Implement tail processing for AudioNodes

third_party/WebKit/Source/platform/audio/HRTFPanner.cpp

 /*
  * Copyright (C) 2010, Google Inc. All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
  * are met:
  * 1.  Redistributions of source code must retain the above copyright
  *    notice, this list of conditions and the following disclaimer.
  * 2.  Redistributions in binary form must reproduce the above copyright
  *    notice, this list of conditions and the following disclaimer in the
@@ skipping 332 matching lines @@
   // We are assuming that the inherent smoothing in the HRTF processing is good
   // enough, and we don't want to increase the complexity of the HRTF panner by
   // 15-20 times.  (We need to compute one output sample for each possibly
   // different impulse response.  That N^2.  Previously, we used an FFT to do
   // them all at once for a complexity of N/log2(N).  Hence, N/log2(N) times
   // more complex.)
   Pan(desired_azimuth[0], elevation[0], input_bus, output_bus,
       frames_to_process, channel_interpretation);
 }
 
+bool HRTFPanner::RequiresTailProcessing() const {
+  // Always return true since the tail and latency are never zero.
+  return true;
+}
+
 double HRTFPanner::TailTime() const {
   // Because HRTFPanner is implemented with a DelayKernel and a FFTConvolver,
   // the tailTime of the HRTFPanner is the sum of the tailTime of the
   // DelayKernel and the tailTime of the FFTConvolver, which is
   // MaxDelayTimeSeconds and fftSize() / 2, respectively.
   return kMaxDelayTimeSeconds +
          (FftSize() / 2) / static_cast<double>(SampleRate());
 }
 
 double HRTFPanner::LatencyTime() const {
   // The latency of a FFTConvolver is also fftSize() / 2, and is in addition to
   // its tailTime of the same value.
   return (FftSize() / 2) / static_cast<double>(SampleRate());
 }
 
 }  // namespace blink