Add dump tests for concatenating 2 audio items.

Audio only tests are now using RAW audio where possible, which is
passed through the Robolectric decoders/encoders, and can be handled by
the AudioProcessor instances accurately.

PiperOrigin-RevId: 541648853

libraries/test_data/src/test/assets/transformerdumps/mp4/sample.mp4.concatenated_audio_high_pitch.dump

@@ -1,207 +0,0 @@
-format 0:
-  averageBitrate = 131072
-  sampleMimeType = audio/mp4a-latm
-  channelCount = 1
-  sampleRate = 44100
-  pcmEncoding = 2
-  metadata = entries=[TSSE: description=null: values=[Lavf56.1.0], xyz: latitude=40.68, longitude=-74.5, Mp4Timestamp: creation time=3547558895, modification time=3547558895, timescale=1000]
-container metadata = entries=[TSSE: description=null: values=[Lavf56.1.0], xyz: latitude=40.68, longitude=-74.5, Mp4Timestamp: creation time=3547558895, modification time=3547558895, timescale=1000]
-sample:
-  trackIndex = 0
-  dataHashCode = 915609509
-  size = 792
-  isKeyFrame = true
-  presentationTimeUs = 0
-sample:
-  trackIndex = 0
-  dataHashCode = -1580893866
-  size = 678
-  isKeyFrame = true
-  presentationTimeUs = 8979
-sample:
-  trackIndex = 0
-  dataHashCode = -31547651
-  size = 304
-  isKeyFrame = true
-  presentationTimeUs = 16666
-sample:
-  trackIndex = 0
-  dataHashCode = 1415140636
-  size = 460
-  isKeyFrame = true
-  presentationTimeUs = 20113
-sample:
-  trackIndex = 0
-  dataHashCode = 1721060815
-  size = 850
-  isKeyFrame = true
-  presentationTimeUs = 25328
-sample:
-  trackIndex = 0
-  dataHashCode = 1707913464
-  size = 446
-  isKeyFrame = true
-  presentationTimeUs = 34965
-sample:
-  trackIndex = 0
-  dataHashCode = -776771764
-  size = 852
-  isKeyFrame = true
-  presentationTimeUs = 40022
-sample:
-  trackIndex = 0
-  dataHashCode = -609146892
-  size = 368
-  isKeyFrame = true
-  presentationTimeUs = 49682
-sample:
-  trackIndex = 0
-  dataHashCode = -2044977387
-  size = 1166
-  isKeyFrame = true
-  presentationTimeUs = 53854
-sample:
-  trackIndex = 0
-  dataHashCode = -753877175
-  size = 506
-  isKeyFrame = true
-  presentationTimeUs = 67074
-sample:
-  trackIndex = 0
-  dataHashCode = 1491046836
-  size = 578
-  isKeyFrame = true
-  presentationTimeUs = 72811
-sample:
-  trackIndex = 0
-  dataHashCode = 621394572
-  size = 668
-  isKeyFrame = true
-  presentationTimeUs = 79365
-sample:
-  trackIndex = 0
-  dataHashCode = -58393202
-  size = 268
-  isKeyFrame = true
-  presentationTimeUs = 86938
-sample:
-  trackIndex = 0
-  dataHashCode = 1253593269
-  size = 318
-  isKeyFrame = true
-  presentationTimeUs = 89977
-sample:
-  trackIndex = 0
-  dataHashCode = -1544714160
-  size = 424
-  isKeyFrame = true
-  presentationTimeUs = 93582
-sample:
-  trackIndex = 0
-  dataHashCode = -2038565545
-  size = 294
-  isKeyFrame = true
-  presentationTimeUs = 98390
-sample:
-  trackIndex = 0
-  dataHashCode = 803611858
-  size = 394
-  isKeyFrame = true
-  presentationTimeUs = 101723
-sample:
-  trackIndex = 0
-  dataHashCode = 890682839
-  size = 812
-  isKeyFrame = true
-  presentationTimeUs = 106190
-sample:
-  trackIndex = 0
-  dataHashCode = 1798765816
-  size = 332
-  isKeyFrame = true
-  presentationTimeUs = 115396
-sample:
-  trackIndex = 0
-  dataHashCode = -155329417
-  size = 250
-  isKeyFrame = true
-  presentationTimeUs = 119160
-sample:
-  trackIndex = 0
-  dataHashCode = 2061435630
-  size = 304
-  isKeyFrame = true
-  presentationTimeUs = 121995
-sample:
-  trackIndex = 0
-  dataHashCode = -667770092
-  size = 1318
-  isKeyFrame = true
-  presentationTimeUs = 125442
-sample:
-  trackIndex = 0
-  dataHashCode = 1947321516
-  size = 224
-  isKeyFrame = true
-  presentationTimeUs = 140385
-sample:
-  trackIndex = 0
-  dataHashCode = 1744495738
-  size = 446
-  isKeyFrame = true
-  presentationTimeUs = 142925
-sample:
-  trackIndex = 0
-  dataHashCode = 801488010
-  size = 838
-  isKeyFrame = true
-  presentationTimeUs = 147981
-sample:
-  trackIndex = 0
-  dataHashCode = -867204691
-  size = 520
-  isKeyFrame = true
-  presentationTimeUs = 157482
-sample:
-  trackIndex = 0
-  dataHashCode = 1994555264
-  size = 230
-  isKeyFrame = true
-  presentationTimeUs = 163378
-sample:
-  trackIndex = 0
-  dataHashCode = -748724753
-  size = 380
-  isKeyFrame = true
-  presentationTimeUs = 165986
-sample:
-  trackIndex = 0
-  dataHashCode = -1557661843
-  size = 692
-  isKeyFrame = true
-  presentationTimeUs = 170294
-sample:
-  trackIndex = 0
-  dataHashCode = 461522726
-  size = 270
-  isKeyFrame = true
-  presentationTimeUs = 178140
-sample:
-  trackIndex = 0
-  dataHashCode = 1058760091
-  size = 238
-  isKeyFrame = true
-  presentationTimeUs = 181201
-sample:
-  trackIndex = 0
-  dataHashCode = 1541647596
-  size = 722
-  isKeyFrame = true
-  presentationTimeUs = 183900
-sample:
-  trackIndex = 0
-  dataHashCode = -2107816707
-  size = 2062
-  isKeyFrame = true
-  presentationTimeUs = 192086
-released = true

