webrtc/audio
webrtc/common_audio
webrtc/media/engine
webrtc/voice_engine
webrtc/module/audio_coding
webrtc/module/audio_conference_mixer
webrtc/module/audio_device
webrtc/module/audio_processing

src/audio/audio_state.cc

void AudioState::AddSendingStream(webrtc::AudioSendStream* stream,
                                  int sample_rate_hz,
                                  size_t num_channels) {
  RTC_DCHECK(thread_checker_.CalledOnValidThread());
  auto& properties = sending_streams_[stream];
  properties.sample_rate_hz = sample_rate_hz;
  properties.num_channels = num_channels;
  UpdateAudioTransportWithSendingStreams();

  // Make sure recording is initialized; start recording if enabled.
  auto* adm = config_.audio_device_module.get();
  if (!adm->Recording()) {
    if (adm->InitRecording() == 0) {
      if (recording_enabled_) {
        adm->StartRecording();
      }
    } else {
      RTC_DLOG_F(LS_ERROR) << "Failed to initialize recording.";
    }
  }
}



src/media/engine/webrtcvoiceengine.cc

void WebRtcVoiceEngine::Init() {
...
#if defined(WEBRTC_INCLUDE_INTERNAL_AUDIO_DEVICE)
  // No ADM supplied? Create a default one.
  if (!adm_) {
    adm_ = webrtc::AudioDeviceModule::Create(
        webrtc::AudioDeviceModule::kPlatformDefaultAudio);
  }
#endif  // WEBRTC_INCLUDE_INTERNAL_AUDIO_DEVICE
  RTC_CHECK(adm());
  webrtc::adm_helpers::Init(adm());
  webrtc::apm_helpers::Init(apm());

  // Set up AudioState.
  {
    webrtc::AudioState::Config config;
    if (audio_mixer_) {
      config.audio_mixer = audio_mixer_;
    } else {
      config.audio_mixer = webrtc::AudioMixerImpl::Create();
    }
    config.audio_processing = apm_;
    config.audio_device_module = adm_;
    audio_state_ = webrtc::AudioState::Create(config);
  }

  // Connect the ADM to our audio path.
  adm()->RegisterAudioCallback(audio_state()->audio_transport());

  // Save the default AGC configuration settings. This must happen before
  // calling ApplyOptions or the default will be overwritten.
  default_agc_config_ = webrtc::apm_helpers::GetAgcConfig(apm());

  // Set default engine options.
  {
    AudioOptions options;
    options.echo_cancellation = true;
    options.auto_gain_control = true;
    options.noise_suppression = true;
    options.highpass_filter = true;
    options.stereo_swapping = false;
    options.audio_jitter_buffer_max_packets = 50;
    options.audio_jitter_buffer_fast_accelerate = false;
    options.audio_jitter_buffer_min_delay_ms = 0;
    options.typing_detection = true;
    options.experimental_agc = false;
    options.extended_filter_aec = false;
    options.delay_agnostic_aec = false;
    options.experimental_ns = false;
    options.residual_echo_detector = true;
    bool error = ApplyOptions(options);
    RTC_DCHECK(error);
  }

  initialized_ = true;
}

enum AudioLayer {
    kPlatformDefaultAudio = 0,
    kWindowsCoreAudio,
    kWindowsCoreAudio2,  // experimental
    kLinuxAlsaAudio,
    kLinuxPulseAudio,
    kAndroidJavaAudio,
    kAndroidOpenSLESAudio,
    kAndroidJavaInputAndOpenSLESOutputAudio,
    kAndroidAAudioAudio,
    kAndroidJavaInputAndAAudioOutputAudio,
    kDummyAudio,
  };

rtc::scoped_refptr<AudioDeviceModule> AudioDeviceModule::Create(
    const AudioLayer audio_layer) {
  RTC_LOG(INFO) << __FUNCTION__;
  return AudioDeviceModule::CreateForTest(audio_layer);
}
rtc::scoped_refptr<AudioDeviceModuleForTest> AudioDeviceModule::CreateForTest(
    const AudioLayer audio_layer) {
  RTC_LOG(INFO) << __FUNCTION__;

  // The "AudioDeviceModule::kWindowsCoreAudio2" audio layer has its own
  // dedicated factory method which should be used instead.
  if (audio_layer == AudioDeviceModule::kWindowsCoreAudio2) {
    RTC_LOG(LS_ERROR) << "Use the CreateWindowsCoreAudioAudioDeviceModule() "
                        "factory method instead for this option.";
    return nullptr;
  }

  // Create the generic reference counted (platform independent) implementation.
  rtc::scoped_refptr<AudioDeviceModuleImpl> audioDevice(
      new rtc::RefCountedObject<AudioDeviceModuleImpl>(audio_layer));

