Handle media item (Effects/Format) changes in AudioSamplePipeline.

On a MediaItem change, the input Format (and Effects to apply) may be
different. Therefore the AudioProcessingPipeline must be reconfigured
to determine what processing is active, and what the AudioFormat of the
data output is. In the event that it is different, additional
AudioProcessor instances must be used to ensure the encoder will still
be able to accept the audio buffers.

PiperOrigin-RevId: 544338451

libraries/common/src/main/java/androidx/media3/common/audio/AudioProcessingPipeline.java

@@ -170,7 +170,12 @@ public final class AudioProcessingPipeline {
     }
   }
 
-  /** Returns the {@link AudioFormat} currently being output. */
+  /**
+   * Returns the {@link AudioFormat} of data being output through {@link #getOutput()}.
+   *
+   * @return The {@link AudioFormat} currently being output, or {@link AudioFormat#NOT_SET} if no
+   *     {@linkplain #configure(AudioFormat) configuration} has been {@linkplain #flush() applied}.
+   */
   public AudioFormat getOutputAudioFormat() {
     return outputAudioFormat;
   }

libraries/test_data/src/test/assets/transformerdumps/mp4/testvid_1022ms.mp4.silence_then_audio_with_effects.dump

@@ -637,110 +637,38 @@ sample:
   presentationTimeUs = 967709
 sample:
   trackIndex = 1
-  dataHashCode = -1759454975
+  dataHashCode = 1916158593
-  size = 440
+  size = 5164
   isKeyFrame = true
   presentationTimeUs = 992653
 sample:
   trackIndex = 1
-  dataHashCode = -1759454975
+  dataHashCode = -1950877403
-  size = 440
+  size = 1120
   isKeyFrame = true
-  presentationTimeUs = 995147
+  presentationTimeUs = 1021927
 sample:
   trackIndex = 1
-  dataHashCode = -1759454975
+  dataHashCode = 1590174306
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 997641
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1000136
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1002630
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1005124
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1007619
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1010113
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1012607
-sample:
-  trackIndex = 1
-  dataHashCode = -1759454975
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1015102
-sample:
-  trackIndex = 1
-  dataHashCode = -2121773972
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1017596
-sample:
-  trackIndex = 1
-  dataHashCode = 1517597454
-  size = 440
-  isKeyFrame = true
-  presentationTimeUs = 1020090
-sample:
-  trackIndex = 1
-  dataHashCode = 1952035733
-  size = 448
-  isKeyFrame = true
-  presentationTimeUs = 1022585
-sample:
-  trackIndex = 1
-  dataHashCode = -1861807761
-  size = 604
-  isKeyFrame = true
-  presentationTimeUs = 1025124
-sample:
-  trackIndex = 1
-  dataHashCode = -1460926592
   size = 880
   isKeyFrame = true
-  presentationTimeUs = 1028548
+  presentationTimeUs = 1028276
 sample:
   trackIndex = 1
-  dataHashCode = 1194085269
+  dataHashCode = -432400310
-  size = 532
+  size = 1588
   isKeyFrame = true
-  presentationTimeUs = 1033537
+  presentationTimeUs = 1033265
 sample:
   trackIndex = 1
-  dataHashCode = -493118955
+  dataHashCode = -270222999
-  size = 2296
+  size = 832
   isKeyFrame = true
-  presentationTimeUs = 1036553
+  presentationTimeUs = 1042267
 sample:
   trackIndex = 1
-  dataHashCode = 90420756
+  dataHashCode = -1941634508
-  size = 4568
+  size = 5028
   isKeyFrame = true
-  presentationTimeUs = 1049569
+  presentationTimeUs = 1046984
 released = true

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

@@ -66,68 +66,74 @@ sample:
   presentationTimeUs = 872766
 sample:
   trackIndex = 0
-  dataHashCode = -949798077
+  dataHashCode = 992130724
-  size = 9424
+  size = 2580
   isKeyFrame = true
   presentationTimeUs = 970680
 sample:
   trackIndex = 0
-  dataHashCode = -1275686831
+  dataHashCode = 2042155098
-  size = 8088
+  size = 6482
   isKeyFrame = true
-  presentationTimeUs = 1077528
+  presentationTimeUs = 999931
 sample:
   trackIndex = 0
-  dataHashCode = -1360039206
+  dataHashCode = 1417355469
-  size = 8858
+  size = 8556
   isKeyFrame = true
-  presentationTimeUs = 1169229
+  presentationTimeUs = 1073424
+sample:
+  trackIndex = 0
+  dataHashCode = -2107697498
+  size = 8754
+  isKeyFrame = true
+  presentationTimeUs = 1170430
 sample:
   trackIndex = 0
   dataHashCode = 736072795
   size = 8908
   isKeyFrame = true
-  presentationTimeUs = 1269659
+  presentationTimeUs = 1269682
 sample:
   trackIndex = 0
   dataHashCode = -1913553170
   size = 9208
   isKeyFrame = true
-  presentationTimeUs = 1370657
+  presentationTimeUs = 1370680
 sample:
   trackIndex = 0
   dataHashCode = 13583718
   size = 8968
   isKeyFrame = true
-  presentationTimeUs = 1475056
+  presentationTimeUs = 1475079
 sample:
   trackIndex = 0
   dataHashCode = -1444602526
   size = 8588
   isKeyFrame = true
-  presentationTimeUs = 1576734
+  presentationTimeUs = 1576757
 sample:
   trackIndex = 0
   dataHashCode = -1693065958
   size = 8778
   isKeyFrame = true
-  presentationTimeUs = 1674104
+  presentationTimeUs = 1674126
 sample:
   trackIndex = 0
   dataHashCode = 2071205641
   size = 8736
   isKeyFrame = true
-  presentationTimeUs = 1773628
+  presentationTimeUs = 1773650
 sample:
   trackIndex = 0
   dataHashCode = 1433538831
   size = 8636
   isKeyFrame = true
-  presentationTimeUs = 1872675
+  presentationTimeUs = 1872698
 sample:
   trackIndex = 0
   dataHashCode = 992130724
   size = 2580
   isKeyFrame = true
-  presentationTimeUs = 1970589
+  presentationTimeUs = 1970612
 released = true