libraries/test_data/src/test/assets/transformerdumps/wav/sample.wav.concatenated.dump

@@ -0,0 +1,127 @@
+format 0:
+  averageBitrate = 131072
+  sampleMimeType = audio/mp4a-latm
+  channelCount = 1
+  sampleRate = 44100
+  pcmEncoding = 2
+sample:
+  trackIndex = 0
+  dataHashCode = -85819864
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 0
+sample:
+  trackIndex = 0
+  dataHashCode = 566487491
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 100000
+sample:
+  trackIndex = 0
+  dataHashCode = -1256531710
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 200000
+sample:
+  trackIndex = 0
+  dataHashCode = 793455796
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 300000
+sample:
+  trackIndex = 0
+  dataHashCode = -268235582
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 400000
+sample:
+  trackIndex = 0
+  dataHashCode = -8136122
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 500000
+sample:
+  trackIndex = 0
+  dataHashCode = 1750866613
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 600000
+sample:
+  trackIndex = 0
+  dataHashCode = -1100753636
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 700000
+sample:
+  trackIndex = 0
+  dataHashCode = 507833230
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 800000
+sample:
+  trackIndex = 0
+  dataHashCode = 1472467506
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 900000
+sample:
+  trackIndex = 0
+  dataHashCode = -85819864
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1000000
+sample:
+  trackIndex = 0
+  dataHashCode = 566487491
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1100000
+sample:
+  trackIndex = 0
+  dataHashCode = -1256531710
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1200000
+sample:
+  trackIndex = 0
+  dataHashCode = 793455796
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1300000
+sample:
+  trackIndex = 0
+  dataHashCode = -268235582
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1400000
+sample:
+  trackIndex = 0
+  dataHashCode = -8136122
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1500000
+sample:
+  trackIndex = 0
+  dataHashCode = 1750866613
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1600000
+sample:
+  trackIndex = 0
+  dataHashCode = -1100753636
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1700000
+sample:
+  trackIndex = 0
+  dataHashCode = 507833230
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1800000
+sample:
+  trackIndex = 0
+  dataHashCode = 1472467506
+  size = 8820
+  isKeyFrame = true
+  presentationTimeUs = 1900000
+released = true

libraries/test_data/src/test/assets/transformerdumps/wav/sample.wav.concatenated_high_pitch.dump