  // Ensure that the current platform is supported.
  if (audioDevice->CheckPlatform() == -1) {
    return nullptr;
  }

  // Create the platform-dependent implementation.
  if (audioDevice->CreatePlatformSpecificObjects() == -1) {
    return nullptr;
  }

  // Ensure that the generic audio buffer can communicate with the platform
  // specific parts.
  if (audioDevice->AttachAudioBuffer() == -1) {
    return nullptr;
  }

  return audioDevice;
}



src/modules/audio_device/audio_device_impl.cc
int32_t AudioDeviceModuleImpl::StartRecording() {
  RTC_LOG(INFO) << __FUNCTION__;
  CHECKinitialized_();
  if (Recording()) {
    return 0;
  }
  audio_device_buffer_.StartRecording();
  int32_t result = audio_device_->StartRecording();
  RTC_LOG(INFO) << "output: " << result;
  RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.StartRecordingSuccess",
                        static_cast<int>(result == 0));
  return result;
}

src/sdk/objc/api/peerconnection/RTCPeerConnectionFactory.mm

- (instancetype)
    initWithNativeAudioEncoderFactory:... {
...
    std::unique_ptr<cricket::MediaEngineInterface> media_engine =
        cricket::WebRtcMediaEngineFactory::Create(audioDeviceModule,
                                                  audioEncoderFactory,
                                                  audioDecoderFactory,
                                                  std::move(videoEncoderFactory),
                                                  std::move(videoDecoderFactory),
                                                  nullptr,  // audio mixer
                                                  audioProcessingModule);
...
  return self;
#endif
}

- (rtc::scoped_refptr<webrtc::AudioDeviceModule>)audioDeviceModule {
#if defined(WEBRTC_IOS)
  return webrtc::CreateAudioDeviceModule();
#else
  return nullptr;
#endif
}

src/sdk/objc/native/api/audio_device_module.mm
namespace webrtc {
rtc::scoped_refptr<AudioDeviceModule> CreateAudioDeviceModule() {
  RTC_LOG(INFO) << __FUNCTION__;
#if defined(WEBRTC_IOS)
  return new rtc::RefCountedObject<ios_adm::AudioDeviceModuleIOS>();
#else
  RTC_LOG(LERROR)
      << "current platform is not supported => this module will self destruct!";
  return nullptr;
#endif
}
}

src/sdk/objc/native/src/audio/audio_device_module_ios.mm
  int32_t AudioDeviceModuleIOS::Init() {
    RTC_LOG(INFO) << __FUNCTION__;
    if (initialized_)
      return 0;

    audio_device_buffer_.reset(new webrtc::AudioDeviceBuffer());
    audio_device_.reset(new ios_adm::AudioDeviceIOS());
    RTC_CHECK(audio_device_);

    this->AttachAudioBuffer();

    AudioDeviceGeneric::InitStatus status = audio_device_->Init();
    RTC_HISTOGRAM_ENUMERATION(
        "WebRTC.Audio.InitializationResult", static_cast<int>(status),
        static_cast<int>(AudioDeviceGeneric::InitStatus::NUM_STATUSES));
    if (status != AudioDeviceGeneric::InitStatus::OK) {
      RTC_LOG(LS_ERROR) << "Audio device initialization failed.";
      return -1;
    }
    initialized_ = true;
    return 0;
  }

src/sdk/objc/native/src/audio/audio_device_ios.mm
  fine_audio_buffer_.reset(new FineAudioBuffer(audio_device_buffer_));

src/sdk/objc/native/src/audio/audio_device_ios.mm

OSStatus AudioDeviceIOS::OnDeliverRecordedData(AudioUnitRenderActionFlags* flags,
                                              const AudioTimeStamp* time_stamp,
                                              UInt32 bus_number,
                                              UInt32 num_frames,
                                              AudioBufferList* /* io_data */) {
  RTC_DCHECK_RUN_ON(&io_thread_checker_);
  OSStatus result = noErr;
  // Simply return if recording is not enabled.
  if (!rtc::AtomicOps::AcquireLoad(&recording_)) return result;

  // Set the size of our own audio buffer and clear it first to avoid copying
  // in combination with potential reallocations.
  // On real iOS devices, the size will only be set once (at first callback).
  record_audio_buffer_.Clear();
  record_audio_buffer_.SetSize(num_frames);

  // Allocate AudioBuffers to be used as storage for the received audio.
  // The AudioBufferList structure works as a placeholder for the
  // AudioBuffer structure, which holds a pointer to the actual data buffer
  // in |record_audio_buffer_|. Recorded audio will be rendered into this memory
  // at each input callback when calling AudioUnitRender().
  AudioBufferList audio_buffer_list;
  audio_buffer_list.mNumberBuffers = 1;
  AudioBuffer* audio_buffer = &audio_buffer_list.mBuffers[0];
  audio_buffer->mNumberChannels = record_parameters_.channels();
  audio_buffer->mDataByteSize =
      record_audio_buffer_.size() * VoiceProcessingAudioUnit::kBytesPerSample;
  audio_buffer->mData = reinterpret_cast<int8_t*>(record_audio_buffer_.data());