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

@@ -0,0 +1,139 @@
+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 = 992130724
+  size = 2580
+  isKeyFrame = true
+  presentationTimeUs = 970680
+sample:
+  trackIndex = 0
+  dataHashCode = 2002762528
+  size = 6450
+  isKeyFrame = true
+  presentationTimeUs = 999931
+sample:
+  trackIndex = 0
+  dataHashCode = -657557348
+  size = 9288
+  isKeyFrame = true
+  presentationTimeUs = 1073061
+sample:
+  trackIndex = 0
+  dataHashCode = 1101737790
+  size = 8026
+  isKeyFrame = true
+  presentationTimeUs = 1178367
+sample:
+  trackIndex = 0
+  dataHashCode = 942749699
+  size = 9580
+  isKeyFrame = true
+  presentationTimeUs = 1269365
+sample:
+  trackIndex = 0
+  dataHashCode = 1696779430
+  size = 9176
+  isKeyFrame = true
+  presentationTimeUs = 1377981
+sample:
+  trackIndex = 0
+  dataHashCode = -1146784158
+  size = 7898
+  isKeyFrame = true
+  presentationTimeUs = 1482018
+sample:
+  trackIndex = 0
+  dataHashCode = -108739018
+  size = 9244
+  isKeyFrame = true
+  presentationTimeUs = 1571564
+sample:
+  trackIndex = 0
+  dataHashCode = 1527742727
+  size = 8614
+  isKeyFrame = true
+  presentationTimeUs = 1676371
+sample:
+  trackIndex = 0
+  dataHashCode = -771072152
+  size = 8408
+  isKeyFrame = true
+  presentationTimeUs = 1774036
+sample:
+  trackIndex = 0
+  dataHashCode = -307907100
+  size = 10212
+  isKeyFrame = true
+  presentationTimeUs = 1869365
+sample:
+  trackIndex = 0
+  dataHashCode = -2024887086
+  size = 1312
+  isKeyFrame = true
+  presentationTimeUs = 1985147
+released = true

libraries/test_data/src/test/assets/transformerdumps/wav/sample_rf64.wav-high_pitch_then_sample.wav-high_pitch.dump

@@ -0,0 +1,109 @@
+format 0:
+  averageBitrate = 131072
+  sampleMimeType = audio/mp4a-latm
+  channelCount = 2
+  sampleRate = 48000
+  pcmEncoding = 2
+sample:
+  trackIndex = 0
+  dataHashCode = -226169394
+  size = 15528
+  isKeyFrame = true
+  presentationTimeUs = 0
+sample:
+  trackIndex = 0
+  dataHashCode = -1654106973
+  size = 17436
+  isKeyFrame = true
+  presentationTimeUs = 80875
+sample:
+  trackIndex = 0
+  dataHashCode = 1078555669
+  size = 19384
+  isKeyFrame = true
+  presentationTimeUs = 171687
+sample:
+  trackIndex = 0
+  dataHashCode = -1783495042
+  size = 8904
+  isKeyFrame = true
+  presentationTimeUs = 272645
+sample:
+  trackIndex = 0
+  dataHashCode = 1369234850
+  size = 5688
+  isKeyFrame = true
+  presentationTimeUs = 319020
+sample:
+  trackIndex = 0
+  dataHashCode = 340823895
+  size = 14108
+  isKeyFrame = true
+  presentationTimeUs = 348645
+sample:
+  trackIndex = 0
+  dataHashCode = 591641755
+  size = 18624
+  isKeyFrame = true
+  presentationTimeUs = 422125
+sample:
+  trackIndex = 0
+  dataHashCode = -909124577
+  size = 19056
+  isKeyFrame = true
+  presentationTimeUs = 519125
+sample:
+  trackIndex = 0
+  dataHashCode = -1644849415
+  size = 19392
+  isKeyFrame = true
+  presentationTimeUs = 618375
+sample:
+  trackIndex = 0
+  dataHashCode = 660809049
+  size = 20044
+  isKeyFrame = true
+  presentationTimeUs = 719375
+sample:
+  trackIndex = 0
+  dataHashCode = -601805369
+  size = 19524
+  isKeyFrame = true
+  presentationTimeUs = 823770
+sample:
+  trackIndex = 0
+  dataHashCode = -1348998373
+  size = 18696
+  isKeyFrame = true
+  presentationTimeUs = 925458
+sample:
+  trackIndex = 0
+  dataHashCode = 1191266277
+  size = 19108
+  isKeyFrame = true
+  presentationTimeUs = 1022833
+sample:
+  trackIndex = 0
+  dataHashCode = -1278785599
+  size = 19016
+  isKeyFrame = true
+  presentationTimeUs = 1122354
+sample:
+  trackIndex = 0
+  dataHashCode = -809901361
+  size = 18800
+  isKeyFrame = true
+  presentationTimeUs = 1221395
+sample:
+  trackIndex = 0
+  dataHashCode = 163546833
+  size = 5616
+  isKeyFrame = true
+  presentationTimeUs = 1319312
+sample:
+  trackIndex = 0
+  dataHashCode = 743627
+  size = 4
+  isKeyFrame = true
+  presentationTimeUs = 1348562
+released = true

