Note: estimation is turned OFF if config_.ep_strength.default_len
is set >= 0 (in this case config_.ep_strength.default_len defines a
constant echo decay factor), and hence turned ON if < 0. In case the
echo tail estimation is turned ON, -config_.ep_strength.default_len is
the starting point for the estimator.
The estimation is done in two passes; first we go through all "sections"
(corresponding to chunks of length kFftLengthBy2) of the filter impulse
response to determine which sections correspond to a "stable" decay",
and then the second pass we go through each stable decay section and
estimate the decay. The actual decay estimation is based on linear
regression of the log magnitude of the squared impulse response.
A bunch of sanity checks are also performed continuously to avoid
estimation error during e.g., filter adaptation.
Bug: webrtc:8924
Change-Id: I686ce3f3e8b6b472348f8d6e01fb44c31e25145d
Reviewed-on: https://webrtc-review.googlesource.com/48440
Commit-Queue: Christian Schuldt <cschuldt@webrtc.org>
Reviewed-by: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22247}
This eliminates a lot of warnings when running AudioMixer unittests.
Bug: None
Change-Id: I978e859e8904bb382843a9c23a35ace1884aa4b4
Reviewed-on: https://webrtc-review.googlesource.com/58981
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22245}
This is a reland of e6cefdf9c572cdce55ff0497ad6e516c76132ee8.
Original change's description:
> Moved congestion controller to goog_cc folder.
>
> Bug: webrtc:8415
> Change-Id: I2070da0cacf1dbfc4b6a89285af3e68fd03497ab
> Reviewed-on: https://webrtc-review.googlesource.com/43841
> Commit-Queue: Sebastian Jansson <srte@webrtc.org>
> Reviewed-by: Björn Terelius <terelius@webrtc.org>
> Reviewed-by: Stefan Holmer <stefan@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#21928}
Bug: webrtc:8415
Change-Id: Ib5cf8641466655d64ac80f720561817f4cab49a9
Reviewed-on: https://webrtc-review.googlesource.com/53062
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22244}
We want to evaluate the impact on battery consumption from using the
fullscreen HW scaling.
Bug: None
Change-Id: If5becf02c6eaf77f9d0877827db39773ae17fc05
Reviewed-on: https://webrtc-review.googlesource.com/59101
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Commit-Queue: Magnus Jedvert <magjed@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22243}
Using unique_ptr of send side congestion controller (SSCC in
RtpTransportControllerSend. This prepares for a later CL where
the implementation of SSCC will be toggled by an experiment.
Bug: webrtc:8415
Change-Id: I52501d470efc96a62d0bd5c35ad361d43bceb9f9
Reviewed-on: https://webrtc-review.googlesource.com/58901
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22241}
The functionality in network_control is by design not rtp specific and
should live outside the rtp directory in congestion_controller.
Bug: webrtc:8415
Change-Id: I68778f1a630c8e93950040bc56ee864c201b5990
Reviewed-on: https://webrtc-review.googlesource.com/55383
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22240}
* Do not simulate freeze in decoded output file when frames have been dropped.
* Add more DCHECKs and consts.
* Remove unused members |num_encoded_frames_| and |num_decoded_frames_|.
* Move SdpVideoFormat conversion to TestConfig.
Bug: webrtc:8448
Change-Id: Ia879141f36dc23427cd1abcaa66716656fbaac2a
Reviewed-on: https://webrtc-review.googlesource.com/56802
Commit-Queue: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22239}
This removes a confusing name collision between webrtc::RateLimiter
and rtc::RateLimiter where the header file names were separated only by
an underscore.
Bug: None
Change-Id: Ifcf0a4e62b2bf3bd9057714d7c536f7609ad1b79
Reviewed-on: https://webrtc-review.googlesource.com/58741
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22237}
Modifying VideoSendStreamTest.RetransmitsNack to test for multiple lost
packets. This covers more failure modes since the RateLimiter class
always allows the first packet to get trough.
Bug: None
Change-Id: I2c408ea10ed4ac130edc55626b5ec03397ac0d9a
Reviewed-on: https://webrtc-review.googlesource.com/58743
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22236}
Makes the new task queue based congestion controller implement the
SendSideCongestionControllerInterface.
Bug: webrtc:8415
Change-Id: I3dfe11c2eb200bc8d85c83edf78d7fdd0129bbff
Reviewed-on: https://webrtc-review.googlesource.com/56781
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22235}
This reverts commit d2ed0a4c9e7f04060d8e3358eb0006c31579bb86.