  // Obtain the recorded audio samples by initiating a rendering cycle.
  // Since it happens on the input bus, the |io_data| parameter is a reference
  // to the preallocated audio buffer list that the audio unit renders into.
  // We can make the audio unit provide a buffer instead in io_data, but we
  // currently just use our own.
  // TODO(henrika): should error handling be improved?
  result = audio_unit_->Render(flags, time_stamp, bus_number, num_frames, &audio_buffer_list);
  if (result != noErr) {
    RTCLogError(@"Failed to render audio.");
    return result;
  }

  // Get a pointer to the recorded audio and send it to the WebRTC ADB.
  // Use the FineAudioBuffer instance to convert between native buffer size
  // and the 10ms buffer size used by WebRTC.
  fine_audio_buffer_->DeliverRecordedData(record_audio_buffer_, kFixedRecordDelayEstimate);
  return noErr;
}

src/sdk/objc/native/src/audio/voice_processing_audio_unit.mm
OSStatus VoiceProcessingAudioUnit::Render(AudioUnitRenderActionFlags* flags,
                                          const AudioTimeStamp* time_stamp,
                                          UInt32 output_bus_number,
                                          UInt32 num_frames,
                                          AudioBufferList* io_data) {
  RTC_DCHECK(vpio_unit_) << "Init() not called.";

  OSStatus result = AudioUnitRender(vpio_unit_, flags, time_stamp,
                                    output_bus_number, num_frames, io_data);
  if (result != noErr) {
    RTCLogError(@"Failed to render audio unit. Error=%ld", (long)result);
  }
  return result;
}

src/modules/audio_device/fine_audio_buffer.cc

void FineAudioBuffer::DeliverRecordedData(
    rtc::ArrayView<const int16_t> audio_buffer,
    int record_delay_ms) {
  RTC_DCHECK(IsReadyForRecord());
  // Always append new data and grow the buffer when needed.
  record_buffer_.AppendData(audio_buffer.data(), audio_buffer.size());
  // Consume samples from buffer in chunks of 10ms until there is not
  // enough data left. The number of remaining samples in the cache is given by
  // the new size of the internal |record_buffer_|.
  const size_t num_elements_10ms =
      record_channels_ * record_samples_per_channel_10ms_;
  while (record_buffer_.size() >= num_elements_10ms) {
    audio_device_buffer_->SetRecordedBuffer(record_buffer_.data(),
                                            record_samples_per_channel_10ms_);
    audio_device_buffer_->SetVQEData(playout_delay_ms_, record_delay_ms);
    audio_device_buffer_->DeliverRecordedData();
    memmove(record_buffer_.data(), record_buffer_.data() + num_elements_10ms,
            (record_buffer_.size() - num_elements_10ms) * sizeof(int16_t));
    record_buffer_.SetSize(record_buffer_.size() - num_elements_10ms);
  }
}

src/modules/audio_device/audio_device_buffer.cc

int32_t AudioDeviceBuffer::DeliverRecordedData() {
  if (!audio_transport_cb_) {
    RTC_LOG(LS_WARNING) << "Invalid audio transport";
    return 0;
  }
  const size_t frames = rec_buffer_.size() / rec_channels_;
  const size_t bytes_per_frame = rec_channels_ * sizeof(int16_t);
  uint32_t new_mic_level_dummy = 0;
  uint32_t total_delay_ms = play_delay_ms_ + rec_delay_ms_;
  int32_t res = audio_transport_cb_->RecordedDataIsAvailable(
      rec_buffer_.data(), frames, bytes_per_frame, rec_channels_,
      rec_sample_rate_, total_delay_ms, 0, 0, typing_status_,
      new_mic_level_dummy);
  if (res == -1) {
    RTC_LOG(LS_ERROR) << "RecordedDataIsAvailable() failed";
  }
  return 0;
}

void WebRtcVoiceEngine::Init() {
...
src/media/engine/webrtcvoiceengine.cc
void WebRtcVoiceEngine::Init() {
...
  // Connect the ADM to our audio path.
  adm()->RegisterAudioCallback(audio_state()->audio_transport());
...
}

src/audio/audio_state.cc
AudioState::AudioState(const AudioState::Config& config)
    : config_(config),
      audio_transport_(config_.audio_mixer, config_.audio_processing.get()) {
  process_thread_checker_.DetachFromThread();
  RTC_DCHECK(config_.audio_mixer);
  RTC_DCHECK(config_.audio_device_module);
}

