Update to current webrtc library
This is from the upstream library commit id 3326535126e435f1ba647885ce43a8f0f3d317eb, corresponding to Chromium 88.0.4290.1.
This commit is contained in:
Arun Raghavan
53
webrtc/audio/utility/BUILD.gn
Normal file
53
webrtc/audio/utility/BUILD.gn
Normal file
@ -0,0 +1,53 @@
|
||||
# Copyright (c) 2016 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.
|
||||
import("../../webrtc.gni")
|
||||
|
||||
group("utility") {
|
||||
deps = [ ":audio_frame_operations" ]
|
||||
}
|
||||
|
||||
rtc_library("audio_frame_operations") {
|
||||
visibility = [ "*" ]
|
||||
sources = [
|
||||
"audio_frame_operations.cc",
|
||||
"audio_frame_operations.h",
|
||||
"channel_mixer.cc",
|
||||
"channel_mixer.h",
|
||||
"channel_mixing_matrix.cc",
|
||||
"channel_mixing_matrix.h",
|
||||
]
|
||||
|
||||
deps = [
|
||||
"../../api/audio:audio_frame_api",
|
||||
"../../common_audio",
|
||||
"../../rtc_base:checks",
|
||||
"../../rtc_base:deprecation",
|
||||
"../../rtc_base:rtc_base_approved",
|
||||
"../../system_wrappers:field_trial",
|
||||
]
|
||||
}
|
||||
|
||||
if (rtc_include_tests) {
|
||||
rtc_library("utility_tests") {
|
||||
testonly = true
|
||||
sources = [
|
||||
"audio_frame_operations_unittest.cc",
|
||||
"channel_mixer_unittest.cc",
|
||||
"channel_mixing_matrix_unittest.cc",
|
||||
]
|
||||
deps = [
|
||||
":audio_frame_operations",
|
||||
"../../api/audio:audio_frame_api",
|
||||
"../../rtc_base:checks",
|
||||
"../../rtc_base:rtc_base_approved",
|
||||
"../../test:field_trial",
|
||||
"../../test:test_support",
|
||||
"//testing/gtest",
|
||||
]
|
||||
}
|
||||
}
|
||||
294
webrtc/audio/utility/audio_frame_operations.cc
Normal file
294
webrtc/audio/utility/audio_frame_operations.cc
Normal file
@ -0,0 +1,294 @@
|
||||
/*
|
||||
* Copyright (c) 2012 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 "audio/utility/audio_frame_operations.h"
|
||||
|
||||
#include <string.h>
|
||||
|
||||
#include <algorithm>
|
||||
#include <cstdint>
|
||||
#include <utility>
|
||||
|
||||
#include "common_audio/include/audio_util.h"
|
||||
#include "rtc_base/checks.h"
|
||||
#include "rtc_base/numerics/safe_conversions.h"
|
||||
|
||||
namespace webrtc {
|
||||
namespace {
|
||||
|
||||
// 2.7ms @ 48kHz, 4ms @ 32kHz, 8ms @ 16kHz.
|
||||
const size_t kMuteFadeFrames = 128;
|
||||
const float kMuteFadeInc = 1.0f / kMuteFadeFrames;
|
||||
|
||||
} // namespace
|
||||
|
||||
void AudioFrameOperations::Add(const AudioFrame& frame_to_add,
|
||||
AudioFrame* result_frame) {
|
||||
// Sanity check.
|
||||
RTC_DCHECK(result_frame);
|
||||
RTC_DCHECK_GT(result_frame->num_channels_, 0);
|
||||
RTC_DCHECK_EQ(result_frame->num_channels_, frame_to_add.num_channels_);
|
||||
|
||||
bool no_previous_data = result_frame->muted();
|
||||
if (result_frame->samples_per_channel_ != frame_to_add.samples_per_channel_) {
|
||||
// Special case we have no data to start with.