libraries/test_data/src/test/assets/transformerdumps/wav/sample_rf64.wav-high_pitch_then_sample.wav-low_pitch.dump

@@ -0,0 +1,109 @@
+format 0:
+  averageBitrate = 131072
+  sampleMimeType = audio/mp4a-latm
+  channelCount = 2
+  sampleRate = 48000
+  pcmEncoding = 2
+sample:
+  trackIndex = 0
+  dataHashCode = -226169394
+  size = 15528
+  isKeyFrame = true
+  presentationTimeUs = 0
+sample:
+  trackIndex = 0
+  dataHashCode = -1654106973
+  size = 17436
+  isKeyFrame = true
+  presentationTimeUs = 80875
+sample:
+  trackIndex = 0
+  dataHashCode = 1078555669
+  size = 19384
+  isKeyFrame = true
+  presentationTimeUs = 171687
+sample:
+  trackIndex = 0
+  dataHashCode = -1783495042
+  size = 8904
+  isKeyFrame = true
+  presentationTimeUs = 272645
+sample:
+  trackIndex = 0
+  dataHashCode = 1369234850
+  size = 5688
+  isKeyFrame = true
+  presentationTimeUs = 319020
+sample:
+  trackIndex = 0
+  dataHashCode = 711439073
+  size = 14040
+  isKeyFrame = true
+  presentationTimeUs = 348645
+sample:
+  trackIndex = 0
+  dataHashCode = -555451391
+  size = 20216
+  isKeyFrame = true
+  presentationTimeUs = 421770
+sample:
+  trackIndex = 0
+  dataHashCode = 1700905759
+  size = 17472
+  isKeyFrame = true
+  presentationTimeUs = 527062
+sample:
+  trackIndex = 0
+  dataHashCode = -416119019
+  size = 20856
+  isKeyFrame = true
+  presentationTimeUs = 618062
+sample:
+  trackIndex = 0
+  dataHashCode = -2060733789
+  size = 19976
+  isKeyFrame = true
+  presentationTimeUs = 726687
+sample:
+  trackIndex = 0
+  dataHashCode = 212454167
+  size = 17192
+  isKeyFrame = true
+  presentationTimeUs = 830729
+sample:
+  trackIndex = 0
+  dataHashCode = 956501205
+  size = 20124
+  isKeyFrame = true
+  presentationTimeUs = 920270
+sample:
+  trackIndex = 0
+  dataHashCode = 299859135
+  size = 18748
+  isKeyFrame = true
+  presentationTimeUs = 1025083
+sample:
+  trackIndex = 0
+  dataHashCode = 636945085
+  size = 18304
+  isKeyFrame = true
+  presentationTimeUs = 1122729
+sample:
+  trackIndex = 0
+  dataHashCode = -754285165
+  size = 22232
+  isKeyFrame = true
+  presentationTimeUs = 1218062
+sample:
+  trackIndex = 0
+  dataHashCode = -1967816889
+  size = 2856
+  isKeyFrame = true
+  presentationTimeUs = 1333854
+sample:
+  trackIndex = 0
+  dataHashCode = 1012987
+  size = 4
+  isKeyFrame = true
+  presentationTimeUs = 1348729
+released = true

libraries/test_data/src/test/assets/transformerdumps/wav/sample_rf64.wav_then_sample.wav.dump

