[Chromoting] Implement down mixing in AudioPump

remoting/host/audio_capturer_win.cc

Index: remoting/host/audio_capturer_win.cc
diff --git a/remoting/host/audio_capturer_win.cc b/remoting/host/audio_capturer_win.cc
index 2610e6e2b692eed2523b0dfe2047d3222b379996..361352a6cc3bf275b3216c294eabbe2af47934ea 100644
--- a/remoting/host/audio_capturer_win.cc
+++ b/remoting/host/audio_capturer_win.cc
@@ -21,7 +21,6 @@
 #include "remoting/host/win/default_audio_device_change_detector.h"
 
 namespace {
-const int kChannels = 2;
 const int kBytesPerSample = 2;
 const int kBitsPerSample = kBytesPerSample * 8;
 // Conversion factor from 100ns to 1ms.
@@ -39,6 +38,21 @@ const int kMinTimerInterval = 30;
 // Upper bound for the timer precision error, in milliseconds.
 // Timers are supposed to be accurate to 20ms, so we use 30ms to be safe.
 const int kMaxExpectedTimerLag = 30;
+
+// Chooses a similar and supported channel configuration. |channels| represents
+// the selected channels. It can be nullptr.
+void ChooseChannels(WORD* channel_count, DWORD* channels) {
+  // We supports mono to 7.1.
+  if (*channel_count >= 1 && *channel_count <= 8) {
+    return;
+  }
+
+  *channel_count = remoting::AudioPacket::CHANNELS_STEREO;
+  if (channels) {
+    *channels = SPEAKER_FRONT_LEFT | SPEAKER_FRONT_RIGHT;
+  }
+}
+
 }  // namespace
 
 namespace remoting {
@@ -163,11 +177,12 @@ bool AudioCapturerWin::Initialize() {
           wave_format_ex_->nSamplesPerSec);
 
       wave_format_ex_->wFormatTag = WAVE_FORMAT_PCM;
-      wave_format_ex_->nChannels = kChannels;
+      ChooseChannels(&wave_format_ex_->nChannels, nullptr);
       wave_format_ex_->wBitsPerSample = kBitsPerSample;
-      wave_format_ex_->nBlockAlign = kChannels * kBytesPerSample;
+      wave_format_ex_->nBlockAlign =
+          wave_format_ex_->nChannels * kBytesPerSample;
       wave_format_ex_->nAvgBytesPerSec =
-          sampling_rate_ * kChannels * kBytesPerSample;
+          sampling_rate_ * wave_format_ex_->nChannels * kBytesPerSample;
       break;
     case WAVE_FORMAT_EXTENSIBLE: {
       PWAVEFORMATEXTENSIBLE wave_format_extensible =
@@ -186,13 +201,15 @@ bool AudioCapturerWin::Initialize() {
         wave_format_extensible->SubFormat = KSDATAFORMAT_SUBTYPE_PCM;
         wave_format_extensible->Samples.wValidBitsPerSample = kBitsPerSample;
 
-        wave_format_extensible->Format.nChannels = kChannels;
+        ChooseChannels(&wave_format_extensible->Format.nChannels,
+                       &wave_format_extensible->dwChannelMask);
         wave_format_extensible->Format.nSamplesPerSec = sampling_rate_;
         wave_format_extensible->Format.wBitsPerSample = kBitsPerSample;
         wave_format_extensible->Format.nBlockAlign =
-            kChannels * kBytesPerSample;
+            wave_format_extensible->Format.nChannels * kBytesPerSample;
         wave_format_extensible->Format.nAvgBytesPerSec =
-            sampling_rate_ * kChannels * kBytesPerSample;
+            sampling_rate_ * wave_format_extensible->Format.nChannels *
+            kBytesPerSample;
       } else {
         LOG(ERROR) << "Failed to force 16-bit samples";
         return false;
@@ -233,7 +250,7 @@ bool AudioCapturerWin::Initialize() {
   }
 
   volume_filter_.ActivateBy(mm_device_.Get());
-  volume_filter_.Initialize(sampling_rate_, kChannels);
+  volume_filter_.Initialize(sampling_rate_, wave_format_ex_->nChannels);
 
   return true;
 }
@@ -280,7 +297,8 @@ void AudioCapturerWin::DoCapture() {
       packet->set_encoding(AudioPacket::ENCODING_RAW);
       packet->set_sampling_rate(sampling_rate_);
       packet->set_bytes_per_sample(AudioPacket::BYTES_PER_SAMPLE_2);
-      packet->set_channels(AudioPacket::CHANNELS_STEREO);
+      packet->set_channels(static_cast<AudioPacket::Channels>(
+          wave_format_ex_->nChannels));
 
       callback_.Run(std::move(packet));
     }