@@ -0,0 +1,133 @@
+format 0:
+  averageBitrate = 131072
+  sampleMimeType = audio/mp4a-latm
+  channelCount = 1
+  sampleRate = 44100
+  pcmEncoding = 2
+sample:
+  trackIndex = 0
+  dataHashCode = 2042155098
+  size = 6482
+  isKeyFrame = true
+  presentationTimeUs = 0
+sample:
+  trackIndex = 0
+  dataHashCode = 1417355469
+  size = 8556
+  isKeyFrame = true
+  presentationTimeUs = 73492
+sample:
+  trackIndex = 0
+  dataHashCode = -2107697498
+  size = 8754
+  isKeyFrame = true
+  presentationTimeUs = 170498
+sample:
+  trackIndex = 0
+  dataHashCode = 736072795
+  size = 8908
+  isKeyFrame = true
+  presentationTimeUs = 269750
+sample:
+  trackIndex = 0
+  dataHashCode = -1913553170
+  size = 9208
+  isKeyFrame = true
+  presentationTimeUs = 370748
+sample:
+  trackIndex = 0
+  dataHashCode = 13583718
+  size = 8968
+  isKeyFrame = true
+  presentationTimeUs = 475147
+sample:
+  trackIndex = 0
+  dataHashCode = -1444602526
+  size = 8588
+  isKeyFrame = true
+  presentationTimeUs = 576825
+sample:
+  trackIndex = 0
+  dataHashCode = -1693065958
+  size = 8778
+  isKeyFrame = true
+  presentationTimeUs = 674195
+sample:
+  trackIndex = 0
+  dataHashCode = 2071205641
+  size = 8736
+  isKeyFrame = true
+  presentationTimeUs = 773718
+sample:
+  trackIndex = 0
+  dataHashCode = 1433538831
+  size = 8636
+  isKeyFrame = true
+  presentationTimeUs = 872766
+sample:
+  trackIndex = 0
+  dataHashCode = -949798077
+  size = 9424
+  isKeyFrame = true
+  presentationTimeUs = 970680
+sample:
+  trackIndex = 0
+  dataHashCode = -1275686831
+  size = 8088
+  isKeyFrame = true
+  presentationTimeUs = 1077528
+sample:
+  trackIndex = 0
+  dataHashCode = -1360039206
+  size = 8858
+  isKeyFrame = true
+  presentationTimeUs = 1169229
+sample:
+  trackIndex = 0
+  dataHashCode = 736072795
+  size = 8908
+  isKeyFrame = true
+  presentationTimeUs = 1269659
+sample:
+  trackIndex = 0
+  dataHashCode = -1913553170
+  size = 9208
+  isKeyFrame = true
+  presentationTimeUs = 1370657
+sample:
+  trackIndex = 0
+  dataHashCode = 13583718
+  size = 8968
+  isKeyFrame = true
+  presentationTimeUs = 1475056
+sample:
+  trackIndex = 0
+  dataHashCode = -1444602526
+  size = 8588
+  isKeyFrame = true
+  presentationTimeUs = 1576734
+sample:
+  trackIndex = 0
+  dataHashCode = -1693065958
+  size = 8778
+  isKeyFrame = true
+  presentationTimeUs = 1674104
+sample:
+  trackIndex = 0
+  dataHashCode = 2071205641
+  size = 8736
+  isKeyFrame = true
+  presentationTimeUs = 1773628
+sample:
+  trackIndex = 0
+  dataHashCode = 1433538831
+  size = 8636
+  isKeyFrame = true
+  presentationTimeUs = 1872675
+sample:
+  trackIndex = 0
+  dataHashCode = 992130724
+  size = 2580
+  isKeyFrame = true
+  presentationTimeUs = 1970589
+released = true

libraries/transformer/src/test/java/androidx/media3/transformer/SequenceExportTest.java

@@ -18,9 +18,11 @@ package androidx.media3.transformer;
 
 import static androidx.media3.common.util.Assertions.checkNotNull;
 import static androidx.media3.transformer.TestUtil.ASSET_URI_PREFIX;
+import static androidx.media3.transformer.TestUtil.FILE_AUDIO_RAW;
 import static androidx.media3.transformer.TestUtil.FILE_AUDIO_VIDEO;
 import static androidx.media3.transformer.TestUtil.FILE_AUDIO_VIDEO_INCREASING_TIMESTAMPS_15S;
 import static androidx.media3.transformer.TestUtil.createEncodersAndDecoders;
+import static androidx.media3.transformer.TestUtil.createPitchChangingAudioProcessor;
 import static androidx.media3.transformer.TestUtil.createTransformerBuilder;
 import static androidx.media3.transformer.TestUtil.getDumpFileName;
 import static androidx.media3.transformer.TestUtil.removeEncodersAndDecoders;
@@ -40,12 +42,20 @@ import java.nio.file.Files;
 import java.nio.file.Paths;
 import org.junit.After;
 import org.junit.Before;
+import org.junit.Ignore;
 import org.junit.Test;
 import org.junit.runner.RunWith;
 