@@ -0,0 +1,97 @@
+format 0:
+  averageBitrate = 131072
+  sampleMimeType = audio/mp4a-latm
+  channelCount = 2
+  sampleRate = 48000
+  pcmEncoding = 2
+sample:
+  trackIndex = 0
+  dataHashCode = -278103001
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 0
+sample:
+  trackIndex = 0
+  dataHashCode = 1522105084
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 100000
+sample:
+  trackIndex = 0
+  dataHashCode = 932319027
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 200000
+sample:
+  trackIndex = 0
+  dataHashCode = 325000240
+  size = 9336
+  isKeyFrame = true
+  presentationTimeUs = 300000
+sample:
+  trackIndex = 0
+  dataHashCode = -717358295
+  size = 19196
+  isKeyFrame = true
+  presentationTimeUs = 348625
+sample:
+  trackIndex = 0
+  dataHashCode = -877978019
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 448604
+sample:
+  trackIndex = 0
+  dataHashCode = 903565693
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 548604
+sample:
+  trackIndex = 0
+  dataHashCode = -1424006305
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 648604
+sample:
+  trackIndex = 0
+  dataHashCode = -1450884853
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 748604
+sample:
+  trackIndex = 0
+  dataHashCode = 1566382623
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 848604
+sample:
+  trackIndex = 0
+  dataHashCode = -510424155
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 948604
+sample:
+  trackIndex = 0
+  dataHashCode = 13706341
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 1048604
+sample:
+  trackIndex = 0
+  dataHashCode = -1421600077
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 1148604
+sample:
+  trackIndex = 0
+  dataHashCode = 722747579
+  size = 19200
+  isKeyFrame = true
+  presentationTimeUs = 1248604
+sample:
+  trackIndex = 0
+  dataHashCode = 326119
+  size = 4
+  isKeyFrame = true
+  presentationTimeUs = 1348604
+released = true

libraries/transformer/src/main/java/androidx/media3/transformer/AudioSamplePipeline.java

@@ -19,10 +19,12 @@ package androidx.media3.transformer;
 import static androidx.media3.common.util.Assertions.checkArgument;
 import static androidx.media3.common.util.Assertions.checkNotNull;
 import static androidx.media3.common.util.Assertions.checkState;
+import static androidx.media3.common.util.Assertions.checkStateNotNull;
 import static androidx.media3.decoder.DecoderInputBuffer.BUFFER_REPLACEMENT_MODE_DIRECT;
 import static androidx.media3.decoder.DecoderInputBuffer.BUFFER_REPLACEMENT_MODE_DISABLED;
 import static java.lang.Math.min;
 
+import android.util.Pair;
 import androidx.annotation.Nullable;
 import androidx.media3.common.C;
 import androidx.media3.common.Format;
@@ -30,7 +32,11 @@ import androidx.media3.common.MimeTypes;
 import androidx.media3.common.audio.AudioProcessingPipeline;
 import androidx.media3.common.audio.AudioProcessor;
 import androidx.media3.common.audio.AudioProcessor.AudioFormat;
+import androidx.media3.common.audio.ChannelMixingAudioProcessor;
+import androidx.media3.common.audio.ChannelMixingMatrix;
+import androidx.media3.common.audio.SonicAudioProcessor;
 import androidx.media3.common.audio.SpeedChangingAudioProcessor;
+import androidx.media3.common.util.NullableType;
 import androidx.media3.common.util.Util;
 import androidx.media3.decoder.DecoderInputBuffer;
 import com.google.common.collect.ImmutableList;
@@ -38,6 +44,7 @@ import java.nio.ByteBuffer;
 import java.nio.ByteOrder;
 import java.util.Queue;
 import java.util.concurrent.ConcurrentLinkedDeque;
+import java.util.concurrent.atomic.AtomicReference;
 import org.checkerframework.dataflow.qual.Pure;
 
 /** Pipeline to process, re-encode and mux raw audio samples. */
@@ -49,12 +56,14 @@ import org.checkerframework.dataflow.qual.Pure;
   private final SilentAudioGenerator silentAudioGenerator;
   private final Queue<DecoderInputBuffer> availableInputBuffers;
   private final Queue<DecoderInputBuffer> pendingInputBuffers;
-  private final AudioProcessingPipeline audioProcessingPipeline;
   private final Codec encoder;
   private final AudioFormat encoderInputAudioFormat;
   private final DecoderInputBuffer encoderInputBuffer;
   private final DecoderInputBuffer encoderOutputBuffer;
-
+  private final AtomicReference<@NullableType Pair<EditedMediaItem, @NullableType Format>>
+      pendingMediaItem;
+  private boolean receivedFirstMediaItemCallback;
+  private AudioProcessingPipeline audioProcessingPipeline;
   private long encoderTotalInputBytes;
 