src/audio/audio_transport_impl.cc

// Not used in Chromium. Process captured audio and distribute to all sending
// streams, and try to do this at the lowest possible sample rate.
int32_t AudioTransportImpl::RecordedDataIsAvailable(
    const void* audio_data,
    const size_t number_of_frames,
    const size_t bytes_per_sample,
    const size_t number_of_channels,
    const uint32_t sample_rate,
    const uint32_t audio_delay_milliseconds,
    const int32_t /*clock_drift*/,
    const uint32_t /*volume*/,
    const bool key_pressed,
    uint32_t& /*new_mic_volume*/) {  // NOLINT: to avoid changing APIs
  RTC_DCHECK(audio_data);
  RTC_DCHECK_GE(number_of_channels, 1);
  RTC_DCHECK_LE(number_of_channels, 2);
  RTC_DCHECK_EQ(2 * number_of_channels, bytes_per_sample);
  RTC_DCHECK_GE(sample_rate, AudioProcessing::NativeRate::kSampleRate8kHz);
  // 100 = 1 second / data duration (10 ms).
  RTC_DCHECK_EQ(number_of_frames * 100, sample_rate);
  RTC_DCHECK_LE(bytes_per_sample * number_of_frames * number_of_channels,
                AudioFrame::kMaxDataSizeBytes);

  int send_sample_rate_hz = 0;
  size_t send_num_channels = 0;
  bool swap_stereo_channels = false;
  {
    rtc::CritScope lock(&capture_lock_);
    send_sample_rate_hz = send_sample_rate_hz_;
    send_num_channels = send_num_channels_;
    swap_stereo_channels = swap_stereo_channels_;
  }

  std::unique_ptr<AudioFrame> audio_frame(new AudioFrame());
  InitializeCaptureFrame(sample_rate, send_sample_rate_hz, number_of_channels,
                        send_num_channels, audio_frame.get());
  voe::RemixAndResample(static_cast<const int16_t*>(audio_data),
                        number_of_frames, number_of_channels, sample_rate,
                        &capture_resampler_, audio_frame.get());
  ProcessCaptureFrame(audio_delay_milliseconds, key_pressed,
                      swap_stereo_channels, audio_processing_,
                      audio_frame.get());

  // Typing detection (utilizes the APM/VAD decision). We let the VAD determine
  // if we're using this feature or not.
  // TODO(solenberg): is_enabled() takes a lock. Work around that.
  bool typing_detected = false;
  if (audio_processing_->voice_detection()->is_enabled()) {
    if (audio_frame->vad_activity_ != AudioFrame::kVadUnknown) {
      bool vad_active = audio_frame->vad_activity_ == AudioFrame::kVadActive;
      typing_detected = typing_detection_.Process(key_pressed, vad_active);
    }
  }

  // Measure audio level of speech after all processing.
  double sample_duration = static_cast<double>(number_of_frames) / sample_rate;
  audio_level_.ComputeLevel(*audio_frame.get(), sample_duration);

  // Copy frame and push to each sending stream. The copy is required since an
  // encoding task will be posted internally to each stream.
  {
    rtc::CritScope lock(&capture_lock_);
    typing_noise_detected_ = typing_detected;

    RTC_DCHECK_GT(audio_frame->samples_per_channel_, 0);
    if (!sending_streams_.empty()) {
      auto it = sending_streams_.begin();
      while (++it != sending_streams_.end()) {
        std::unique_ptr<AudioFrame> audio_frame_copy(new AudioFrame());
        audio_frame_copy->CopyFrom(*audio_frame.get());
        (*it)->SendAudioData(std::move(audio_frame_copy));
      }
      // Send the original frame to the first stream w/o copying.
      (*sending_streams_.begin())->SendAudioData(std::move(audio_frame));
    }
  }

  return 0;
}

ProcessCaptureFrame(audio_delay_milliseconds, key_pressed,
                    swap_stereo_channels, audio_processing_,
                    audio_frame.get());

if (!audioProcessingModule) audioProcessingModule = webrtc::AudioProcessingBuilder().Create();

src/modules/audio_processing/audio_processing_impl.cc

AudioProcessing* AudioProcessingBuilder::Create(const webrtc::Config& config) {
  AudioProcessingImpl* apm = new rtc::RefCountedObject<AudioProcessingImpl>(
      config, std::move(capture_post_processing_),
      std::move(render_pre_processing_), std::move(echo_control_factory_),
      std::move(echo_detector_), std::move(capture_analyzer_));
  if (apm->Initialize() != AudioProcessing::kNoError) {
    delete apm;
    apm = nullptr;
  }
  return apm;
}