Reason for revert: Breaks downstream projects.
Original change's description:
> Replaced temporal_layer_thresholds_bps[] field with num_temporal_layers.
>
> temporal_layer_thresholds_bps served only one purpose: its size was used
> to infer number of temporal layers. I replaced it with num_temporal_layers,
> which does what is says.
>
> The practical reason for this change is the need to have possibility to
> distinguish between cases when VP9 SVC temporal layering was/not set
> through field trial. That was not possible with
> temporal_layer_thresholds_bps[] because empty vector means 1 temporal
> layer.
>
> Bug: webrtc:8518
> Change-Id: I275ec3a8c74e8ba409eb049878199f132a20ec51
> Reviewed-on: https://webrtc-review.googlesource.com/58084
> Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
> Reviewed-by: Erik Språng <sprang@webrtc.org>
> Reviewed-by: Stefan Holmer <stefan@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22230}
TBR=sprang@webrtc.org,stefan@webrtc.org,ssilkin@webrtc.org
Change-Id: Ic2940f7f78a74312170940d51ad8967cde8ad42f
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:8518
Reviewed-on: https://webrtc-review.googlesource.com/58902
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22234}
Allows passing in the application context to NetworkMonitor in
startMonitoring. The audio code will refactored once it is moved under
sdk/android.
Bug: webrtc:8937
Change-Id: I50c917a845fc4f711899a97d34c04813cc68b68c
Reviewed-on: https://webrtc-review.googlesource.com/58091
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Commit-Queue: Sami Kalliomäki <sakal@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22231}
temporal_layer_thresholds_bps served only one purpose: its size was used
to infer number of temporal layers. I replaced it with num_temporal_layers,
which does what is says.
The practical reason for this change is the need to have possibility to
distinguish between cases when VP9 SVC temporal layering was/not set
through field trial. That was not possible with
temporal_layer_thresholds_bps[] because empty vector means 1 temporal
layer.
Bug: webrtc:8518
Change-Id: I275ec3a8c74e8ba409eb049878199f132a20ec51
Reviewed-on: https://webrtc-review.googlesource.com/58084
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22230}
This CL activates unit tests for the task queue based send side
congestion controller that will replace the current one in the future.
To be able to have the tests run side by side with the tests of the old
congestion controller, the old tests have been prefixed with "Legacy".
This CL also contains some minor fixes to the new congestion controller
code.
Bug: webrtc:8415
Change-Id: I5e7474d42f17fcbfef402e26f638846fa3424695
Reviewed-on: https://webrtc-review.googlesource.com/55381
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22229}
The additional include is needed in order to use EXPECT_NONFATAL_FAILURE()
in unit tests.
Bug: webrtc:8948
Change-Id: If5b9ceb89a3a36480657d094cfabc81c9b0e15b7
Reviewed-on: https://webrtc-review.googlesource.com/58096
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22227}
SetTransportOverhead was used by send streams to signal the packet
overhead that they received from Call. However, call receives the value
from OnNetworkRouteChanged in WebRtcVideoChannel and
WebRtcVoiceMediaChannel which is already propagated to
RtpTransportControllerSend. By skipping the round trip, the interface on
the rtp transport controller can be reduced.
Bug: None
Change-Id: I759b1207aab214bbc2b993106f6ff7cc24e177f7
Reviewed-on: https://webrtc-review.googlesource.com/57182
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22226}
The construction task did not provide much value while making it harder
to follow what happens in the code. Replacing it with a lambda closure.
This reduces the amount of arguments to keep track of and moves the
initialization code to the context of the VideoSendStream constructor.
A slight difference in behavior is that ownership of fec_controller
is kept in the VideoSendStream constructor scope until it is moved
in the initialization task. Since the construction waits for completion
of the task, this should not change any actual behavior.
Bug: None
Change-Id: I5ff84587486b77d855cc6c178fb626124ec65030
Reviewed-on: https://webrtc-review.googlesource.com/58642
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22225}
Uploading of the file to device causes timeouts. I removed it from
resources for iOS and Android builds and disabled the test that used
the file since we don't really need to run it on test bots.