   private volatile boolean queueEndOfStreamAfterSilence;
@@ -64,13 +73,13 @@ import org.checkerframework.dataflow.qual.Pure;
       Format firstAssetLoaderInputFormat,
       Format firstPipelineInputFormat,
       TransformationRequest transformationRequest,
-      boolean flattenForSlowMotion,
+      EditedMediaItem firstEditedMediaItem,
-      ImmutableList<AudioProcessor> audioProcessors,
       Codec.EncoderFactory encoderFactory,
       MuxerWrapper muxerWrapper,
       FallbackListener fallbackListener)
       throws ExportException {
     super(firstAssetLoaderInputFormat, muxerWrapper);
+    checkArgument(firstPipelineInputFormat.pcmEncoding != Format.NO_VALUE);
 
     availableInputBuffers = new ConcurrentLinkedDeque<>();
     ByteBuffer emptyBuffer = ByteBuffer.allocateDirect(0).order(ByteOrder.nativeOrder());
@@ -80,36 +89,21 @@ import org.checkerframework.dataflow.qual.Pure;
       availableInputBuffers.add(inputBuffer);
     }
     pendingInputBuffers = new ConcurrentLinkedDeque<>();
-
     encoderInputBuffer = new DecoderInputBuffer(BUFFER_REPLACEMENT_MODE_DISABLED);
     encoderOutputBuffer = new DecoderInputBuffer(BUFFER_REPLACEMENT_MODE_DISABLED);
-
+    pendingMediaItem = new AtomicReference<>();
-    checkArgument(firstPipelineInputFormat.pcmEncoding != Format.NO_VALUE);
     AudioFormat inputAudioFormat = new AudioFormat(firstPipelineInputFormat);
-
     silentAudioGenerator = new SilentAudioGenerator(inputAudioFormat);
+    audioProcessingPipeline =
+        configureProcessing(
+            /* editedMediaItem= */ firstEditedMediaItem,
+            /* trackFormat= */ firstPipelineInputFormat,
+            /* inputAudioFormat= */ inputAudioFormat,
+            /* requiredOutputAudioFormat= */ AudioFormat.NOT_SET);
+    AudioFormat outputAudioFormat = audioProcessingPipeline.getOutputAudioFormat();
+    checkState(!outputAudioFormat.equals(AudioFormat.NOT_SET));
 
-    if (flattenForSlowMotion && firstAssetLoaderInputFormat.metadata != null) {
+    encoderInputAudioFormat = outputAudioFormat;
-      audioProcessors =
-          new ImmutableList.Builder<AudioProcessor>()
-              .add(
-                  new SpeedChangingAudioProcessor(
-                      new SegmentSpeedProvider(firstAssetLoaderInputFormat.metadata)))
-              .addAll(audioProcessors)
-              .build();
-    }
-
-    audioProcessingPipeline = new AudioProcessingPipeline(audioProcessors);
-
-    try {
-      encoderInputAudioFormat = audioProcessingPipeline.configure(inputAudioFormat);
-    } catch (AudioProcessor.UnhandledAudioFormatException unhandledAudioFormatException) {
-      throw ExportException.createForAudioProcessing(
-          unhandledAudioFormatException, inputAudioFormat);
-    }
-
-    audioProcessingPipeline.flush();
-
     Format requestedEncoderFormat =
         new Format.Builder()
             .setSampleMimeType(
@@ -153,16 +147,22 @@ import org.checkerframework.dataflow.qual.Pure;
       if (isLast) {
         queueEndOfStreamAfterSilence = true;
       }
+    } else {
+      checkState(MimeTypes.isAudio(trackFormat.sampleMimeType));
+      checkState(trackFormat.pcmEncoding != Format.NO_VALUE);
+    }
+
+    if (!receivedFirstMediaItemCallback) {
+      receivedFirstMediaItemCallback = true;
       return;
     }
-    checkState(MimeTypes.isAudio(trackFormat.sampleMimeType));
+    pendingMediaItem.set(Pair.create(editedMediaItem, trackFormat));
-    checkState(trackFormat.pcmEncoding != Format.NO_VALUE);
   }
 
   @Override
   @Nullable
   public DecoderInputBuffer getInputBuffer() {
-    if (shouldGenerateSilence()) {
+    if (shouldGenerateSilence() || pendingMediaItem.get() != null) {
       return null;
     }
     return availableInputBuffers.peek();
@@ -170,6 +170,7 @@ import org.checkerframework.dataflow.qual.Pure;
 
   @Override
   public boolean queueInputBuffer() {
+    checkState(pendingMediaItem.get() == null);
     DecoderInputBuffer inputBuffer = availableInputBuffers.remove();
     pendingInputBuffers.add(inputBuffer);
     return true;
@@ -218,10 +219,34 @@ import org.checkerframework.dataflow.qual.Pure;
     return encoder.isEnded();
   }
 