|
||||
RTC_DCHECK_EQ(result_frame->samples_per_channel_, 0);
|
||||
result_frame->samples_per_channel_ = frame_to_add.samples_per_channel_;
|
||||
no_previous_data = true;
|
||||
}
|
||||
|
||||
if (result_frame->vad_activity_ == AudioFrame::kVadActive ||
|
||||
frame_to_add.vad_activity_ == AudioFrame::kVadActive) {
|
||||
result_frame->vad_activity_ = AudioFrame::kVadActive;
|
||||
} else if (result_frame->vad_activity_ == AudioFrame::kVadUnknown ||
|
||||
frame_to_add.vad_activity_ == AudioFrame::kVadUnknown) {
|
||||
result_frame->vad_activity_ = AudioFrame::kVadUnknown;
|
||||
}
|
||||
|
||||
if (result_frame->speech_type_ != frame_to_add.speech_type_)
|
||||
result_frame->speech_type_ = AudioFrame::kUndefined;
|
||||
|
||||
if (!frame_to_add.muted()) {
|
||||
const int16_t* in_data = frame_to_add.data();
|
||||
int16_t* out_data = result_frame->mutable_data();
|
||||
size_t length =
|
||||
frame_to_add.samples_per_channel_ * frame_to_add.num_channels_;
|
||||
if (no_previous_data) {
|
||||
std::copy(in_data, in_data + length, out_data);
|
||||
} else {
|
||||
for (size_t i = 0; i < length; i++) {
|
||||
const int32_t wrap_guard = static_cast<int32_t>(out_data[i]) +
|
||||
static_cast<int32_t>(in_data[i]);
|
||||
out_data[i] = rtc::saturated_cast<int16_t>(wrap_guard);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
int AudioFrameOperations::MonoToStereo(AudioFrame* frame) {
|
||||
if (frame->num_channels_ != 1) {
|
||||
return -1;
|
||||
}
|
||||
UpmixChannels(2, frame);
|
||||
return 0;
|
||||
}
|
||||
|
||||
int AudioFrameOperations::StereoToMono(AudioFrame* frame) {
|
||||
if (frame->num_channels_ != 2) {
|
||||
return -1;
|
||||
}
|
||||
DownmixChannels(1, frame);
|
||||
return frame->num_channels_ == 1 ? 0 : -1;
|
||||
}
|
||||
|
||||
void AudioFrameOperations::QuadToStereo(const int16_t* src_audio,
|
||||
size_t samples_per_channel,
|
||||
int16_t* dst_audio) {
|
||||
for (size_t i = 0; i < samples_per_channel; i++) {
|
||||
dst_audio[i * 2] =
|
||||
(static_cast<int32_t>(src_audio[4 * i]) + src_audio[4 * i + 1]) >> 1;
|
||||
dst_audio[i * 2 + 1] =
|
||||
(static_cast<int32_t>(src_audio[4 * i + 2]) + src_audio[4 * i + 3]) >>
|
||||
1;
|
||||
}
|
||||
}
|
||||
|
||||
int AudioFrameOperations::QuadToStereo(AudioFrame* frame) {
|
||||
if (frame->num_channels_ != 4) {
|
||||
return -1;
|
||||
}
|
||||
|
||||
RTC_DCHECK_LE(frame->samples_per_channel_ * 4,
|
||||
AudioFrame::kMaxDataSizeSamples);
|
||||
|
||||
if (!frame->muted()) {
|
||||
QuadToStereo(frame->data(), frame->samples_per_channel_,
|
||||
frame->mutable_data());
|
||||
}
|
||||
frame->num_channels_ = 2;
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
void AudioFrameOperations::DownmixChannels(const int16_t* src_audio,
|
||||
size_t src_channels,
|
||||
size_t samples_per_channel,
|
||||
size_t dst_channels,
|
||||
int16_t* dst_audio) {
|
||||
if (src_channels > 1 && dst_channels == 1) {
|
||||
DownmixInterleavedToMono(src_audio, samples_per_channel, src_channels,
|
||||
dst_audio);
|
||||
return;
|
||||
} else if (src_channels == 4 && dst_channels == 2) {
|
||||
QuadToStereo(src_audio, samples_per_channel, dst_audio);
|
||||
return;
|
||||
}
|
||||
|
||||
RTC_NOTREACHED() << "src_channels: " << src_channels
|
||||