Bug: webrtc:8936
Change-Id: Ia5e04c4630544eca8e56826c9e89c9c9f4dcb600
Reviewed-on: https://webrtc-review.googlesource.com/58090
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Commit-Queue: Sergey Silkin <ssilkin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22222}
This makes it easier to have the new and the old send side congestion
controller side by side. This namespace is only temporary. As soon the
new task queue based congestion controller is fully functional, the old
will be deprecated and removed together with the temporary namespace.
Bug: webrtc:8415
Change-Id: Ie817511345c91cab2ebca68f038075875c7e6529
Reviewed-on: https://webrtc-review.googlesource.com/56720
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22221}
Added sanity checks to interval budget to protect against integer
overflow. To avoid tests failing due to initializing paced sender
in another time domain, the elapsed time that is input into the
interval budget is limited to max 2 seconds.
Bug: webrtc:8942
Change-Id: I9ed32f059e65df7898c37bb34a008189ce79dc60
Reviewed-on: https://webrtc-review.googlesource.com/58087
Reviewed-by: Tommi <tommi@webrtc.org>
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22220}
This CL adds wrappers for the following PeerConnection native
APIs to the Objective C API:
- SdpSemantics enum added to the RTCConfiguration
- RTCRtpTransceiver
- RTCPeerConnection.addTrack
- RTCPeerConnection.removeTrack
- RTCPeerConnection.addTransceiver
- RTCPeerConnection.transceivers
Bug: webrtc:8870
Change-Id: I9449df9742a59e90894712dc7749ca30b569d94b
Reviewed-on: https://webrtc-review.googlesource.com/54780
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22214}
This allows clients to move to these new accessors and off of the
sync_label field which is deprecated.
Bug: webrtc:7932
Change-Id: I32b30087bf1be380d607a649bd90fa9617dafeb9
Reviewed-on: https://webrtc-review.googlesource.com/58020
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22213}
To be able to safely move over to the new code, the revised
code is added alongside the old code. Most of the files added
in this CL are more or less direct copies of the previous code.
This new version of send side congestion controller will be
activated under a field trial in a followup CL.
Bug: webrtc:8415
Change-Id: I034e583cf891a8f6357119739a1517cc0a4abe88
Reviewed-on: https://webrtc-review.googlesource.com/53322
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22212}
The flag is passed as --isolated-script-test-perf-output=/b/whatever
on the bots, but this code expected a blank space instead of =.
Bug: webrtc:8932
Change-Id: I9ca48c9b285e365ac23a04ea2e89d9a8e75f5540
Reviewed-on: https://webrtc-review.googlesource.com/58088
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Commit-Queue: Patrik Höglund <phoglund@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22211}
Now VCMTiming::MaxWaitingTime will not clip negative values. Thus frame
buffer will be able to distinguish between late frames and when waiting
cycle was simply interrupted by a new inserted frame right before the
waiting timer would expire.
Bug: webrtc:8917
Change-Id: I6b253f459fcb3a346064a103cc92ee332b074e1b
Reviewed-on: https://webrtc-review.googlesource.com/57741
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22210}
Allows mixing up to 4 input streams. Useful for profiling and manual
tests. Allows testing different combinations of input/output rates and
number of channels. Reads and writes WAV files. Can also configure
whether to use the Limiter component of the AudioMixer.
Bug: webrtc:8925
Change-Id: Iaf4fee5284980f6ed01f4bb721e49bb1af8dd392
Reviewed-on: https://webrtc-review.googlesource.com/56842
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22209}
This is a reland of 12dc1842d62ee8df1e462f9b6a617fef9ab8b3b7.
Original change's description:
> Some cleanup for the logging code:
>
> * Only include 'tag' for Android. Before there was an
> extra std::string variable per log statement for all
> platforms.
> * Remove unused logging macro for Windows and 'module' ctor argument.
> * Move httpcommon code out of logging and to where it's used.
>
> Change-Id: I347805d3df2cee2840c0d2eef5bfefaff1cdbf37
> Bug: webrtc:8928
> Reviewed-on: https://webrtc-review.googlesource.com/57183
> Reviewed-by: Jonas Olsson <jonasolsson@webrtc.org>
> Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
> Commit-Queue: Tommi <tommi@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22184}
Bug: webrtc:8928
Change-Id: Ib97895aaeb376e19f136d258c0259a340235a5d1
Reviewed-on: https://webrtc-review.googlesource.com/58200
Commit-Queue: Tommi <tommi@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22208}
The webrtc::AudioMixer uses a limiter component. This CL allows
changes the APM-AGC limiter to the APM-AGC2 limiter though a Chrome
field trial.
