Add NetEq::FilteredCurrentDelayMs() and use it in VoiceEngine

The new method returns the current total delay (packet buffer and sync
buffer) in ms, with smoothing applied to even out short-time
fluctuations due to jitter. The packet buffer part of the delay is not
updated during DTX/CNG periods.

This CL also pipes the new metric through ACM and uses it in
VoiceEngine. It replaces the previous method of estimating the buffer
delay (where an inserted packet's RTP timestamp was compared with the
last played timestamp from NetEq). The new method works better under
periods of DTX/CNG.

Review-Url: https://codereview.webrtc.org/2262203002
Cr-Commit-Position: refs/heads/master@{#13855}

webrtc/modules/audio_coding/acm2/acm_receiver.cc

@@ -309,6 +309,10 @@ rtc::Optional<uint32_t> AcmReceiver::GetPlayoutTimestamp() {
   return neteq_->GetPlayoutTimestamp();
 }
 
+int AcmReceiver::FilteredCurrentDelayMs() const {
+  return neteq_->FilteredCurrentDelayMs();
+}
+
 int AcmReceiver::LastAudioCodec(CodecInst* codec) const {
   rtc::CritScope lock(&crit_sect_);
   if (!last_audio_decoder_) {

webrtc/modules/audio_coding/acm2/acm_receiver.h

@@ -201,6 +201,13 @@ class AcmReceiver {
   // The return value will be empty if no valid timestamp is available.
   rtc::Optional<uint32_t> GetPlayoutTimestamp();
 
+  // Returns the current total delay from NetEq (packet buffer and sync buffer)
+  // in ms, with smoothing applied to even out short-time fluctuations due to
+  // jitter. The packet buffer part of the delay is not updated during DTX/CNG
+  // periods.
+  //
+  int FilteredCurrentDelayMs() const;
+
   //
   // Get the audio codec associated with the last non-CNG/non-DTMF received
   // payload. If no non-CNG/non-DTMF packet is received -1 is returned,

webrtc/modules/audio_coding/acm2/audio_coding_module.cc

@@ -159,6 +159,8 @@ class AudioCodingModuleImpl final : public AudioCodingModule {
 
   rtc::Optional<uint32_t> PlayoutTimestamp() override;
 
+  int FilteredCurrentDelayMs() const override;
+
   // Get 10 milliseconds of raw audio data to play out, and
   // automatic resample to the requested frequency if > 0.
   int PlayoutData10Ms(int desired_freq_hz,
@@ -1225,6 +1227,10 @@ rtc::Optional<uint32_t> AudioCodingModuleImpl::PlayoutTimestamp() {
   return receiver_.GetPlayoutTimestamp();
 }
 
+int AudioCodingModuleImpl::FilteredCurrentDelayMs() const {
+  return receiver_.FilteredCurrentDelayMs();
+}
+
 bool AudioCodingModuleImpl::HaveValidEncoder(const char* caller_name) const {
   if (!encoder_stack_) {
     WEBRTC_TRACE(webrtc::kTraceError, webrtc::kTraceAudioCoding, id_,

webrtc/modules/audio_coding/include/audio_coding_module.h

@@ -678,6 +678,15 @@ class AudioCodingModule {
   //
   virtual rtc::Optional<uint32_t> PlayoutTimestamp() = 0;
 
+  ///////////////////////////////////////////////////////////////////////////
+  // int FilteredCurrentDelayMs()
+  // Returns the current total delay from NetEq (packet buffer and sync buffer)
+  // in ms, with smoothing applied to even out short-time fluctuations due to
+  // jitter. The packet buffer part of the delay is not updated during DTX/CNG
+  // periods.
+  //
+  virtual int FilteredCurrentDelayMs() const = 0;
+
   ///////////////////////////////////////////////////////////////////////////
   // int32_t PlayoutData10Ms(
   // Get 10 milliseconds of raw audio data for playout, at the given sampling

webrtc/modules/audio_coding/neteq/include/neteq.h

@@ -223,6 +223,11 @@ class NetEq {
   // Returns the current total delay (packet buffer and sync buffer) in ms.
   virtual int CurrentDelayMs() const = 0;
 
+  // Returns the current total delay (packet buffer and sync buffer) in ms,
+  // with smoothing applied to even out short-time fluctuations due to jitter.
+  // The packet buffer part of the delay is not updated during DTX/CNG periods.
+  virtual int FilteredCurrentDelayMs() const = 0;
+
   // Sets the playout mode to |mode|.
   // Deprecated. Set the mode in the Config struct passed to the constructor.
   // TODO(henrik.lundin) Delete.

webrtc/modules/audio_coding/neteq/neteq_impl.cc

@@ -353,6 +353,22 @@ int NetEqImpl::CurrentDelayMs() const {
   return delay_ms;
 }
 
+int NetEqImpl::FilteredCurrentDelayMs() const {
+  rtc::CritScope lock(&crit_sect_);
+  // Calculate the filtered packet buffer level in samples. The value from
+  // |buffer_level_filter_| is in number of packets, represented in Q8.
+  const size_t packet_buffer_samples =
+      (buffer_level_filter_->filtered_current_level() *
+       decoder_frame_length_) >>
+      8;
+  // Sum up the filtered packet buffer level with the future length of the sync
+  // buffer, and divide the sum by the sample rate.
+  const size_t delay_samples =
+      packet_buffer_samples + sync_buffer_->FutureLength();
+  // The division below will truncate. The return value is in ms.
+  return static_cast<int>(delay_samples) / rtc::CheckedDivExact(fs_hz_, 1000);
+}
+
 // Deprecated.
 // TODO(henrik.lundin) Delete.
 void NetEqImpl::SetPlayoutMode(NetEqPlayoutMode mode) {

webrtc/modules/audio_coding/neteq/neteq_impl.h

@@ -146,6 +146,8 @@ class NetEqImpl : public webrtc::NetEq {
 
   int CurrentDelayMs() const override;
 
+  int FilteredCurrentDelayMs() const override;
+
   // Sets the playout mode to |mode|.
   // Deprecated.
   // TODO(henrik.lundin) Delete.

webrtc/voice_engine/channel.cc

@@ -3173,11 +3173,7 @@ void Channel::GetDecodingCallStatistics(AudioDecodingCallStats* stats) const {
 bool Channel::GetDelayEstimate(int* jitter_buffer_delay_ms,
                                int* playout_buffer_delay_ms) const {
   rtc::CritScope lock(&video_sync_lock_);
-  if (_average_jitter_buffer_delay_us == 0) {
+  *jitter_buffer_delay_ms = audio_coding_->FilteredCurrentDelayMs();
-    return false;
-  }
-  *jitter_buffer_delay_ms =
-      (_average_jitter_buffer_delay_us + 500) / 1000 + _recPacketDelayMs;
   *playout_buffer_delay_ms = playout_delay_ms_;
   return true;
 }
@@ -3390,6 +3386,9 @@ void Channel::UpdatePlayoutTimestamp(bool rtcp) {
 }
 
 // Called for incoming RTP packets after successful RTP header parsing.
+// TODO(henrik.lundin): Clean out this method. With the introduction of
+// AudioCoding::FilteredCurrentDelayMs() most (if not all) of this method can
+// be deleted, along with a few member variables. (WebRTC issue 6237.)
 void Channel::UpdatePacketDelay(uint32_t rtp_timestamp,
                                 uint16_t sequence_number) {
   WEBRTC_TRACE(kTraceStream, kTraceVoice, VoEId(_instanceId, _channelId),