From 54293fdad034eb35da373753e01b47598e32381b Mon Sep 17 00:00:00 2001
From: Ken Cooke <ken@highfidelity.io>
Date: Fri, 3 Jun 2016 18:39:43 -0700
Subject: [PATCH] Replace the linear interpolation embedded in the .WAV/.RAW
 loader with high quality polyphase resampling. When downsampling 48khz to
 24khz, linear interpolation creates aliasing distortion that is quite
 audible.  This change greatly improves the audio quality for all sound
 assets.

---
 libraries/audio/src/Sound.cpp | 55 ++++++++++-------------------------
 1 file changed, 15 insertions(+), 40 deletions(-)

diff --git a/libraries/audio/src/Sound.cpp b/libraries/audio/src/Sound.cpp
index 806e33819e..c3e936e33a 100644
--- a/libraries/audio/src/Sound.cpp
+++ b/libraries/audio/src/Sound.cpp
@@ -25,6 +25,8 @@
 
 #include "AudioRingBuffer.h"
 #include "AudioLogging.h"
+#include "AudioSRC.h"
+
 #include "Sound.h"
 
 QScriptValue soundSharedPointerToScriptValue(QScriptEngine* engine, const SharedSoundPointer& in) {
@@ -89,49 +91,22 @@ void Sound::downSample(const QByteArray& rawAudioByteArray) {
     // we want to convert it to the format that the audio-mixer wants
     // which is signed, 16-bit, 24Khz
 
-    int numSourceSamples = rawAudioByteArray.size() / sizeof(AudioConstants::AudioSample);
+    int numChannels = _isStereo ? 2 : 1;
+    AudioSRC resampler(48000, AudioConstants::SAMPLE_RATE, numChannels);
 
-    if (_isStereo && numSourceSamples % 2 != 0){
-        // in the unlikely case that we have stereo audio but we seem to be missing a sample
-        // (the sample for one channel is missing in a set of interleaved samples)
-        // then drop the odd sample
-        --numSourceSamples;
-    }
+    // resize to max possible output
+    int numSourceFrames = rawAudioByteArray.size() / (numChannels * sizeof(AudioConstants::AudioSample));
+    int maxDestinationFrames = resampler.getMaxOutput(numSourceFrames);
+    int maxDestinationBytes = maxDestinationFrames * numChannels * sizeof(AudioConstants::AudioSample);
+    _byteArray.resize(maxDestinationBytes);
 
-    int numDestinationSamples = numSourceSamples / 2.0f;
+    int numDestinationFrames = resampler.render((int16_t*)rawAudioByteArray.data(), 
+                                                (int16_t*)_byteArray.data(), 
+                                                numSourceFrames);
 
-    if (_isStereo && numDestinationSamples % 2 != 0) {
-        // if this is stereo we need to make sure we produce stereo output
-        // which means we should have an even number of output samples
-        numDestinationSamples += 1;
-    }
-
-    int numDestinationBytes = numDestinationSamples * sizeof(AudioConstants::AudioSample);
-
-    _byteArray.resize(numDestinationBytes);
-
-    int16_t* sourceSamples = (int16_t*) rawAudioByteArray.data();
-    int16_t* destinationSamples = (int16_t*) _byteArray.data();
-
-    if (_isStereo) {
-        for (int i = 0; i < numSourceSamples; i += 4) {
-            if (i + 2 >= numSourceSamples) {
-                destinationSamples[i / 2] = sourceSamples[i];
-                destinationSamples[(i / 2) + 1] = sourceSamples[i + 1];
-            } else {
-                destinationSamples[i / 2] = (sourceSamples[i] + sourceSamples[i + 2]) / 2;
-                destinationSamples[(i / 2) + 1] = (sourceSamples[i + 1] + sourceSamples[i + 3]) / 2;
-            }
-        }
-    } else {
-        for (int i = 1; i < numSourceSamples; i += 2) {
-            if (i + 1 >= numSourceSamples) {
-                destinationSamples[(i - 1) / 2] = (sourceSamples[i - 1] + sourceSamples[i]) / 2;
-            } else {
-                destinationSamples[(i - 1) / 2] = ((sourceSamples[i - 1] + sourceSamples[i + 1]) / 4) + (sourceSamples[i] / 2);
-            }
-        }
-    }
+    // truncate to actual output
+    int numDestinationBytes = numDestinationFrames * numChannels * sizeof(AudioConstants::AudioSample);
+    _byteArray.resize(maxDestinationBytes);
 }
 
 //