The AGC2 limiter has a float interface. We plan to eventually switch
to the AGC2 limiter. Therefore, we will now mix in de-interleaved
floats. Float mixing will happen both when using the old limiter and
when using the new one.
After this CL the mixer will support two limiters. The limiters have
different interfaces and need different processing steps. Because of
that, we make (rather big) changes to the control flow in
FrameCombiner. For a short while, we will mix in deinterleaved floats
when using any limiter.
Originally landed in https://webrtc-review.googlesource.com/c/src/+/56141/
Reverted in https://webrtc-review.googlesource.com/c/src/+/57940
because of both breaking compilation and having a severe error. The
error is fixed and a test is added. The compilation issue is fixed.
Bug: webrtc:8925
Change-Id: Ieba138dee9652c826459fe637ae2dccbbc06bcf0
Reviewed-on: https://webrtc-review.googlesource.com/58085
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22207}
This reverts commit 63e83c77ae81730a78ec4d5bf0465f25970f867a.
Reason for revert: JNI generator is not using the heap profiler
anymore.
Original change's description:
> Forward fix jni_generator_helper.h.
>
> In crrev.com/531028, the JNI generator starts to add heap profiler
> events to JNI generated functions.
>
> This will cause a ~80KiB regression and at the moment it is breaking
> the Chromium Roll into WebRTC.
>
> This CL defines a void macro to re-enable the Chromium Roll avoiding
> the size regression.
>
> Bug: chromium:801260
> Change-Id: I9543299199c4e14b6b9b235c5cb98c0d53cf29ea
> Reviewed-on: https://webrtc-review.googlesource.com/43021
> Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
> Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#21730}
TBR=mbonadei@webrtc.org,magjed@webrtc.org,sakal@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: chromium:801260
Change-Id: I7dac211b89d8206dc461af0a17b6d53cc8661b2a
Reviewed-on: https://webrtc-review.googlesource.com/58040
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22206}
Summary:
The implementation of H264AnnexBBufferHasVideoFormatDescription was
assuming that the SPS NALU is either the first NALU in the stream, or
the second one, in case an AUD NALU is present in the first location.
This change removes this assumption and instead searches for the SPS
NALU, failing only if we can't find one.
In addition, it cleans up some binary buffer manipulation code, using the
the parsed NALU indices we already have in AnnexBBufferReader instead.
Test Plan: Unit tests
Change-Id: Id9715aa1d751f0ba1a1992def2b690607896df56
bug: webrtc:8922
Change-Id: Id9715aa1d751f0ba1a1992def2b690607896df56
Reviewed-on: https://webrtc-review.googlesource.com/49982
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22205}
This is a fairly minimalistic string building class that
can be used instead of stringstream, which is discouraged
but tempting to use due to its convenient interface and
familiarity for anyone using our logging macros.
As a starter, I'm changing the string building code in
ReceiveStatisticsProxy and SendStatisticsProxy from using
stringstream and using SimpleStringBuilder instead.
In the case of SimpleStringBuilder, there's a single allocation,
it's done on the stack (fast), and minimal code is required for
each concatenation. The developer is responsible for ensuring
that the buffer size is adequate but the class won't overflow
the buffer. In dcheck-enabled builds, a check will go off if
we run out of buffer space.
As part of using SimpleStringBuilder for a small part of
rtc::LogMessage, a few more changes were made:
- SimpleStringBuilder is used for formatting errors instead of ostringstream.
- A new 'noop' state has been introduced for log messages that will be dropped.
- Use a static (singleton) noop ostream object for noop logging messages
instead of building up an actual ostringstream object that will be dropped.
- Add a LogMessageForTest class for better state inspection/testing.
- Fix benign bug in LogTest.Perf, change the test to not use File IO and
always enable it.
- Ensure that minimal work is done for noop messages.
- Remove dependency on rtc::Thread.
- Add tests for the extra_ field, correctly parsed paths and noop handling.
Bug: webrtc:8529, webrtc:4364, webrtc:8933
Change-Id: Ifa258c135135945e4560d9e24315f7d96f784acb
Reviewed-on: https://webrtc-review.googlesource.com/55520
Commit-Queue: Tommi <tommi@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Jonas Olsson <jonasolsson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22203}