+  /**
+   * Reconfigures audio processing based on the pending {@linkplain #onMediaItemChanged media item
+   * change}.
+   *
+   * <p>Before reconfiguration, all pending buffers must be fully processed and drained to the
+   * encoder, however end of stream buffers should be handled so the encoder is not {@link
+   * #queueEndOfStreamToEncoder() queued end of stream}.
+   */
+  private void reconfigureProcessingForPendingMediaItem() throws ExportException {
+    Pair<EditedMediaItem, @NullableType Format> pendingChange =
+        checkStateNotNull(pendingMediaItem.get());
+    AudioFormat pendingAudioFormat =
+        pendingChange.second != null
+            ? new AudioFormat(pendingChange.second)
+            : silentAudioGenerator.audioFormat;
+    audioProcessingPipeline =
+        configureProcessing(
+            /* editedMediaItem= */ pendingChange.first,
+            /* trackFormat= */ pendingChange.second,
+            /* inputAudioFormat= */ pendingAudioFormat,
+            /* requiredOutputAudioFormat= */ encoderInputAudioFormat);
+    pendingMediaItem.set(null);
+  }
+
   /**
    * Attempts to pass input data to the encoder.
    *
-   * @return Whether it may be possible to feed more data immediately by calling this method again.
+   * @return Whether the {@link AudioSamplePipeline} may be able to continue processing data.
    */
   private boolean feedEncoderFromInput() throws ExportException {
     if (!encoder.maybeDequeueInputBuffer(encoderInputBuffer)) {
@@ -234,6 +259,10 @@ import org.checkerframework.dataflow.qual.Pure;
     }
 
     if (pendingInputBuffers.isEmpty()) {
+      if (pendingMediaItem.get() != null) {
+        reconfigureProcessingForPendingMediaItem();
+        return true;
+      }
       // Only read volatile variable queueEndOfStreamAfterSilence if there is a chance that end of
       // stream should be queued.
       if (!silentAudioGenerator.hasRemaining() && queueEndOfStreamAfterSilence) {
@@ -244,7 +273,9 @@ import org.checkerframework.dataflow.qual.Pure;
 
     DecoderInputBuffer pendingInputBuffer = pendingInputBuffers.element();
     if (pendingInputBuffer.isEndOfStream()) {
-      queueEndOfStreamToEncoder();
+      if (pendingMediaItem.get() == null) {
+        queueEndOfStreamToEncoder();
+      }
       removePendingInputBuffer();
       return false;
     }
@@ -260,7 +291,7 @@ import org.checkerframework.dataflow.qual.Pure;
   /**
    * Attempts to feed audio processor output data to the encoder.
    *
-   * @return Whether it may be possible to feed more data immediately by calling this method again.
+   * @return Whether the {@link AudioSamplePipeline} may be able to continue processing data.
    */
   private boolean feedEncoderFromProcessingPipeline() throws ExportException {
     if (!encoder.maybeDequeueInputBuffer(encoderInputBuffer)) {
@@ -268,9 +299,12 @@ import org.checkerframework.dataflow.qual.Pure;
     }
 
     ByteBuffer processingPipelineOutputBuffer = audioProcessingPipeline.getOutput();
-
     if (!processingPipelineOutputBuffer.hasRemaining()) {
       if (audioProcessingPipeline.isEnded()) {
+        if (pendingMediaItem.get() != null) {
+          reconfigureProcessingForPendingMediaItem();
+          return true;
+        }
         queueEndOfStreamToEncoder();
       }
       return false;
@@ -283,7 +317,7 @@ import org.checkerframework.dataflow.qual.Pure;
   /**
    * Attempts to feed input data to the {@link AudioProcessingPipeline}.
    *
-   * @return Whether it may be possible to feed more data immediately by calling this method again.
+   * @return Whether the {@link AudioSamplePipeline} may be able to continue processing data.
    */
   private boolean feedProcessingPipelineFromInput() {
     if (shouldGenerateSilence()) {
@@ -295,7 +329,8 @@ import org.checkerframework.dataflow.qual.Pure;
     if (pendingInputBuffers.isEmpty()) {
       // Only read volatile variable queueEndOfStreamAfterSilence if there is a chance that end of
       // stream should be queued.
-      if (!silentAudioGenerator.hasRemaining() && queueEndOfStreamAfterSilence) {
+      if (pendingMediaItem.get() != null
+          || (!silentAudioGenerator.hasRemaining() && queueEndOfStreamAfterSilence)) {
         audioProcessingPipeline.queueEndOfStream();
       }
       return false;
@@ -370,4 +405,57 @@ import org.checkerframework.dataflow.qual.Pure;
   private boolean shouldGenerateSilence() {
     return silentAudioGenerator.hasRemaining() && pendingInputBuffers.isEmpty();
   }
+
+  private static AudioProcessingPipeline configureProcessing(
+      EditedMediaItem editedMediaItem,
+      @Nullable Format trackFormat,
+      AudioFormat inputAudioFormat,
+      AudioFormat requiredOutputAudioFormat)
+      throws ExportException {
+    ImmutableList.Builder<AudioProcessor> audioProcessors = new ImmutableList.Builder<>();
+    if (editedMediaItem.flattenForSlowMotion
+        && trackFormat != null
+        && trackFormat.metadata != null) {
+      audioProcessors.add(
+          new SpeedChangingAudioProcessor(new SegmentSpeedProvider(trackFormat.metadata)));
+    }
+    audioProcessors.addAll(editedMediaItem.effects.audioProcessors);
+    // Ensure the output from APP matches what the encoder is configured to receive.
+    if (!requiredOutputAudioFormat.equals(AudioFormat.NOT_SET)) {
+      SonicAudioProcessor sampleRateChanger = new SonicAudioProcessor();
+      sampleRateChanger.setOutputSampleRateHz(requiredOutputAudioFormat.sampleRate);
+      audioProcessors.add(sampleRateChanger);
+
+      // TODO(b/262706549): Handle channel mixing with AudioMixer.
+      if (requiredOutputAudioFormat.channelCount <= 2) {
+        // ChannelMixingMatrix.create only has defaults for mono/stereo input/output.
+        ChannelMixingAudioProcessor channelCountChanger = new ChannelMixingAudioProcessor();
+        channelCountChanger.putChannelMixingMatrix(
+            ChannelMixingMatrix.create(
+                /* inputChannelCount= */ 1, requiredOutputAudioFormat.channelCount));
+        channelCountChanger.putChannelMixingMatrix(
+            ChannelMixingMatrix.create(
+                /* inputChannelCount= */ 2, requiredOutputAudioFormat.channelCount));
+        audioProcessors.add(channelCountChanger);
+      }
+    }
+
+    AudioProcessingPipeline audioProcessingPipeline =
+        new AudioProcessingPipeline(audioProcessors.build());
+    try {
+      AudioFormat outputAudioFormat = audioProcessingPipeline.configure(inputAudioFormat);
+      if (!requiredOutputAudioFormat.equals(AudioFormat.NOT_SET)
+          && !outputAudioFormat.equals(requiredOutputAudioFormat)) {
+        throw new AudioProcessor.UnhandledAudioFormatException(
+            "Audio format can not be modified to match existing downstream format",
+            inputAudioFormat);
+      }
+    } catch (AudioProcessor.UnhandledAudioFormatException unhandledAudioFormatException) {
+      throw ExportException.createForAudioProcessing(
+          unhandledAudioFormatException, inputAudioFormat);
+    }
+
+    audioProcessingPipeline.flush();
+    return audioProcessingPipeline;
+  }
 }