<< ", dst_channels: " << dst_channels;
|
||||
}
|
||||
|
||||
void AudioFrameOperations::DownmixChannels(size_t dst_channels,
|
||||
AudioFrame* frame) {
|
||||
RTC_DCHECK_LE(frame->samples_per_channel_ * frame->num_channels_,
|
||||
AudioFrame::kMaxDataSizeSamples);
|
||||
if (frame->num_channels_ > 1 && dst_channels == 1) {
|
||||
if (!frame->muted()) {
|
||||
DownmixInterleavedToMono(frame->data(), frame->samples_per_channel_,
|
||||
frame->num_channels_, frame->mutable_data());
|
||||
}
|
||||
frame->num_channels_ = 1;
|
||||
} else if (frame->num_channels_ == 4 && dst_channels == 2) {
|
||||
int err = QuadToStereo(frame);
|
||||
RTC_DCHECK_EQ(err, 0);
|
||||
} else {
|
||||
RTC_NOTREACHED() << "src_channels: " << frame->num_channels_
|
||||
<< ", dst_channels: " << dst_channels;
|
||||
}
|
||||
}
|
||||
|
||||
void AudioFrameOperations::UpmixChannels(size_t target_number_of_channels,
|
||||
AudioFrame* frame) {
|
||||
RTC_DCHECK_EQ(frame->num_channels_, 1);
|
||||
RTC_DCHECK_LE(frame->samples_per_channel_ * target_number_of_channels,
|
||||
AudioFrame::kMaxDataSizeSamples);
|
||||
|
||||
if (frame->num_channels_ != 1 ||
|
||||
frame->samples_per_channel_ * target_number_of_channels >
|
||||
AudioFrame::kMaxDataSizeSamples) {
|
||||
return;
|
||||
}
|
||||
|
||||
if (!frame->muted()) {
|
||||
// Up-mixing done in place. Going backwards through the frame ensure nothing
|
||||
// is irrevocably overwritten.
|
||||
for (int i = frame->samples_per_channel_ - 1; i >= 0; i--) {
|
||||
for (size_t j = 0; j < target_number_of_channels; ++j) {
|
||||
frame->mutable_data()[target_number_of_channels * i + j] =
|
||||
frame->data()[i];
|
||||
}
|
||||
}
|
||||
}
|
||||
frame->num_channels_ = target_number_of_channels;
|
||||
}
|
||||
|
||||
void AudioFrameOperations::SwapStereoChannels(AudioFrame* frame) {
|
||||
RTC_DCHECK(frame);
|
||||
if (frame->num_channels_ != 2 || frame->muted()) {
|
||||
return;
|
||||
}
|
||||
|
||||
int16_t* frame_data = frame->mutable_data();
|
||||
for (size_t i = 0; i < frame->samples_per_channel_ * 2; i += 2) {
|
||||
std::swap(frame_data[i], frame_data[i + 1]);
|
||||
}
|
||||
}
|
||||
|
||||
void AudioFrameOperations::Mute(AudioFrame* frame,
|
||||
bool previous_frame_muted,
|
||||
bool current_frame_muted) {
|
||||
RTC_DCHECK(frame);
|
||||
if (!previous_frame_muted && !current_frame_muted) {
|
||||
// Not muted, don't touch.
|
||||
} else if (previous_frame_muted && current_frame_muted) {
|
||||
// Frame fully muted.
|
||||
size_t total_samples = frame->samples_per_channel_ * frame->num_channels_;
|
||||
RTC_DCHECK_GE(AudioFrame::kMaxDataSizeSamples, total_samples);
|
||||
frame->Mute();
|
||||
} else {
|
||||
// Fade is a no-op on a muted frame.
|
||||
if (frame->muted()) {
|
||||
return;
|
||||
}
|
||||
|
||||
// Limit number of samples to fade, if frame isn't long enough.
|
||||
size_t count = kMuteFadeFrames;
|
||||
float inc = kMuteFadeInc;
|
||||
if (frame->samples_per_channel_ < kMuteFadeFrames) {
|
||||
count = frame->samples_per_channel_;
|
||||
if (count > 0) {
|
||||
inc = 1.0f / count;
|
||||
}
|
||||
}
|
||||
|
||||
size_t start = 0;
|
||||
size_t end = count;
|
||||
float start_g = 0.0f;
|
||||
if (current_frame_muted) {
|
||||
// Fade out the last |count| samples of frame.
