CL as possible for passing audio through
the new audio mixer in WebRTC. The dependent CL https://codereview.webrtc.org/2436033002/
enables the mixer.
An object of class AudioState is shared across different webrtc audio
connections. It is created in tests and in
WebRTCVoiceEngine. AudioState is constructed by passing a Config
struct, where one argument is scoped_refptr<AudioMixer>.
Populating this field has now been mandatory. Tests and
WebRTCVoiceEngine create and pass either a AudioMixerImpl.
WebRTCVoiceEngine passes a real AudioMixer, which is
currently unused.
An alternative would have tests pass a mocked audio mixer. We
chose not to do that, because we believe that tests should use
the real thing unless there are reasons against it. Construction
time is not an issue, because the real mixer is relatively
lightweight.
We couldn't find a way to test any mixer-related changes in AudioState
before the mixes is connected. The next dependent CL
https://codereview.webrtc.org/2436033002/ contains unit tests for
mixer usage.
BUG=webrtc:6346
Review-Url: https://codereview.webrtc.org/2469743002
Cr-Commit-Position: refs/heads/master@{#15134}
Will be used by full stack tests and video_loopback.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2500373002
Cr-Commit-Position: refs/heads/master@{#15114}
This is needed for the following coming tests: VideoSendStream, end-to-end,
full stack, and video_loopback.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2500943002
Cr-Commit-Position: refs/heads/master@{#15087}
After the removal of the RED/RTX workaround, we now need to explicitly
enable RTX for RED. Prior to the removal of the workaround, RED over RTX
was implicitly enabled whenever media over RTX was enabled.
BUG=webrtc:6650
Review-Url: https://codereview.webrtc.org/2493723002
Cr-Commit-Position: refs/heads/master@{#15061}
The unit was kbps but the one default use of it is in bps. The inconsistency should be fixed.
BUG=webrtc:6670
Review-Url: https://codereview.webrtc.org/2247213005
Cr-Commit-Position: refs/heads/master@{#14955}
This removes the VideoSendStream::LoadObserver interface and the implementation in WebrtcVideoSendStream and replace it with VideoSinkWants through the VideoSourceInterface.
To do that that, some stats for CPU adaptation is moved into VideoSendStream. Also handling of the CVO rtp header extension is moved to VideoSendStreamImpl.
BUG=webrtc:5687
TBR=mflodman@webrtc.org
Review-Url: https://codereview.webrtc.org/2304363002
Cr-Commit-Position: refs/heads/master@{#14877}
- The legacy API is not used in WVoE/MC.
- Removed use of the API (along with StartReceive()) from unit tests.
BUG=webrtc:4690
Review-Url: https://codereview.webrtc.org/2453243003
Cr-Commit-Position: refs/heads/master@{#14858}
This CL is a pure refactoring which should not result in any functinal
changes. It moves ownership of the RtcEventLog from webrtc::Call to the
webrtc::PeerConnection object.
This is done so that we can add RtcEventLog support for ICE events -
which will require the TransportController to have a pointer to the
RtcEventLog. PeerConnection is the closest common owner of both Call and
TransportController (through WebRtcSession).
BUG=webrtc:6393
Review-Url: https://codereview.webrtc.org/2353033005
Cr-Commit-Position: refs/heads/master@{#14578}
Also rename some related minor methods. No functional changes
are intended/expected.
BUG=webrtc:5654
Review-Url: https://codereview.webrtc.org/2391963002
Cr-Commit-Position: refs/heads/master@{#14513}