libraries/transformer/src/main/java/androidx/media3/transformer/SilentAudioGenerator.java

@@ -26,7 +26,9 @@ import java.util.concurrent.atomic.AtomicLong;
 /* package */ final class SilentAudioGenerator {
   private static final int DEFAULT_BUFFER_SIZE_FRAMES = 1024;
 
-  private final AudioFormat audioFormat;
+  /** The {@link AudioFormat} of the silent audio generated. */
+  public final AudioFormat audioFormat;
+
   private final ByteBuffer internalBuffer;
   private final AtomicLong remainingBytesToOutput;
 

libraries/transformer/src/main/java/androidx/media3/transformer/TransformerInternal.java

@@ -555,8 +555,7 @@ import org.checkerframework.checker.nullness.qual.MonotonicNonNull;
               addedTrackInfo.firstAssetLoaderInputFormat,
               /* firstPipelineInputFormat= */ firstAssetLoaderOutputFormat,
               transformationRequest,
-              firstEditedMediaItem.flattenForSlowMotion,
+              firstEditedMediaItem,
-              firstEditedMediaItem.effects.audioProcessors,
               encoderFactory,
               muxerWrapper,
               fallbackListener);

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

@@ -19,6 +19,7 @@ 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_RAW_STEREO_48000KHZ;
 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.FILE_AUDIO_VIDEO_STEREO;
@@ -43,7 +44,6 @@ 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;
 
@@ -233,7 +233,7 @@ public final class SequenceExportTest {
   }
 
   @Test
-  public void start_concatenateSameAudioItem_completesSuccessfully() throws Exception {
+  public void concatenateTwoAudioItems_withSameFormat_completesSuccessfully() throws Exception {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
     MediaItem audioOnlyMediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
@@ -253,7 +253,8 @@ public final class SequenceExportTest {
   }
 
   @Test
-  public void start_concatenateSameAudioItemWithEffects_completesSuccessfully() throws Exception {
+  public void concatenateTwoAudioItems_withSameFormatAndSameEffects_completesSuccessfully()
+      throws Exception {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
     MediaItem audioOnlyMediaItem = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
@@ -277,8 +278,7 @@ public final class SequenceExportTest {
   }
 