src/modules/audio_processing/audio_processing_impl.cc

int AudioProcessingImpl::ProcessStream(AudioFrame* frame) {
...
  capture_.capture_audio->DeinterleaveFrom(frame);
  RETURN_ON_ERR(ProcessCaptureStreamLocked());
  capture_.capture_audio->InterleaveTo(
      frame, submodule_states_.CaptureMultiBandProcessingActive() ||
                 submodule_states_.CaptureFullBandProcessingActive());

src/media/engine/webrtcvoiceengine.cc
WebRtcAudioSendStream(...){
...
    stream_ = call_->CreateAudioSendStream(config_);
...
}

void ChannelSend::ProcessAndEncodeAudio(
    std::unique_ptr<AudioFrame> audio_frame) {
  RTC_DCHECK_RUNS_SERIALIZED(&audio_thread_race_checker_);
  // Avoid posting any new tasks if sending was already stopped in StopSend().
  rtc::CritScope cs(&encoder_queue_lock_);
  if (!encoder_queue_is_active_) {
    return;
  }
  // Profile time between when the audio frame is added to the task queue and
  // when the task is actually executed.
  audio_frame->UpdateProfileTimeStamp();
  encoder_queue_->PostTask(std::unique_ptr<rtc::QueuedTask>(
      new ProcessAndEncodeAudioTask(std::move(audio_frame), this)));
}

src/pc/peerconnectionfactory.cc

rtc::scoped_refptr<PeerConnectionInterface>
PeerConnectionFactory::CreatePeerConnection(
    const PeerConnectionInterface::RTCConfiguration& configuration,
    PeerConnectionDependencies dependencies) {
  RTC_DCHECK(signaling_thread_->IsCurrent());

  // Set internal defaults if optional dependencies are not set.
  if (!dependencies.cert_generator) {
    dependencies.cert_generator =
        absl::make_unique<rtc::RTCCertificateGenerator>(signaling_thread_,
                                                        network_thread_);
  }
  if (!dependencies.allocator) {
    network_thread_->Invoke<void>(RTC_FROM_HERE, [this, &configuration,
                                                  &dependencies]() {
      dependencies.allocator = absl::make_unique<cricket::BasicPortAllocator>(
          default_network_manager_.get(), default_socket_factory_.get(),
          configuration.turn_customizer);
    });
  }

  // TODO(zstein): Once chromium injects its own AsyncResolverFactory, set
  // |dependencies.async_resolver_factory| to a new
  // |rtc::BasicAsyncResolverFactory| if no factory is provided.

  network_thread_->Invoke<void>(
      RTC_FROM_HERE,
      rtc::Bind(&cricket::PortAllocator::SetNetworkIgnoreMask,
                dependencies.allocator.get(), options_.network_ignore_mask));

  std::unique_ptr<RtcEventLog> event_log =
      worker_thread_->Invoke<std::unique_ptr<RtcEventLog>>(
          RTC_FROM_HERE,
          rtc::Bind(&PeerConnectionFactory::CreateRtcEventLog_w, this));

  std::unique_ptr<Call> call = worker_thread_->Invoke<std::unique_ptr<Call>>(
      RTC_FROM_HERE,
      rtc::Bind(&PeerConnectionFactory::CreateCall_w, this, event_log.get()));

  rtc::scoped_refptr<PeerConnection> pc(
      new rtc::RefCountedObject<PeerConnection>(this, std::move(event_log),
                                                std::move(call)));
  ActionsBeforeInitializeForTesting(pc);
  if (!pc->Initialize(configuration, std::move(dependencies))) {
    return nullptr;
  }
  return PeerConnectionProxy::Create(signaling_thread(), pc);
}

Call* Call::Create(const Call::Config& config) {
  return new internal::Call(
      config, absl::make_unique<RtpTransportControllerSend>(
                  Clock::GetRealTimeClock(), config.event_log,
                  config.network_controller_factory, config.bitrate_config));
}

process_thread_->RegisterModule(&pacer_, RTC_FROM_HERE);
  process_thread_->RegisterModule(send_side_cc_.get(), RTC_FROM_HERE);

encoder_queue_->PostTask(std::unique_ptr<rtc::QueuedTask>(
      new ProcessAndEncodeAudioTask(std::move(audio_frame), this)));

src/audio/channel_send.cc

void ChannelSend::ProcessAndEncodeAudioOnTaskQueue(AudioFrame* audio_input) {
  RTC_DCHECK_RUN_ON(encoder_queue_);
  RTC_DCHECK_GT(audio_input->samples_per_channel_, 0);
  RTC_DCHECK_LE(audio_input->num_channels_, 2);