|
||||
RTC_DCHECK(!previous_frame_muted);
|
||||
start = frame->samples_per_channel_ - count;
|
||||
end = frame->samples_per_channel_;
|
||||
start_g = 1.0f;
|
||||
inc = -inc;
|
||||
} else {
|
||||
// Fade in the first |count| samples of frame.
|
||||
RTC_DCHECK(previous_frame_muted);
|
||||
}
|
||||
|
||||
// Perform fade.
|
||||
int16_t* frame_data = frame->mutable_data();
|
||||
size_t channels = frame->num_channels_;
|
||||
for (size_t j = 0; j < channels; ++j) {
|
||||
float g = start_g;
|
||||
for (size_t i = start * channels; i < end * channels; i += channels) {
|
||||
g += inc;
|
||||
frame_data[i + j] *= g;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
void AudioFrameOperations::Mute(AudioFrame* frame) {
|
||||
Mute(frame, true, true);
|
||||
}
|
||||
|
||||
void AudioFrameOperations::ApplyHalfGain(AudioFrame* frame) {
|
||||
RTC_DCHECK(frame);
|
||||
RTC_DCHECK_GT(frame->num_channels_, 0);
|
||||
if (frame->num_channels_ < 1 || frame->muted()) {
|
||||
return;
|
||||
}
|
||||
|
||||
int16_t* frame_data = frame->mutable_data();
|
||||
for (size_t i = 0; i < frame->samples_per_channel_ * frame->num_channels_;
|
||||
i++) {
|
||||
frame_data[i] = frame_data[i] >> 1;
|
||||
}
|
||||
}
|
||||
|
||||
int AudioFrameOperations::Scale(float left, float right, AudioFrame* frame) {
|
||||
if (frame->num_channels_ != 2) {
|
||||
return -1;
|
||||
} else if (frame->muted()) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
int16_t* frame_data = frame->mutable_data();
|
||||
for (size_t i = 0; i < frame->samples_per_channel_; i++) {
|
||||
frame_data[2 * i] = static_cast<int16_t>(left * frame_data[2 * i]);
|
||||
frame_data[2 * i + 1] = static_cast<int16_t>(right * frame_data[2 * i + 1]);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
int AudioFrameOperations::ScaleWithSat(float scale, AudioFrame* frame) {
|
||||
if (frame->muted()) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
int16_t* frame_data = frame->mutable_data();
|
||||
for (size_t i = 0; i < frame->samples_per_channel_ * frame->num_channels_;
|
||||
i++) {
|
||||
frame_data[i] = rtc::saturated_cast<int16_t>(scale * frame_data[i]);
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
} // namespace webrtc
|
||||
105
webrtc/audio/utility/audio_frame_operations.h
Normal file
105
webrtc/audio/utility/audio_frame_operations.h
Normal file
@ -0,0 +1,105 @@
|
||||
/*
|
||||
* Copyright (c) 2012 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.
|
||||
*/
|
||||
|
||||
#ifndef AUDIO_UTILITY_AUDIO_FRAME_OPERATIONS_H_
|
||||
#define AUDIO_UTILITY_AUDIO_FRAME_OPERATIONS_H_
|
||||
|
||||
#include <stddef.h>
|
||||
#include <stdint.h>
|
||||
|
||||
#include "api/audio/audio_frame.h"
|
||||
#include "rtc_base/deprecation.h"
|
||||
|
||||
namespace webrtc {
|
||||
|
||||
// TODO(andrew): consolidate this with utility.h and audio_frame_manipulator.h.
|
||||
// Change reference parameters to pointers. Consider using a namespace rather
|
||||
// than a class.
|
||||
class AudioFrameOperations {
|
||||
public:
|
||||
// Add samples in |frame_to_add| with samples in |result_frame|
|
||||
// putting the results in |results_frame|. The fields
|
||||
// |vad_activity_| and |speech_type_| of the result frame are
|
||||
// updated. If |result_frame| is empty (|samples_per_channel_|==0),
|
||||
// the samples in |frame_to_add| are added to it. The number of
|
||||
// channels and number of samples per channel must match except when
|
||||
// |result_frame| is empty.
