Adding a temporary workaround in the WebRTC Opus decoder wrapper to fix
https://issues.webrtc.org/376493209. Once the issue is fixed in libopus,
the workaround must be removed (TODO added in the code).
The workaround keeps track of the number of channels for the last
decoded packet and, if the decoder operates in stereo mode and the last
packet was a mono one, the left channel is copied into the right one
when comfort noise / PLC audio is generated.
Bug: webrtc:376493209
Change-Id: Iad3bfb1b393bd68833decf51b69b5238cb0ec4b7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/367740
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43371}
- With mono encoding and stereo decoding check that the decoded
signal is trivial stereo
- DTX tests
- With mono encoding and stereo decoding check that the comfort
noise generated by Opus is NOT(*) trivially stereo
- With stereo encoding and stereo decoding check that the comfort
noise generated by Opus is not trivially stereo
*: the test shows the behavior described in [1] and that needs to
be fixed.
[1] https://issues.webrtc.org/376493209
Bug: webrtc:376493209
Change-Id: I34aacd4bd7c79be9df05c242e912c9981896a73d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/367206
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43363}
- WebRTC does use the libopus DTX implementation
- The removed detail is anyways irrelevant in a docstring
Bug: webrtc:376493209
Change-Id: I3dfe1521259e596dbfa0db97f91ffb75deeb16b8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/367200
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43344}
Indirect input deps for imported proto is now handled by deps in
proto_library template, so we don't need to use proto_data_sources
anymore after https://crrev.com/c/5919027.
To remove proto_data_sources from proto_library template, let me clean
up proto_data_sources usages from this repository.
Bug: chromium:366137880
Change-Id: I288a30004f7d622be502477a0567b00d19432e89
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/366060
Auto-Submit: Takuto Ikuta <tikuta@google.com>
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43261}
This makes the delay manager interface significantly simpler and easier to expose.
Bug: None
Change-Id: Ie3d37c3b869eb17ca421a76e9d1af8f0a1a36ee5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/364781
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43186}
A slight behavior change is that we only increment total samples received when GetAudio is successful.
Bug: webrtc:370424996
Change-Id: I8607418c179ca3bc22963b98792a9e8b9af2d451
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/364220
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43139}
The code now uses NetEq directly instead of AcmReceiver.
Bug: webrtc:14867
Change-Id: I11c7e2ca00060ab15bba5ec67dfd92ec413196f6
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/364140
Commit-Queue: Henrik Lundin <henrik.lundin@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43108}
We need to have imported proto as proto_data_sources in BUILD.gn to
run the action remotely without workaround config in siso.
Bug: b/366137880
Change-Id: I053774f00b761520a8a85154e386da3edb8f39b4
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/362680
Reviewed-by: Jeremy Leconte <jleconte@webrtc.org>
Reviewed-by: Jeremy Leconte <jleconte@google.com>
Auto-Submit: Takuto Ikuta <tikuta@google.com>
Commit-Queue: Takuto Ikuta <tikuta@google.com>
Cr-Commit-Position: refs/heads/main@{#43040}
Use the DTX state from inside the Opus encoder instead of trying to
mimic the logic outside.
Bug: None
Change-Id: I852044fee261a5b7f9255c557a27adfd0b1701bd
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/362640
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Commit-Queue: Lionel Koenig <lionelk@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43034}
The class itself and its unit test remains, for now, but will be removed
later.
Bug: webrtc:14867
Change-Id: I36cec8fca7913663f63c53622ed2760e5e048c2f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/362580
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#43023}
ChannelReceive is now owning and interfacing with NetEq directly.
A new ResamplerHelper is added to acm_resampler.cc/.h, to do the
audio resampling that was previously done inside AcmReceiver.
AcmReceiver still remains, since it is used in other places for now.
