Update a ton of audio code to use size_t more correctly and in general reduce

webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h

Index: webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h
diff --git a/webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h b/webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h
index 3e1345af542a662e64b5f205b823e7ad36f6a26b..9e82fbc0a5b3419bd4d0a8a7cfe255458a4e985b 100644
--- a/webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h
+++ b/webrtc/voice_engine/test/auto_test/fakes/fake_media_process.h
@@ -18,10 +18,10 @@ class FakeMediaProcess : public webrtc::VoEMediaProcess {
   virtual void Process(int channel,
                        const webrtc::ProcessingTypes type,
                        int16_t audio_10ms[],
-                       int length,
+                       size_t length,
                        int sampling_freq_hz,
                        bool stereo) {
-    for (int i = 0; i < length; i++) {
+    for (size_t i = 0; i < length; i++) {
       if (!stereo) {
         audio_10ms[i] = static_cast<int16_t>(audio_10ms[i] *
             sin(2.0 * 3.14 * frequency * 400.0 / sampling_freq_hz));