   @Test
-  @Ignore("Handle MediaItem effects changes (See [internal: b/274093424]).")
+  public void concatenateTwoAudioItems_withSameFormatAndDiffEffects_completesSuccessfully()
-  public void start_concatenateSameAudioItemWithDifferentEffects_completesSuccessfully()
       throws Exception {
     Transformer transformer =
         createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
@@ -313,4 +313,99 @@ public final class SequenceExportTest {
         checkNotNull(testMuxerHolder.testMuxer),
         getDumpFileName(FILE_AUDIO_RAW + ".high_pitch_then_low_pitch"));
   }
+
+  @Test
+  public void concatenateTwoAudioItems_withDiffFormat_completesSuccessfully() throws Exception {
+    Transformer transformer =
+        createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
+    MediaItem stereo48000Audio =
+        MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW_STEREO_48000KHZ);
+    MediaItem mono44100Audio = MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW);
+    EditedMediaItemSequence editedMediaItemSequence =
+        new EditedMediaItemSequence(
+            ImmutableList.of(
+                new EditedMediaItem.Builder(stereo48000Audio).build(),
+                new EditedMediaItem.Builder(mono44100Audio).build()));
+    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_STEREO_48000KHZ + "_then_sample.wav"));
+  }
+
+  @Test
+  public void concatenateTwoAudioItems_withDiffFormatAndSameEffects_completesSuccessfully()
+      throws Exception {
+    Transformer transformer =
+        createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
+
+    Effects highPitch =
+        new Effects(
+            ImmutableList.of(createPitchChangingAudioProcessor(/* pitch= */ 2f)),
+            /* videoEffects= */ ImmutableList.of());
+
+    EditedMediaItem stereo48000Audio =
+        new EditedMediaItem.Builder(
+                MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW_STEREO_48000KHZ))
+            .setEffects(highPitch)
+            .build();
+    EditedMediaItem mono44100Audio =
+        new EditedMediaItem.Builder(MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW))
+            .setEffects(highPitch)
+            .build();
+
+    EditedMediaItemSequence editedMediaItemSequence =
+        new EditedMediaItemSequence(ImmutableList.of(stereo48000Audio, mono44100Audio));
+    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_STEREO_48000KHZ + "-high_pitch_then_sample.wav-high_pitch"));
+  }
+
+  @Test
+  public void concatenateTwoAudioItems_withDiffFormatAndDiffEffects_completesSuccessfully()
+      throws Exception {
+    Transformer transformer =
+        createTransformerBuilder(testMuxerHolder, /* enableFallback= */ false).build();
+
+    EditedMediaItem stereo48000AudioHighPitch =
+        new EditedMediaItem.Builder(
+                MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW_STEREO_48000KHZ))
+            .setEffects(
+                new Effects(
+                    ImmutableList.of(createPitchChangingAudioProcessor(/* pitch= */ 2f)),
+                    /* videoEffects= */ ImmutableList.of()))
+            .build();
+    EditedMediaItem mono44100AudioLowPitch =
+        new EditedMediaItem.Builder(MediaItem.fromUri(ASSET_URI_PREFIX + FILE_AUDIO_RAW))
+            .setEffects(
+                new Effects(
+                    ImmutableList.of(createPitchChangingAudioProcessor(/* pitch= */ 0.5f)),
+                    /* videoEffects= */ ImmutableList.of()))
+            .build();
+    EditedMediaItemSequence editedMediaItemSequence =
+        new EditedMediaItemSequence(
+            ImmutableList.of(stereo48000AudioHighPitch, mono44100AudioLowPitch));
+    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_STEREO_48000KHZ + "-high_pitch_then_sample.wav-low_pitch"));
+  }
 }

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

@@ -153,6 +153,7 @@ public final class TestUtil {
   public static final String FILE_AUDIO_VIDEO_INCREASING_TIMESTAMPS_15S =
       "mp4/sample_with_increasing_timestamps_320w_240h.mp4";
   public static final String FILE_AUDIO_RAW = "wav/sample.wav";
+  public static final String FILE_AUDIO_RAW_STEREO_48000KHZ = "wav/sample_rf64.wav";
   public static final String FILE_WITH_SUBTITLES = "mkv/sample_with_srt.mkv";
   public static final String FILE_WITH_SEF_SLOW_MOTION = "mp4/sample_sef_slow_motion.mp4";
   public static final String FILE_AUDIO_UNSUPPORTED_BY_DECODER = "amr/sample_wb.amr";
@@ -168,8 +169,8 @@ public final class TestUtil {
   public static void createEncodersAndDecoders() {
     ShadowMediaCodec.CodecConfig codecConfig =
         new ShadowMediaCodec.CodecConfig(
-            /* inputBufferSize= */ 10_000,
+            /* inputBufferSize= */ 100_000,
-            /* outputBufferSize= */ 10_000,
+            /* outputBufferSize= */ 100_000,
             /* codec= */ (in, out) -> out.put(in));
     addCodec(
         MimeTypes.AUDIO_AAC,
@@ -194,8 +195,8 @@ public final class TestUtil {
 
     ShadowMediaCodec.CodecConfig throwingCodecConfig =
         new ShadowMediaCodec.CodecConfig(
-            /* inputBufferSize= */ 10_000,
+            /* inputBufferSize= */ 100_000,
-            /* outputBufferSize= */ 10_000,
+            /* outputBufferSize= */ 100_000,
             new ShadowMediaCodec.CodecConfig.Codec() {
 
               @Override