mirror of
https://github.com/mollyim/webrtc.git
synced 2025-05-17 15:47:53 +01:00
ee84d39fce
Multichannel signals are downmixed to mono before decimation and delay estimation. This is useful when not all channels play audio content. The feature can be toggled in the AEC3 configuration. Bug: webrtc:10913 Change-Id: I7d40edf7732bb51fec69e7f3ca063d821c5069c4 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/151762 Commit-Queue: Gustaf Ullberg <gustaf@webrtc.org> Reviewed-by: Per Åhgren <peah@webrtc.org> Cr-Commit-Position: refs/heads/master@{#29126}
120 lines
4.7 KiB
C++
120 lines
4.7 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/echo_path_delay_estimator.h"
|
|
|
|
#include <array>
|
|
|
|
#include "api/audio/echo_canceller3_config.h"
|
|
#include "modules/audio_processing/aec3/aec3_common.h"
|
|
#include "modules/audio_processing/aec3/downsampled_render_buffer.h"
|
|
#include "modules/audio_processing/logging/apm_data_dumper.h"
|
|
#include "rtc_base/checks.h"
|
|
|
|
namespace webrtc {
|
|
|
|
EchoPathDelayEstimator::EchoPathDelayEstimator(
|
|
ApmDataDumper* data_dumper,
|
|
const EchoCanceller3Config& config)
|
|
: data_dumper_(data_dumper),
|
|
down_sampling_factor_(config.delay.down_sampling_factor),
|
|
sub_block_size_(down_sampling_factor_ != 0
|
|
? kBlockSize / down_sampling_factor_
|
|
: kBlockSize),
|
|
capture_decimator_(down_sampling_factor_),
|
|
matched_filter_(
|
|
data_dumper_,
|
|
DetectOptimization(),
|
|
sub_block_size_,
|
|
kMatchedFilterWindowSizeSubBlocks,
|
|
config.delay.num_filters,
|
|
kMatchedFilterAlignmentShiftSizeSubBlocks,
|
|
config.delay.down_sampling_factor == 8
|
|
? config.render_levels.poor_excitation_render_limit_ds8
|
|
: config.render_levels.poor_excitation_render_limit,
|
|
config.delay.delay_estimate_smoothing,
|
|
config.delay.delay_candidate_detection_threshold),
|
|
matched_filter_lag_aggregator_(data_dumper_,
|
|
matched_filter_.GetMaxFilterLag(),
|
|
config.delay.delay_selection_thresholds),
|
|
downmix_(config.delay.downmix_before_delay_estimation) {
|
|
RTC_DCHECK(data_dumper);
|
|
RTC_DCHECK(down_sampling_factor_ > 0);
|
|
}
|
|
|
|
EchoPathDelayEstimator::~EchoPathDelayEstimator() = default;
|
|
|
|
void EchoPathDelayEstimator::Reset(bool reset_delay_confidence) {
|
|
Reset(true, reset_delay_confidence);
|
|
}
|
|
|
|
absl::optional<DelayEstimate> EchoPathDelayEstimator::EstimateDelay(
|
|
const DownsampledRenderBuffer& render_buffer,
|
|
const std::vector<std::vector<float>>& capture) {
|
|
RTC_DCHECK_EQ(kBlockSize, capture[0].size());
|
|
|
|
std::array<float, kBlockSize> downsampled_capture_data;
|
|
rtc::ArrayView<float> downsampled_capture(downsampled_capture_data.data(),
|
|
sub_block_size_);
|
|
capture_decimator_.Decimate(capture, downmix_, downsampled_capture);
|
|
data_dumper_->DumpWav("aec3_capture_decimator_output",
|
|
downsampled_capture.size(), downsampled_capture.data(),
|
|
16000 / down_sampling_factor_, 1);
|
|
matched_filter_.Update(render_buffer, downsampled_capture);
|
|
|
|
absl::optional<DelayEstimate> aggregated_matched_filter_lag =
|
|
matched_filter_lag_aggregator_.Aggregate(
|
|
matched_filter_.GetLagEstimates());
|
|
|
|
// Run clockdrift detection.
|
|
if (aggregated_matched_filter_lag &&
|
|
(*aggregated_matched_filter_lag).quality ==
|
|
DelayEstimate::Quality::kRefined)
|
|
clockdrift_detector_.Update((*aggregated_matched_filter_lag).delay);
|
|
|
|
// TODO(peah): Move this logging outside of this class once EchoCanceller3
|
|
// development is done.
|
|
data_dumper_->DumpRaw(
|
|
"aec3_echo_path_delay_estimator_delay",
|
|
aggregated_matched_filter_lag
|
|
? static_cast<int>(aggregated_matched_filter_lag->delay *
|
|
down_sampling_factor_)
|
|
: -1);
|
|
|
|
// Return the detected delay in samples as the aggregated matched filter lag
|
|
// compensated by the down sampling factor for the signal being correlated.
|
|
if (aggregated_matched_filter_lag) {
|
|
aggregated_matched_filter_lag->delay *= down_sampling_factor_;
|
|
}
|
|
|
|
if (old_aggregated_lag_ && aggregated_matched_filter_lag &&
|
|
old_aggregated_lag_->delay == aggregated_matched_filter_lag->delay) {
|
|
++consistent_estimate_counter_;
|
|
} else {
|
|
consistent_estimate_counter_ = 0;
|
|
}
|
|
old_aggregated_lag_ = aggregated_matched_filter_lag;
|
|
constexpr size_t kNumBlocksPerSecondBy2 = kNumBlocksPerSecond / 2;
|
|
if (consistent_estimate_counter_ > kNumBlocksPerSecondBy2) {
|
|
Reset(false, false);
|
|
}
|
|
|
|
return aggregated_matched_filter_lag;
|
|
}
|
|
|
|
void EchoPathDelayEstimator::Reset(bool reset_lag_aggregator,
|
|
bool reset_delay_confidence) {
|
|
if (reset_lag_aggregator) {
|
|
matched_filter_lag_aggregator_.Reset(reset_delay_confidence);
|
|
}
|
|
matched_filter_.Reset();
|
|
old_aggregated_lag_ = absl::nullopt;
|
|
consistent_estimate_counter_ = 0;
|
|
}
|
|
} // namespace webrtc
|