  // Measure time between when the audio frame is added to the task queue and
  // when the task is actually executed. Goal is to keep track of unwanted
  // extra latency added by the task queue.
  RTC_HISTOGRAM_COUNTS_10000("WebRTC.Audio.EncodingTaskQueueLatencyMs",
                            audio_input->ElapsedProfileTimeMs());

  bool is_muted = InputMute();
  AudioFrameOperations::Mute(audio_input, previous_frame_muted_, is_muted);

  if (_includeAudioLevelIndication) {
    size_t length =
        audio_input->samples_per_channel_ * audio_input->num_channels_;
    RTC_CHECK_LE(length, AudioFrame::kMaxDataSizeBytes);
    if (is_muted && previous_frame_muted_) {
      rms_level_.AnalyzeMuted(length);
    } else {
      rms_level_.Analyze(
          rtc::ArrayView<const int16_t>(audio_input->data(), length));
    }
  }
  previous_frame_muted_ = is_muted;

  // Add 10ms of raw (PCM) audio data to the encoder @ 32kHz.

  // The ACM resamples internally.
  audio_input->timestamp_ = _timeStamp;
  // This call will trigger AudioPacketizationCallback::SendData if encoding
  // is done and payload is ready for packetization and transmission.
  // Otherwise, it will return without invoking the callback.
  if (audio_coding_->Add10MsData(*audio_input) < 0) {
    RTC_DLOG(LS_ERROR) << "ACM::Add10MsData() failed.";
    return;
  }

  _timeStamp += static_cast<uint32_t>(audio_input->samples_per_channel_);
}

int32_t ChannelSend::SendData(FrameType frameType,
                              uint8_t payloadType,
                              uint32_t timeStamp,
                              const uint8_t* payloadData,
                              size_t payloadSize,
                              const RTPFragmentationHeader* fragmentation) {
  RTC_DCHECK_RUN_ON(encoder_queue_);
  rtc::ArrayView<const uint8_t> payload(payloadData, payloadSize);

  if (media_transport() != nullptr) {
    return SendMediaTransportAudio(frameType, payloadType, timeStamp, payload,
                                   fragmentation);
  } else {
    return SendRtpAudio(frameType, payloadType, timeStamp, payload,
                        fragmentation);
  }
}

src/api/peerconnectioninterface.h
PeerConnectionInterface::
// If MediaTransportFactory is provided in PeerConnectionFactory, this flag
    // informs PeerConnection that it should use the MediaTransportInterface.
    // It's invalid to set it to |true| if the MediaTransportFactory wasn't
    // provided.
    bool use_media_transport = false;

src/media/engine/webrtcvoiceengine.cc

bool WebRtcVoiceEngine::ApplyOptions(const AudioOptions& options_in) {
  RTC_DCHECK(worker_thread_checker_.CalledOnValidThread());
  RTC_LOG(LS_INFO) << "WebRtcVoiceEngine::ApplyOptions: "
                   << options_in.ToString();
  AudioOptions options = options_in;  // The options are modified below.

  // Set and adjust echo canceller options.
  // kEcConference is AEC with high suppression.
  webrtc::EcModes ec_mode = webrtc::kEcConference;

#if defined(WEBRTC_IOS)
  // wbt: 在ios上强制软件回声消除
  if (options.ios_force_software_aec_HACK &&
      *options.ios_force_software_aec_HACK) {
    // EC may be forced on for a device known to have non-functioning platform
    // AEC.
    options.echo_cancellation = true;
    options.extended_filter_aec = true;
    RTC_LOG(LS_WARNING)
        << "Force software AEC on iOS. May conflict with platform AEC.";
  } else {// wbt: 默认使用VPIO内置的回声消除
    // On iOS, VPIO provides built-in EC.
    options.echo_cancellation = false;
    options.extended_filter_aec = false;
    RTC_LOG(LS_INFO) << "Always disable AEC on iOS. Use built-in instead.";
  }
#elif defined(WEBRTC_ANDROID)
  ec_mode = webrtc::kEcAecm;
  options.extended_filter_aec = false;
#endif

  // wbt :除iOS平台外：如果启用了估计延时不确定性Delay Agnostic设置则自动开启aec
  // Delay Agnostic AEC automatically turns on EC if not set except on iOS
  // where the feature is not supported.
  bool use_delay_agnostic_aec = false;
#if !defined(WEBRTC_IOS)
  if (options.delay_agnostic_aec) {
    use_delay_agnostic_aec = *options.delay_agnostic_aec;
    if (use_delay_agnostic_aec) {
      options.echo_cancellation = true;
      options.extended_filter_aec = true;
      ec_mode = webrtc::kEcConference;
    }
  }
#endif