Bug: webrtc:14867
Change-Id: If3eb6415e06b9b5e729d393713f3fccb31b0570f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/361820
Auto-Submit: Henrik Lundin <henrik.lundin@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Henrik Lundin <henrik.lundin@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42974}
We can avoid using the global now that field trials from Environment are
used in NetEq. This allows running multiple instances in parallel with
different settings.
Bug: webrtc:42220378
Change-Id: Icff8539e3ae9b61c86bb393d9a313e786e032b93
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/359720
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42894}
This replaces the payload type tracking in AcmReceiver with the one in
NetEq and should be a noop.
Bug: None
Change-Id: Iaf124b5e56a646f994b5c2af65d349ede550b7fd
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/360840
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Tomas Lundqvist <tomasl@google.com>
Cr-Commit-Position: refs/heads/main@{#42875}
This fixes an edge case where the frame size changes for a DTX packet.
We should avoid having the frame size larger than the timestamp gap.
Bug: None
Change-Id: I0a384cfb06f5aebc1654c1e3d127541fd24e05c7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/360722
Reviewed-by: Tomas Lundqvist <tomasl@google.com>
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42873}
This is currenly tracked in both AcmReceiver and NetEq. Adding this API
enables us to have it in just one place.
Bug: None
Change-Id: Ia537f87f36b0aedf19c00a57bd6cec4425a49df1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/360743
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Reviewed-by: Tomas Lundqvist <tomasl@google.com>
Cr-Commit-Position: refs/heads/main@{#42872}
Mark old overload deprecated.
This allows to migrate both calls through AudioDecoderFactory and direct calls to AudioDecpderOpus trait.
Bug: webrtc:356878416
Change-Id: I1502aee5b18aac43a8258e77b770c8e73a056f92
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/359741
Commit-Queue: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Auto-Submit: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42793}
Finalize change started in https://webrtc-review.googlesource.com/c/src/+/359243
Remove fallback to old interface and unneeded clock member in the config struct.
Bug: None
Change-Id: I4c2b65a09dd1c8a0d44ee76320b095516e2c61fb
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/359561
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42782}
to make it available for creating AudioDecoders
Bug: webrtc:356878416
Change-Id: Ibd24a55df70985dfe02d924da037618f13661032
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/359241
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42771}
To propagate field trials in addition to clock
Bug: webrtc:356878416
Change-Id: Idefc4848ec4af30c8aed0f93b7fadfc3181bddb1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/358980
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42761}
To propagate field trials into the NetEq and further towards Audio Decoders
Bug: webrtc:356878416
Change-Id: Ia7cf18451aef70441ca958bf652f492138c6051a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/358620
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42739}
Passing Environment would allow to propage field trials with it further to NetEq and AudioDecoders
Bug: webrtc:356878416
Change-Id: Ic68420df3b157ed341146207a2c45cb49e59a931
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/358501
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42728}
To allow delete old signature of the AudioEncoderOpus::MakeAudioEncoder function and thus guarantee Opus AudioEncoder always has an Environment
Bug: webrtc:343086059
Change-Id: Ib660678aeb5a549dddd1dffa3d8c28b2ec6b9d0b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/356981
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42690}
This switches from accepting a sample rate and convert to channel
size over to accepting the channel size.
Instead of InitializeIfNeeded:
* Offer a way to explicitly initialize PushResampler via the ctor
(needed for VoiceActivityDetectorWrapper)
* Implicitly check for the right configuration from within Resample().
(All calls to Resample() were preceded by a call to Initialize)
As part of this, refactor VoiceActivityDetectorWrapper (VADW):
* VADW is now initialized in the constructor and more const.
* Remove VADW::Initialize() and instead reconstruct VADW if needed.
Add constants for max sample rate and num channels to audio_util.h
In many cases the numbers for these values are embedded in the code
which has led to some inconsistency.
Bug: chromium:335805780
Change-Id: Iead0d52eb1b261a8d64e93f51401147c8fba32f0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/353360
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42587}
