Add raw audio support to Mp4Muxer.

The fourcc used is `sowt` for little endian 16 bit PCM and `twos` for big endian 16 bit PCM.

PiperOrigin-RevId: 724391195

libraries/muxer/src/main/java/androidx/media3/muxer/Boxes.java

@@ -685,6 +685,7 @@ import org.checkerframework.checker.nullness.qual.PolyNull;
   }
 
   /** Returns a codec specific box. */
+  @SuppressWarnings("MergeCases")
   public static ByteBuffer codecSpecificBox(Format format) {
     String mimeType = checkNotNull(format.sampleMimeType);
     switch (mimeType) {
@@ -697,6 +698,8 @@ import org.checkerframework.checker.nullness.qual.PolyNull;
         return damrBox(/* mode= */ (short) 0x83FF); // mode set: all enabled for AMR-WB
       case MimeTypes.AUDIO_OPUS:
         return dOpsBox(format);
+      case MimeTypes.AUDIO_RAW:
+        return ByteBuffer.allocate(0); // No codec specific box for raw audio.
       case MimeTypes.VIDEO_H263:
         return d263Box(format);
       case MimeTypes.VIDEO_H264:
@@ -1690,6 +1693,14 @@ import org.checkerframework.checker.nullness.qual.PolyNull;
         return "s263";
       case MimeTypes.AUDIO_OPUS:
         return "Opus";
+      case MimeTypes.AUDIO_RAW:
+        if (format.pcmEncoding == C.ENCODING_PCM_16BIT) {
+          return "sowt";
+        } else if (format.pcmEncoding == C.ENCODING_PCM_16BIT_BIG_ENDIAN) {
+          return "twos";
+        } else {
+          throw new IllegalArgumentException("Unsupported PCM encoding: " + format.pcmEncoding);
+        }
       case MimeTypes.VIDEO_H264:
         return "avc1";
       case MimeTypes.VIDEO_H265:

libraries/muxer/src/main/java/androidx/media3/muxer/FragmentedMp4Muxer.java

@@ -58,6 +58,7 @@ import java.nio.ByteBuffer;
  *         <li>AMR-WB (Wideband AMR)
  *         <li>Opus
  *         <li>Vorbis
+ *         <li>Raw Audio
  *       </ul>
  *   <li>Metadata
  * </ul>
@@ -158,7 +159,8 @@ public final class FragmentedMp4Muxer implements AutoCloseable {
           MimeTypes.AUDIO_AMR_NB,
           MimeTypes.AUDIO_AMR_WB,
           MimeTypes.AUDIO_OPUS,
-          MimeTypes.AUDIO_VORBIS);
+          MimeTypes.AUDIO_VORBIS,
+          MimeTypes.AUDIO_RAW);
 
   private final FragmentedMp4Writer fragmentedMp4Writer;
   private final MetadataCollector metadataCollector;

libraries/muxer/src/main/java/androidx/media3/muxer/Mp4Muxer.java

@@ -82,6 +82,7 @@ import org.checkerframework.checker.nullness.qual.EnsuresNonNull;
  *         <li>AMR-WB (Wideband AMR)
  *         <li>Opus
  *         <li>Vorbis
+ *         <li>Raw Audio
  *       </ul>
  *   <li>Metadata
  * </ul>
@@ -361,7 +362,8 @@ public final class Mp4Muxer implements AutoCloseable {
           MimeTypes.AUDIO_AMR_NB,
           MimeTypes.AUDIO_AMR_WB,
           MimeTypes.AUDIO_OPUS,
-          MimeTypes.AUDIO_VORBIS);
+          MimeTypes.AUDIO_VORBIS,
+          MimeTypes.AUDIO_RAW);
 
   private static final String TAG = "Mp4Muxer";
 

libraries/muxer/src/test/java/androidx/media3/muxer/Mp4MuxerEndToEndParameterizedTest.java

@@ -73,6 +73,7 @@ public class Mp4MuxerEndToEndParameterizedTest {
   private static final String AMR_WB_3GP = "bbb_mono_16kHz_23.05kbps_amrwb.3gp";
   private static final String OPUS_OGG = "bbb_6ch_8kHz_opus.ogg";
   private static final String VORBIS_OGG = "bbb_1ch_16kHz_q10_vorbis.ogg";
+  private static final String RAW_WAV = "bbb_2ch_44kHz.wav";
 
   @Parameters(name = "{0}")
   public static ImmutableList<String> mediaSamples() {
@@ -92,7 +93,8 @@ public class Mp4MuxerEndToEndParameterizedTest {
         AMR_NB_3GP,
         AMR_WB_3GP,
         OPUS_OGG,
-        VORBIS_OGG);
+        VORBIS_OGG,
+        RAW_WAV);
   }
 
   @Parameter public @MonotonicNonNull String inputFile;

libraries/test_data/src/test/assets/muxerdumps/bbb_2ch_44kHz.wav.dump

@@ -0,0 +1,63 @@
+seekMap:
+  isSeekable = true
+  duration = 1000000
+  getPosition(0) = [[timeUs=0, position=400052]]
+  getPosition(1) = [[timeUs=1, position=400052]]
+  getPosition(500000) = [[timeUs=500000, position=488252]]
+  getPosition(1000000) = [[timeUs=1000000, position=558812]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 40
+  sample count = 10
+  track duration = 1000000
+  format 0:
+    id = 1
+    containerMimeType = audio/mp4
+    sampleMimeType = audio/raw
+    maxInputSize = 34
+    channelCount = 2
+    sampleRate = 44100
+    pcmEncoding = 2
+    language = ```
+    metadata = entries=[Mp4Timestamp: creation time=100000000, modification time=500000000, timescale=10000]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 4, hash F0FD1
+  sample 1:
+    time = 100000
+    flags = 1
+    data = length 4, hash FFDB3408
+  sample 2:
+    time = 200000
+    flags = 1
+    data = length 4, hash 47510F
+  sample 3:
+    time = 300000
+    flags = 1
+    data = length 4, hash 35B613
+  sample 4:
+    time = 400000
+    flags = 1
+    data = length 4, hash 333956
+  sample 5:
+    time = 500000
+    flags = 1
+    data = length 4, hash FFF14579
+  sample 6:
+    time = 600000
+    flags = 1
+    data = length 4, hash 3620FE
+  sample 7:
+    time = 700000
+    flags = 1
+    data = length 4, hash C488D
+  sample 8:
+    time = 800000
+    flags = 1
+    data = length 4, hash FFF6FBD8
+  sample 9:
+    time = 900000
+    flags = 536870913
+    data = length 4, hash FFDD4C2A
+tracksEnded = true