// Set and adjust noise suppressor options.
#if defined(WEBRTC_IOS)
  // On iOS, VPIO provides built-in NS.
  // wbt: iOS平台使用VIPIO内置的降噪。
  // 关闭键盘声音检测（因为触摸设备没有键盘)，并且关闭实验性降噪。
  options.noise_suppression = false;
  options.typing_detection = false;
  options.experimental_ns = false;
  RTC_LOG(LS_INFO) << "Always disable NS on iOS. Use built-in instead.";
#elif defined(WEBRTC_ANDROID)
  // wbt: Android平台
  // 关闭键盘声音检测（因为触摸设备没有键盘)，并且关闭实验性降噪。
  options.typing_detection = false;
  options.experimental_ns = false;
#endif

// Set and adjust gain control options.
#if defined(WEBRTC_IOS)
  // On iOS, VPIO provides built-in AGC.
  // wbt: iOS平台使用VIPIO内置的AGC增益。并且关闭试验性AGC.
  options.auto_gain_control = false;
  options.experimental_agc = false;
  RTC_LOG(LS_INFO) << "Always disable AGC on iOS. Use built-in instead.";
#elif defined(WEBRTC_ANDROID)
  // wbt: Android平台关闭试验性AGC.
  options.experimental_agc = false;
#endif

#if defined(WEBRTC_IOS) || defined(WEBRTC_ANDROID)
  // wbt: 手机平台
  // 如果设置了"WebRTC-Audio-MinimizeResamplingOnMobile"则关闭AGC。
  // 然后如果开启降噪且未开启回声消除则关闭高通滤波器。

  // Turn off the gain control if specified by the field trial.
  // The purpose of the field trial is to reduce the amount of resampling
  // performed inside the audio processing module on mobile platforms by
  // whenever possible turning off the fixed AGC mode and the high-pass filter.
  // (https://bugs.chromium.org/p/webrtc/issues/detail?id=6181).
  if (webrtc::field_trial::IsEnabled(
          "WebRTC-Audio-MinimizeResamplingOnMobile")) {
    options.auto_gain_control = false;
    RTC_LOG(LS_INFO) << "Disable AGC according to field trial.";
    if (!(options.noise_suppression.value_or(false) ||
          options.echo_cancellation.value_or(false))) {
      // If possible, turn off the high-pass filter.
      RTC_LOG(LS_INFO)
          << "Disable high-pass filter in response to field trial.";
      options.highpass_filter = false;
    }
  }
#endif

  if (options.echo_cancellation) {
    // Check if platform supports built-in EC. Currently only supported on
    // Android and in combination with Java based audio layer.
    // TODO(henrika): investigate possibility to support built-in EC also
    // in combination with Open SL ES audio.
    // wbt: 目前只有android支持内置aec
    // 如果支持内置aec而且开启回声消除echo_cancellation和未开启use_delay_agnostic_aec
    // 且内置EnableBuiltInAEC时关闭软件回声消除从而使用平台内置回声消除。
    const bool built_in_aec = adm()->BuiltInAECIsAvailable();//除了Android平台为true，其他平台默认为false
    if (built_in_aec) {
      // Built-in EC exists on this device and use_delay_agnostic_aec is not
      // overriding it. Enable/Disable it according to the echo_cancellation
      // audio option.
      const bool enable_built_in_aec =
          *options.echo_cancellation && !use_delay_agnostic_aec;
      if (adm()->EnableBuiltInAEC(enable_built_in_aec) == 0 &&
          enable_built_in_aec) {
        // Disable internal software EC if built-in EC is enabled,
        // i.e., replace the software EC with the built-in EC.
        options.echo_cancellation = false;
        RTC_LOG(LS_INFO)
            << "Disabling EC since built-in EC will be used instead";
      }
    }
    // 这里的ec_mode在上面会根据情况设置为默认会议积极的aec还是手机型aec。
    webrtc::apm_helpers::SetEcStatus(apm(), *options.echo_cancellation,
                                     ec_mode);
  }
  // 判断是否支持平台内置agc,如果支持则关闭软件agc
  if (options.auto_gain_control) {
    bool built_in_agc_avaliable = adm()->BuiltInAGCIsAvailable();
    if (built_in_agc_avaliable) {
      if (adm()->EnableBuiltInAGC(*options.auto_gain_control) == 0 &&
          *options.auto_gain_control) {
        // Disable internal software AGC if built-in AGC is enabled,
        // i.e., replace the software AGC with the built-in AGC.
        options.auto_gain_control = false;
        RTC_LOG(LS_INFO)
            << "Disabling AGC since built-in AGC will be used instead";
      }
    }
    webrtc::apm_helpers::SetAgcStatus(apm(), *options.auto_gain_control);
  }