 /**
  * End-to-end test for exporting a single {@link EditedMediaItemSequence} containing multiple {@link
  * EditedMediaItem} instances with {@link Transformer}.
+ *
+ * <p>Video tracks can not be processed by Robolectric, as the muxer audio/video interleaving means
+ * it waits for more audio samples before writing video samples. Robolectric decoders (currently)
+ * just copy input buffers to the output. Audio timestamps are computed based on the amount of data
+ * passed through (see [internal: b/178685617]), so are much smaller than expected because they are
+ * based on encoded samples. As a result, input files with video and audio must either remove or
+ * transmux the video.
  */
 @RunWith(AndroidJUnit4.class)
 public final class SequenceExportTest {
@@ -97,8 +107,7 @@ public final class SequenceExportTest {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
     MediaItem mediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_VIDEO);
-    SonicAudioProcessor sonicAudioProcessor = new SonicAudioProcessor();
+    SonicAudioProcessor sonicAudioProcessor = createPitchChangingAudioProcessor(/* pitch= */ 2f);
-    sonicAudioProcessor.setPitch(2f);
     Effect videoEffect = RgbFilter.createGrayscaleFilter();
     Effects effects =
         new Effects(ImmutableList.of(sonicAudioProcessor), ImmutableList.of(videoEffect));
@@ -172,11 +181,11 @@ public final class SequenceExportTest {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
     MediaItem mediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_VIDEO);
-    EditedMediaItem noAudioEditedMediaItem =
+    EditedMediaItem videoOnlyMediaItem =
         new EditedMediaItem.Builder(mediaItem).setRemoveAudio(true).build();
-    EditedMediaItem audioEditedMediaItem = new EditedMediaItem.Builder(mediaItem).build();
+    EditedMediaItem audioVideoMediaItem = new EditedMediaItem.Builder(mediaItem).build();
     EditedMediaItemSequence sequence =
-        new EditedMediaItemSequence(ImmutableList.of(noAudioEditedMediaItem, audioEditedMediaItem));
+        new EditedMediaItemSequence(ImmutableList.of(videoOnlyMediaItem, audioVideoMediaItem));
     Composition composition =
         new Composition.Builder(ImmutableList.of(sequence))
             .experimentalSetForceAudioTrack(true)
@@ -198,8 +207,7 @@ public final class SequenceExportTest {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
     MediaItem mediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_VIDEO);
-    SonicAudioProcessor sonicAudioProcessor = new SonicAudioProcessor();
+    SonicAudioProcessor sonicAudioProcessor = createPitchChangingAudioProcessor(/* pitch= */ 2f);
-    sonicAudioProcessor.setPitch(2f);
     Effects effects =
         new Effects(ImmutableList.of(sonicAudioProcessor), /* videoEffects= */ ImmutableList.of());
     EditedMediaItem noAudioEditedMediaItem =
@@ -224,22 +232,11 @@ public final class SequenceExportTest {
   }
 
