- Bug fix: the desired initial gain quickly dropped to 0 dB hence
starting a call with a too low level
- New tuning to make AGC2 more robust to VAD mistakes
- Smarter max gain increase speed: to deal with an increased threshold
of adjacent speech frames, the gain applier temporarily allows a
faster gain increase to deal with a longer time spent waiting for
enough speech frames in a row to be observed
- Saturation protector isolated from `AdaptiveModeLevelEstimator` to
simplify the unit tests for the latter (non bit-exact change)
- AGC2 adaptive digital config: unnecessary params deprecated
- Code readability improvements
- Data dumps clean-up and better naming
Bug: webrtc:7494
Change-Id: I4e36059bdf2566cc2a7e1a7e95b7430ba9ae9844
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/215140
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Jesus de Vicente Pena <devicentepena@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33736}
The current noise level estimator has a bug due to which the estimated
level decays to the lower bound in a few seconds when speech is observed.
Instead of fixing the current implementation, which is based on a
stationarity classifier, an alternative, lightweight, noise floor
estimator has been added and tuned for AGC2.
Tested on several AEC dumps including HW mute, music and fast talking.
Bug: webrtc:7494
Change-Id: Iae4cff9fc955a716878f830957e893cd5bc59446
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214133
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33733}
Done in preparation for the child CL which adds an alternative
implementation.
Bug: webrtc:7494
Change-Id: I4963376afc917eae434a0d0ccee18f21880eefe0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/214125
Reviewed-by: Jakob Ivarsson <jakobi@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#33646}
In preparation for adding AVX2 code, a safe scheme to support
different SIMD optimizations is added.
Safety features:
- AVX2 kill switch to stop using it even if supported by the
architecture
- struct indicating the available CPU features propagated from
AGC2 to each component; in this way
- better control over the unit tests
- no need to propagate individual kill switches but just
set to false features that are turned off
Note that (i) this CL does not change the performance of the RNN VAD
and (ii) no AVX2 optimization is added yet.
Bug: webrtc:10480
Change-Id: I0e61f3311ecd140f38369cf68b6e5954f3dc1f5a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/193140
Reviewed-by: Per Åhgren <peah@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32739}
This CL adds and wires up the following parameters:
- VAD probability attack used in `VadLevelAnalyzer`
- Adjacent spech frames threshold used in `AdaptiveModeLevelEstimator`
- Initial saturation margin used in `AdaptiveModeLevelEstimator`
The deprecated ctor in `AdaptiveModeLevelEstimator` is removed.
Tested: bit-exactness verified with audioproc_f
Bug: webrtc:7494
Change-Id: Idf94aaadba1476757f845e696bfb47ff6252d5f0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/186048
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32265}
This CL adds and wires up a parameter (namely, adjacent speech
frames threshold) used in `AdaptiveDigitalGainApplier`.
Tested: bit-exactness verified with audioproc_f
Bug: webrtc:7494
Change-Id: I751cd91f08a6e98ee20f767c8df0ed121c8d4b68
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/186049
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32264}
This CL was written in preparation for the next CL in the chain and
it contains the following changes:
- SignalWithLevels -> AdaptiveDigitalGainApplier::FrameInfo
- Frame view removed from AdaptiveDigitalGainApplier::FrameInfo
- AdaptiveDigitalGainApplier::Process now gets side info as const& to
avoid unnecessary copies
- AdaptiveAgc::Process: `last_audio_level` renamed to `limiter_envelope`
to better reflect what that actually is
- Missing class/method docstrings added
Tested: bit-exactness verified with audioproc_f
Bug: webrtc:7494
Change-Id: Ie25dcd389d6eed74ea9a65f0720eeb8f20f0096b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/186040
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32251}
`AdaptiveModeLevelEstimator::last_level_dbfs_` doesn't need to be optional.
Note: this CL breaks the chain of 3 CLs titled
"AGC2 AdaptiveModeLevelEstimator min consecutive speech frames".
Bug: webrtc:7494
Change-Id: Id5b409ca5cb5f11ed132c861b7995b9721e167bb
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/185809
Reviewed-by: Minyue Li <minyue@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32237}
This is the first CL needed to add a new `AdaptiveModeLevelEstimator`
feature that makes AGC2 more robus to VAD mistakes: the level estimator
discards estimation updates when too few consecutive speech frames are
observed.
In this CL, the state of the estimator is defined in a separate struct
so that in a follow-up CL a new member of that type can be added to
hold a temporary state (that can be either confirmed or discarded).
Tested: Bit-exactness verified with audioproc_f
Bug: webrtc:7494
Change-Id: Ic2ea5ed63c493b9f3a79f19e7f5eaecaa6808ace
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/184931
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Reviewed-by: Ivo Creusen <ivoc@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32199}
Refactoring CL to improve names and allow to inject a VAD into
`VadLevelAnalyzer` (new name for `VadWithLevel`).
The injectable VAD is needed to inject a mock VAD and write better
unit tests as new features are going to be added to the class.