|
||||
static void Add(const AudioFrame& frame_to_add, AudioFrame* result_frame);
|
||||
|
||||
// |frame.num_channels_| will be updated. This version checks for sufficient
|
||||
// buffer size and that |num_channels_| is mono. Use UpmixChannels
|
||||
// instead. TODO(bugs.webrtc.org/8649): remove.
|
||||
RTC_DEPRECATED static int MonoToStereo(AudioFrame* frame);
|
||||
|
||||
// |frame.num_channels_| will be updated. This version checks that
|
||||
// |num_channels_| is stereo. Use DownmixChannels
|
||||
// instead. TODO(bugs.webrtc.org/8649): remove.
|
||||
RTC_DEPRECATED static int StereoToMono(AudioFrame* frame);
|
||||
|
||||
// Downmixes 4 channels |src_audio| to stereo |dst_audio|. This is an in-place
|
||||
// operation, meaning |src_audio| and |dst_audio| may point to the same
|
||||
// buffer.
|
||||
static void QuadToStereo(const int16_t* src_audio,
|
||||
size_t samples_per_channel,
|
||||
int16_t* dst_audio);
|
||||
|
||||
// |frame.num_channels_| will be updated. This version checks that
|
||||
// |num_channels_| is 4 channels.
|
||||
static int QuadToStereo(AudioFrame* frame);
|
||||
|
||||
// Downmixes |src_channels| |src_audio| to |dst_channels| |dst_audio|.
|
||||
// This is an in-place operation, meaning |src_audio| and |dst_audio|
|
||||
// may point to the same buffer. Supported channel combinations are
|
||||
// Stereo to Mono, Quad to Mono, and Quad to Stereo.
|
||||
static void DownmixChannels(const int16_t* src_audio,
|
||||
size_t src_channels,
|
||||
size_t samples_per_channel,
|
||||
size_t dst_channels,
|
||||
int16_t* dst_audio);
|
||||
|
||||
// |frame.num_channels_| will be updated. This version checks that
|
||||
// |num_channels_| and |dst_channels| are valid and performs relevant downmix.
|
||||
// Supported channel combinations are N channels to Mono, and Quad to Stereo.
|
||||
static void DownmixChannels(size_t dst_channels, AudioFrame* frame);
|
||||
|
||||
// |frame.num_channels_| will be updated. This version checks that
|
||||
// |num_channels_| and |dst_channels| are valid and performs relevant
|
||||
// downmix. Supported channel combinations are Mono to N
|
||||
// channels. The single channel is replicated.
|
||||
static void UpmixChannels(size_t target_number_of_channels,
|
||||
AudioFrame* frame);
|
||||
|
||||
// Swap the left and right channels of |frame|. Fails silently if |frame| is
|
||||
// not stereo.
|
||||
static void SwapStereoChannels(AudioFrame* frame);
|
||||
|
||||
// Conditionally zero out contents of |frame| for implementing audio mute:
|
||||
// |previous_frame_muted| && |current_frame_muted| - Zero out whole frame.
|
||||
// |previous_frame_muted| && !|current_frame_muted| - Fade-in at frame start.
|
||||
// !|previous_frame_muted| && |current_frame_muted| - Fade-out at frame end.
|
||||
// !|previous_frame_muted| && !|current_frame_muted| - Leave frame untouched.
|
||||
static void Mute(AudioFrame* frame,
|
||||
bool previous_frame_muted,
|
||||
bool current_frame_muted);
|
||||
|
||||
// Zero out contents of frame.
|
||||
static void Mute(AudioFrame* frame);
|
||||
|
||||
// Halve samples in |frame|.
|
||||
static void ApplyHalfGain(AudioFrame* frame);
|
||||
|
||||
static int Scale(float left, float right, AudioFrame* frame);
|
||||
|
||||
static int ScaleWithSat(float scale, AudioFrame* frame);
|
||||
};
|
||||
|
||||
} // namespace webrtc
|
||||
|
||||
#endif // AUDIO_UTILITY_AUDIO_FRAME_OPERATIONS_H_
|
||||
Reference in New Issue
Block a user