Compensate for the IntelligibilityEnhancer processing delay in high bands

webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h

Index: webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
diff --git a/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h b/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
index 3af1190cf0c167f0ab53e16e157f6df5ae916fbc..c8a4133883a78ed010ea6f4209647e33421918c9 100644
--- a/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
+++ b/webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h
@@ -16,8 +16,10 @@
 #include <vector>
 
 #include "webrtc/base/swap_queue.h"
-#include "webrtc/common_audio/lapped_transform.h"
+#include "webrtc/common_audio/audio_ring_buffer.h"
 #include "webrtc/common_audio/channel_buffer.h"
+#include "webrtc/common_audio/lapped_transform.h"
+#include "webrtc/modules/audio_processing/audio_buffer.h"
 #include "webrtc/modules/audio_processing/intelligibility/intelligibility_utils.h"
 #include "webrtc/modules/audio_processing/render_queue_item_verifier.h"
 #include "webrtc/modules/audio_processing/vad/voice_activity_detector.h"
@@ -33,6 +35,7 @@ class IntelligibilityEnhancer : public LappedTransform::Callback {
  public:
   IntelligibilityEnhancer(int sample_rate_hz,
                           size_t num_render_channels,
+                          size_t num_bands,
                           size_t num_noise_bins);
 
   ~IntelligibilityEnhancer() override;
@@ -41,9 +44,7 @@ class IntelligibilityEnhancer : public LappedTransform::Callback {
   void SetCaptureNoiseEstimate(std::vector<float> noise, float gain);
 
   // Reads chunk of speech in time domain and updates with modified signal.
-  void ProcessRenderAudio(float* const* audio,
-                          int sample_rate_hz,
-                          size_t num_channels);
+  void ProcessRenderAudio(AudioBuffer* audio);
   bool active() const;
 
  protected:
@@ -56,6 +57,7 @@ class IntelligibilityEnhancer : public LappedTransform::Callback {
                          std::complex<float>* const* out_block) override;
 
  private:
+  FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest, TestRenderUpdate);
   FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest, TestErbCreation);
   FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest, TestSolveForGains);
   FRIEND_TEST_ALL_PREFIXES(IntelligibilityEnhancerTest,
@@ -84,6 +86,10 @@ class IntelligibilityEnhancer : public LappedTransform::Callback {
   // Returns true if the audio is speech.
   bool IsSpeech(const float* audio);
 
+  // Delays the high bands to compensate for the processing delay in the low
+  // band.
+  void DelayHighBands(AudioBuffer* audio);
+
   static const size_t kMaxNumNoiseEstimatesToBuffer = 5;
 
   const size_t freqs_;         // Num frequencies in frequency domain.
@@ -120,6 +126,8 @@ class IntelligibilityEnhancer : public LappedTransform::Callback {
   std::vector<float> noise_estimation_buffer_;
   SwapQueue<std::vector<float>, RenderQueueItemVerifier<float>>
       noise_estimation_queue_;
+
+  std::vector<std::unique_ptr<AudioRingBuffer>> high_bands_buffers_;
 };
 
 }  // namespace webrtc