Bug: webrtc:7494
Change-Id: Ic0cea1e86a19a82533bd40fa04c061be3c44f068
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/185180
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32195}
This CL makes possible to choose the level estimation for the adaptive
digital GC of AGC2. The options are RMS (default and currently used
estimator) and peak-based (already computed, but not used).
Besides adding the new AGC2 config param for the level estimator, this CL
also refactors the config class by making it more structured.
Bug: webrtc:7494
Change-Id: I20eb558ca50f13536aa7bdea08d21de3b630f8bc
Reviewed-on: https://webrtc-review.googlesource.com/c/110144
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Reviewed-by: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25620}
The extra saturation margin is a setting for the SaturationProtector
in GainController2. The higher it is, the less gain GC2 will apply. In
this CL we pipe the setting up to audio_processing.h. Now the setting
can be set at a high level.
Also in this CL add a few (missing, they should have been there
already) tests for the GC2 and GC2 with saturation margin.
Bug: webrtc:7494
Change-Id: I1b61f1662e6c6a8817fd5b0e845339694bf8d50d
Reviewed-on: https://webrtc-review.googlesource.com/c/109001
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25470}
This CL is the result of running include-what-you-use tool on part
of the code base (audio target and dependencies) plus manual fixes.
bug: webrtc:8311
Change-Id: I277d281ce943c3ecc1bd45fd8d83055931743604
Reviewed-on: https://webrtc-review.googlesource.com/c/106280
Commit-Queue: Yves Gerey <yvesg@google.com>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25311}
Hypothetical scenario: short weak speech at start of call, then high
noise. The digital adaptive AGC2 would pick a high gain, and then
continue to apply it on the noise. Unless the noise is detected by the
noise estimator, the gain would never be reduced.
This CL addresses the issue by sending limiter gain info to the
adaptive digital AGC2.
Bug: webrtc:7494
Change-Id: Idf5c2686af0f5e5bad981d39a95b8efc9ffb9d64
Reviewed-on: https://webrtc-review.googlesource.com/102641
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#24922}
Agc2 applies a digital gain to the nearend signal.
When the analog level changes, the digital gain calculation is no
longer valid. Therefore Agc2 should be notified to analog gain
changes.
This CL also allow audioproc_f to chain AGC1 and AGC2. In a dependent
CL we will allow using AGC1 for analog gain and AGC2 for digital
gain.
Bug: webrtc:7494
Change-Id: Id75b3728fbf2de1d84b7fba005e4670c7a2985d9
Reviewed-on: https://webrtc-review.googlesource.com/89387
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#24231}
* Move 'VadWithLevel' to AGC2 where it belongs.
* Remove the vectors from VadWithLevel. They were there to make it work
with modules/audio_processing/vad, which we don't need any longer.
* Remove the vector handling from AGC2. It was spread out across
AdaptiveDigitalGainApplier, AdaptiveAGC and their unit tests.
* Hack the RNN VAD into VadWithLevel. The main issue is the resampling.
Bug: webrtc:9076
Change-Id: I13056c985d0ec41269735150caf4aaeb6ff9281e
Reviewed-on: https://webrtc-review.googlesource.com/77364
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23688}
AGC2 component that computes and applies the digital gain.
The gain is computed from an estimated speech and noise level.
This component decides how fast the gain can change and what it
should be.
Bug: webrtc:7494
Change-Id: If55b6e5c765f958e433730cd9e3b2b93c14a7910
Reviewed-on: https://webrtc-review.googlesource.com/64985
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22741}
We put back the old noise estimator from LevelController. We add a few
new unit tests. We also re-arrange the code so that it fits with how
it is used in AGC2. The differences are:
1. The NoiseLevelEstimator is now fully self-contained.
2. The NoiseLevelEstimator is responsible for calling SignalClassifier
and computing the signal energy. Previously the signal type and
energy were used in several places. It made sense to compute the
values independently of the noise calculation.
3. Re-initialization doesn't have to be done by the caller.
4. The interface is AudioFrameView instead of AudioBuffer.
# Bots are green, nothing should break internal stuff
NOTRY=True
Bug: webrtc:7494
Change-Id: I442bdbbeb3796eb2518e96000aec9dc5a039ae6d
Reviewed-on: https://webrtc-review.googlesource.com/66380
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22738}
This CL defines the control flow of the adaptive AGC. It also defines
method and class stubs.
Contents:
1. Divide the 'agc2' build target into 'fixed_digital' and
'adaptive_digital'.
1. Update the dependencies of everything that depended on 'agc2'.
2. Define the sub-modules of the adaptive digital AGC 2. They are:
1. Level Estimator - it gets the energy and a speech probability
and updates a speech level estimate.
2. Noise Estimator - it gets an immutable view of the speech frame
and updates the noise level estimate
3. Gain applier - it gets the speech frame, the current speech and
noise estimates, and the speech probability. It finds a gain to
apply and applies it.
4. AdaptiveAgc - sets up and controls the sub-modules described
above.
Bug: webrtc:7494
Change-Id: Ib7ccd8924e94eead0bc5f935b5d8a12e06e24fd1
Reviewed-on: https://webrtc-review.googlesource.com/64440
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Commit-Queue: Alex Loiko <aleloi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22628}