   @Test
-  public void start_concatenateSameAudioItemWithEffects_completesSuccessfully() throws Exception {
+  public void start_concatenateSameAudioItem_completesSuccessfully() throws Exception {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
-    MediaItem mediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_VIDEO);
+    MediaItem audioOnlyMediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
-    SonicAudioProcessor sonicAudioProcessor = new SonicAudioProcessor();
+    EditedMediaItem editedMediaItem = new EditedMediaItem.Builder(audioOnlyMediaItem).build();
-    sonicAudioProcessor.setPitch(2f);
-    Effects effects =
-        new Effects(ImmutableList.of(sonicAudioProcessor), /* videoEffects= */ ImmutableList.of());
-
-    // The video track must be removed in order for the export to end. Indeed, the
-    // Robolectric decoder just copies the input buffers to the output and the audio timestamps are
-    // therefore computed based on the encoded samples (see [internal: b/178685617]). As a result,
-    // the audio timestamps are much smaller than they should be and the muxer waits for more audio
-    // samples before writing video samples.
-    EditedMediaItem editedMediaItem =
-        new EditedMediaItem.Builder(mediaItem).setEffects(effects).setRemoveVideo(true).build();
     EditedMediaItemSequence editedMediaItemSequence =
         new EditedMediaItemSequence(ImmutableList.of(editedMediaItem, editedMediaItem));
     Composition composition =
@@ -251,6 +248,68 @@ public final class SequenceExportTest {
     DumpFileAsserts.assertOutput(
         context,
         checkNotNull(testMuxerHolder.testMuxer),
-        getDumpFileName(FILE_AUDIO_VIDEO + ".concatenated_audio_high_pitch"));
+        getDumpFileName(FILE_AUDIO_RAW + ".concatenated"));
+  }
+
+  @Test
+  public void start_concatenateSameAudioItemWithEffects_completesSuccessfully() throws Exception {
+    Transformer transformer =
+        createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
+    MediaItem audioOnlyMediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
+    SonicAudioProcessor sonicAudioProcessor = createPitchChangingAudioProcessor(/* pitch= */ 2f);
+    Effects effects =
+        new Effects(ImmutableList.of(sonicAudioProcessor), /* videoEffects= */ ImmutableList.of());
+    EditedMediaItem editedMediaItem =
+        new EditedMediaItem.Builder(audioOnlyMediaItem).setEffects(effects).build();
+    EditedMediaItemSequence editedMediaItemSequence =
+        new EditedMediaItemSequence(ImmutableList.of(editedMediaItem, editedMediaItem));
+    Composition composition =
+        new Composition.Builder(ImmutableList.of(editedMediaItemSequence)).build();
+
+    transformer.start(composition, outputPath);
+    TransformerTestRunner.runLooper(transformer);
+
+    DumpFileAsserts.assertOutput(
+        context,
+        checkNotNull(testMuxerHolder.testMuxer),
+        getDumpFileName(FILE_AUDIO_RAW + ".concatenated_high_pitch"));
+  }
+
+  @Test
+  @Ignore("Handle MediaItem effects changes (See [internal: b/274093424]).")
+  public void start_concatenateSameAudioItemWithDifferentEffects_completesSuccessfully()
+      throws Exception {
+    Transformer transformer =
+        createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
+    MediaItem audioOnlyMediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
+    Effects highPitchEffects =
+        new Effects(
+            ImmutableList.of(createPitchChangingAudioProcessor(/* pitch= */ 2f)),
+            /* videoEffects= */ ImmutableList.of());
+    EditedMediaItem highPitchMediaItem =
+        new EditedMediaItem.Builder(audioOnlyMediaItem)
+            .setRemoveVideo(true)
+            .setEffects(highPitchEffects)
+            .build();
+    Effects lowPitchEffects =
+        new Effects(
+            ImmutableList.of(createPitchChangingAudioProcessor(/* pitch= */ 0.5f)),
+            /* videoEffects= */ ImmutableList.of());
+    EditedMediaItem lowPitchMediaItem =
+        new EditedMediaItem.Builder(audioOnlyMediaItem)
+            .setRemoveVideo(true)
+            .setEffects(lowPitchEffects)
+            .build();
+    EditedMediaItemSequence sequence =
+        new EditedMediaItemSequence(ImmutableList.of(highPitchMediaItem, lowPitchMediaItem));
+    Composition composition = new Composition.Builder(ImmutableList.of(sequence)).build();
+
+    transformer.start(composition, outputPath);
+    TransformerTestRunner.runLooper(transformer);
+
+    DumpFileAsserts.assertOutput(
+        context,
+        checkNotNull(testMuxerHolder.testMuxer),
+        getDumpFileName(FILE_AUDIO_RAW + ".high_pitch_then_low_pitch"));
   }
 }

libraries/transformer/src/test/java/androidx/media3/transformer/TestUtil.java

@@ -24,6 +24,7 @@ import androidx.annotation.Nullable;
 import androidx.media3.common.C;
 import androidx.media3.common.Format;
 import androidx.media3.common.MimeTypes;
+import androidx.media3.common.audio.SonicAudioProcessor;
 import androidx.media3.common.util.UnstableApi;
 import androidx.media3.common.util.Util;
 import androidx.media3.test.utils.FakeClock;
@@ -239,6 +240,12 @@ public final class TestUtil {
             new DefaultEncoderFactory.Builder(context).setEnableFallback(enableFallback).build());
   }
 
+  public static SonicAudioProcessor createPitchChangingAudioProcessor(float pitch) {
+    SonicAudioProcessor sonicAudioProcessor = new SonicAudioProcessor();
+    sonicAudioProcessor.setPitch(pitch);
+    return sonicAudioProcessor;
+  }
+
   public static String getDumpFileName(String originalFileName) {
     return DUMP_FILE_OUTPUT_DIRECTORY + '/' + originalFileName + '.' + DUMP_FILE_EXTENSION;
   }