Per Åhgren d20639f1f6 Correct the FFT windowing when computing the AEC NLP gain
This CL adds an nonwindowed spectrum of the linear filter error
to use in the NLP computation.

Bug: webrtc:8661
Change-Id: I45bc9bb3eb8eeac0c5d6adb414638eb12b635a27
Reviewed-on: https://webrtc-review.googlesource.com/38701
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#21583}
2018-01-11 14:41:11 +00:00

163 lines
5.9 KiB
C++

/*
* Copyright (c) 2017 The WebRTC project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "modules/audio_processing/aec3/subtractor.h"
#include <algorithm>
#include <numeric>
#include "api/array_view.h"
#include "modules/audio_processing/logging/apm_data_dumper.h"
#include "rtc_base/checks.h"
#include "rtc_base/numerics/safe_minmax.h"
namespace webrtc {
namespace {
void PredictionError(const Aec3Fft& fft,
const FftData& S,
rtc::ArrayView<const float> y,
std::array<float, kBlockSize>* e,
std::array<float, kBlockSize>* s) {
std::array<float, kFftLength> tmp;
fft.Ifft(S, &tmp);
constexpr float kScale = 1.0f / kFftLengthBy2;
std::transform(y.begin(), y.end(), tmp.begin() + kFftLengthBy2, e->begin(),
[&](float a, float b) { return a - b * kScale; });
if (s) {
for (size_t k = 0; k < s->size(); ++k) {
(*s)[k] = kScale * tmp[k + kFftLengthBy2];
}
}
std::for_each(e->begin(), e->end(),
[](float& a) { a = rtc::SafeClamp(a, -32768.f, 32767.f); });
}
} // namespace
Subtractor::Subtractor(const EchoCanceller3Config& config,
ApmDataDumper* data_dumper,
Aec3Optimization optimization)
: fft_(),
data_dumper_(data_dumper),
optimization_(optimization),
main_filter_(config.filter.length_blocks, optimization, data_dumper_),
shadow_filter_(config.filter.length_blocks, optimization, data_dumper_),
G_main_(config.filter.leakage_converged,
config.filter.leakage_diverged,
config.filter.main_noise_gate,
config.filter.error_floor),
G_shadow_(config.filter.shadow_rate, config.filter.shadow_noise_gate) {
RTC_DCHECK(data_dumper_);
}
Subtractor::~Subtractor() = default;
void Subtractor::HandleEchoPathChange(
const EchoPathVariability& echo_path_variability) {
const auto full_reset = [&]() {
main_filter_.HandleEchoPathChange();
shadow_filter_.HandleEchoPathChange();
G_main_.HandleEchoPathChange(echo_path_variability);
G_shadow_.HandleEchoPathChange();
converged_filter_ = false;
};
// TODO(peah): Add delay-change specific reset behavior.
if ((echo_path_variability.delay_change ==
EchoPathVariability::DelayAdjustment::kBufferFlush) ||
(echo_path_variability.delay_change ==
EchoPathVariability::DelayAdjustment::kDelayReset)) {
full_reset();
} else if (echo_path_variability.delay_change ==
EchoPathVariability::DelayAdjustment::kNewDetectedDelay) {
full_reset();
} else if (echo_path_variability.delay_change ==
EchoPathVariability::DelayAdjustment::kBufferReadjustment) {
full_reset();
}
}
void Subtractor::Process(const RenderBuffer& render_buffer,
const rtc::ArrayView<const float> capture,
const RenderSignalAnalyzer& render_signal_analyzer,
const AecState& aec_state,
SubtractorOutput* output) {
RTC_DCHECK_EQ(kBlockSize, capture.size());
rtc::ArrayView<const float> y = capture;
FftData& E_main = output->E_main;
FftData& E_main_nonwindowed = output->E_main_nonwindowed;
FftData E_shadow;
std::array<float, kBlockSize>& e_main = output->e_main;
std::array<float, kBlockSize>& e_shadow = output->e_shadow;
FftData S;
FftData& G = S;
// Form the output of the main filter.
main_filter_.Filter(render_buffer, &S);
PredictionError(fft_, S, y, &e_main, &output->s_main);
fft_.ZeroPaddedFft(e_main, Aec3Fft::Window::kHanning, &E_main);
fft_.ZeroPaddedFft(e_main, Aec3Fft::Window::kRectangular,
&E_main_nonwindowed);
// Form the output of the shadow filter.
shadow_filter_.Filter(render_buffer, &S);
PredictionError(fft_, S, y, &e_shadow, nullptr);
fft_.ZeroPaddedFft(e_shadow, Aec3Fft::Window::kHanning, &E_shadow);
if (!converged_filter_) {
const auto sum_of_squares = [](float a, float b) { return a + b * b; };
const float e2_main =
std::accumulate(e_main.begin(), e_main.end(), 0.f, sum_of_squares);
const float e2_shadow =
std::accumulate(e_shadow.begin(), e_shadow.end(), 0.f, sum_of_squares);
const float y2 = std::accumulate(y.begin(), y.end(), 0.f, sum_of_squares);
if (y2 > kBlockSize * 50.f * 50.f) {
converged_filter_ = (e2_main > 0.3 * y2 || e2_shadow > 0.1 * y2);
}
}
// Compute spectra for future use.
E_main.Spectrum(optimization_, output->E2_main);
E_main_nonwindowed.Spectrum(optimization_, output->E2_main_nonwindowed);
E_shadow.Spectrum(optimization_, output->E2_shadow);
// Update the main filter.
std::array<float, kFftLengthBy2Plus1> X2;
render_buffer.SpectralSum(main_filter_.SizePartitions(), &X2);
G_main_.Compute(X2, render_signal_analyzer, *output, main_filter_,
aec_state.SaturatedCapture(), &G);
main_filter_.Adapt(render_buffer, G);
data_dumper_->DumpRaw("aec3_subtractor_G_main", G.re);
data_dumper_->DumpRaw("aec3_subtractor_G_main", G.im);
// Update the shadow filter.
if (shadow_filter_.SizePartitions() != main_filter_.SizePartitions()) {
render_buffer.SpectralSum(shadow_filter_.SizePartitions(), &X2);
}
G_shadow_.Compute(X2, render_signal_analyzer, E_shadow,
shadow_filter_.SizePartitions(),
aec_state.SaturatedCapture(), &G);
shadow_filter_.Adapt(render_buffer, G);
data_dumper_->DumpRaw("aec3_subtractor_G_shadow", G.re);
data_dumper_->DumpRaw("aec3_subtractor_G_shadow", G.im);
main_filter_.DumpFilter("aec3_subtractor_H_main");
shadow_filter_.DumpFilter("aec3_subtractor_H_shadow");
}
} // namespace webrtc