  //agc参数设置
  if (options.tx_agc_target_dbov || options.tx_agc_digital_compression_gain ||
      options.tx_agc_limiter) {
    // Override default_agc_config_. Generally, an unset option means "leave
    // the VoE bits alone" in this function, so we want whatever is set to be
    // stored as the new "default". If we didn't, then setting e.g.
    // tx_agc_target_dbov would reset digital compression gain and limiter
    // settings.
    default_agc_config_.targetLeveldBOv = options.tx_agc_target_dbov.value_or(
        default_agc_config_.targetLeveldBOv);
    default_agc_config_.digitalCompressionGaindB =
        options.tx_agc_digital_compression_gain.value_or(
            default_agc_config_.digitalCompressionGaindB);
    default_agc_config_.limiterEnable =
        options.tx_agc_limiter.value_or(default_agc_config_.limiterEnable);
    webrtc::apm_helpers::SetAgcConfig(apm(), default_agc_config_);
  }

  // 如果支持平台内置降噪则关闭软件降噪
  if (options.noise_suppression) {
    if (adm()->BuiltInNSIsAvailable()) {
      bool builtin_ns = *options.noise_suppression;
      if (adm()->EnableBuiltInNS(builtin_ns) == 0 && builtin_ns) {
        // Disable internal software NS if built-in NS is enabled,
        // i.e., replace the software NS with the built-in NS.
        options.noise_suppression = false;
        RTC_LOG(LS_INFO)
            << "Disabling NS since built-in NS will be used instead";
      }
    }
    webrtc::apm_helpers::SetNsStatus(apm(), *options.noise_suppression);
  }

  // 立体声通道交换
  if (options.stereo_swapping) {
    RTC_LOG(LS_INFO) << "Stereo swapping enabled? " << *options.stereo_swapping;
    audio_state()->SetStereoChannelSwapping(*options.stereo_swapping);
  }

  // jitter buffer最大包数设置
  if (options.audio_jitter_buffer_max_packets) {
    RTC_LOG(LS_INFO) << "NetEq capacity is "
                     << *options.audio_jitter_buffer_max_packets;
    audio_jitter_buffer_max_packets_ =
        std::max(20, *options.audio_jitter_buffer_max_packets);
  }
  // jitter buffer加速设置
  if (options.audio_jitter_buffer_fast_accelerate) {
    RTC_LOG(LS_INFO) << "NetEq fast mode? "
                     << *options.audio_jitter_buffer_fast_accelerate;
    audio_jitter_buffer_fast_accelerate_ =
        *options.audio_jitter_buffer_fast_accelerate;
  }
  // jitter buffer最小延迟（毫秒）
  if (options.audio_jitter_buffer_min_delay_ms) {
    RTC_LOG(LS_INFO) << "NetEq minimum delay is "
                     << *options.audio_jitter_buffer_min_delay_ms;
    audio_jitter_buffer_min_delay_ms_ =
        *options.audio_jitter_buffer_min_delay_ms;
  }

  // 键盘声音检测
  if (options.typing_detection) {
    RTC_LOG(LS_INFO) << "Typing detection is enabled? "
                     << *options.typing_detection;
    webrtc::apm_helpers::SetTypingDetectionStatus(apm(),
                                                  *options.typing_detection);
  }

  webrtc::Config config;

  if (options.delay_agnostic_aec)
    delay_agnostic_aec_ = options.delay_agnostic_aec;
  if (delay_agnostic_aec_) {
    RTC_LOG(LS_INFO) << "Delay agnostic aec is enabled? "
                     << *delay_agnostic_aec_;
    config.Set<webrtc::DelayAgnostic>(
        new webrtc::DelayAgnostic(*delay_agnostic_aec_));
  }

  if (options.extended_filter_aec) {
    extended_filter_aec_ = options.extended_filter_aec;
  }
  if (extended_filter_aec_) {
    RTC_LOG(LS_INFO) << "Extended filter aec is enabled? "
                     << *extended_filter_aec_;
    config.Set<webrtc::ExtendedFilter>(
        new webrtc::ExtendedFilter(*extended_filter_aec_));
  }

  if (options.experimental_ns) {
    experimental_ns_ = options.experimental_ns;
  }
  if (experimental_ns_) {
    RTC_LOG(LS_INFO) << "Experimental ns is enabled? " << *experimental_ns_;
    config.Set<webrtc::ExperimentalNs>(
        new webrtc::ExperimentalNs(*experimental_ns_));
  }

  webrtc::AudioProcessing::Config apm_config = apm()->GetConfig();

  if (options.highpass_filter) {
    apm_config.high_pass_filter.enabled = *options.highpass_filter;
  }

  // 残留回声检测
  if (options.residual_echo_detector) {
    apm_config.residual_echo_detector.enabled = *options.residual_echo_detector;
  }

  apm()->SetExtraOptions(config);
  apm()->ApplyConfig(apm_config);
  return true;
}