deprecate AudioSample and AudioBuffer

Start the transition to a full AVFrame-based pipeline
that would allow to support arbitrary sample formats,
like floating point or spatial audio.

Change-Id: I46cf567a60cd6cd2b01c7f18ccee5acfbe66b5eb

src/media/audio/audiobuffer.h

@@ -93,7 +93,8 @@ struct AudioFormat
 
 std::ostream& operator<<(std::ostream& stream, const AudioFormat& f);
 
-class AudioBuffer
+class
+[[deprecated("Use AudioFrame instead")]] AudioBuffer
 {
 public:
     /**

src/media/media_buffer.h

@@ -36,6 +36,9 @@ using MediaFrame = libjami::MediaFrame;
 using AudioFrame = libjami::AudioFrame;
 using MediaObserver = std::function<void(std::shared_ptr<MediaFrame>&&)>;
 
+[[deprecated("Use AudioFrame and AVSampleFormat")]]
+typedef int16_t AudioSample;
+
 #ifdef ENABLE_VIDEO
 
 using VideoFrame = libjami::VideoFrame;

src/ring_types.h

@@ -31,7 +31,6 @@
 
 namespace jami {
 
-typedef int16_t AudioSample;
 
 static constexpr size_t SIZEBUF = 16000; /** About 62.5ms of buffering at 48kHz */