diff --git a/RELEASENOTES.md b/RELEASENOTES.md
index 48e5b6b1e9..2483a1954e 100644
--- a/RELEASENOTES.md
+++ b/RELEASENOTES.md
@@ -2,6 +2,9 @@
 
 ### Unreleased changes
 
+*   Extractors:
+    *   Add AC-4 Level-4 ISO base media file format support
+        ([#1265](https://github.com/androidx/media/pull/1265)).
 *   Text:
     *   Fix garbled CEA-608 subtitles in content with more than one SEI message
         per sample.
diff --git a/libraries/common/src/main/java/androidx/media3/common/util/Util.java b/libraries/common/src/main/java/androidx/media3/common/util/Util.java
index 45a408a24e..bec66b385b 100644
--- a/libraries/common/src/main/java/androidx/media3/common/util/Util.java
+++ b/libraries/common/src/main/java/androidx/media3/common/util/Util.java
@@ -2250,6 +2250,24 @@ public static int getAudioTrackChannelConfig(int channelCount) {
         }
       case 12:
         return AudioFormat.CHANNEL_OUT_7POINT1POINT4;
+      case 24:
+        if (Util.SDK_INT >= 32) {
+          return AudioFormat.CHANNEL_OUT_7POINT1POINT4
+              | AudioFormat.CHANNEL_OUT_FRONT_LEFT_OF_CENTER
+              | AudioFormat.CHANNEL_OUT_FRONT_RIGHT_OF_CENTER
+              | AudioFormat.CHANNEL_OUT_BACK_CENTER
+              | AudioFormat.CHANNEL_OUT_TOP_CENTER
+              | AudioFormat.CHANNEL_OUT_TOP_FRONT_CENTER
+              | AudioFormat.CHANNEL_OUT_TOP_BACK_CENTER
+              | AudioFormat.CHANNEL_OUT_TOP_SIDE_LEFT
+              | AudioFormat.CHANNEL_OUT_TOP_SIDE_RIGHT
+              | AudioFormat.CHANNEL_OUT_BOTTOM_FRONT_LEFT
+              | AudioFormat.CHANNEL_OUT_BOTTOM_FRONT_RIGHT
+              | AudioFormat.CHANNEL_OUT_BOTTOM_FRONT_CENTER
+              | AudioFormat.CHANNEL_OUT_LOW_FREQUENCY_2;
+        } else {
+          return AudioFormat.CHANNEL_INVALID;
+        }
       default:
         return AudioFormat.CHANNEL_INVALID;
     }
diff --git a/libraries/exoplayer/src/main/java/androidx/media3/exoplayer/trackselection/DefaultTrackSelector.java b/libraries/exoplayer/src/main/java/androidx/media3/exoplayer/trackselection/DefaultTrackSelector.java
index 0bd8b7ac81..ec51047be3 100644
--- a/libraries/exoplayer/src/main/java/androidx/media3/exoplayer/trackselection/DefaultTrackSelector.java
+++ b/libraries/exoplayer/src/main/java/androidx/media3/exoplayer/trackselection/DefaultTrackSelector.java
@@ -4224,18 +4224,22 @@ public boolean isEnabled() {
 
     public boolean canBeSpatialized(AudioAttributes audioAttributes, Format format) {
       int linearChannelCount;
-      if (Objects.equals(format.sampleMimeType, MimeTypes.AUDIO_E_AC3_JOC)
-          && format.channelCount == 16) {
+      if (Objects.equals(format.sampleMimeType, MimeTypes.AUDIO_E_AC3_JOC)) {
         // For E-AC3 JOC, the format is object based. When the channel count is 16, this maps to 12
         // linear channels and the rest are used for objects. See
         // https://github.com/google/ExoPlayer/pull/10322#discussion_r895265881
-        linearChannelCount = 12;
-      } else if (Objects.equals(format.sampleMimeType, MimeTypes.AUDIO_IAMF)
-          && format.channelCount == Format.NO_VALUE) {
+        linearChannelCount = format.channelCount == 16 ? 12 : format.channelCount;
+      } else if (Objects.equals(format.sampleMimeType, MimeTypes.AUDIO_IAMF)) {
         // IAMF with no channel count specified, assume 5.1 channels. This depends on
         // IamfDecoder.SPATIALIZED_OUTPUT_LAYOUT being set to AudioFormat.CHANNEL_OUT_5POINT1. Any
         // changes to that constant will require updates to this logic.
-        linearChannelCount = 6;
+        linearChannelCount = format.channelCount == Format.NO_VALUE ? 6 : format.channelCount;
+      } else if (Objects.equals(format.sampleMimeType, MimeTypes.AUDIO_AC4)) {
+        // For AC-4 level 3 or level 4, the format may be object based. When the channel count is
+        // 18 (level 3 17.1 OBI) or 21 (level 4 20.1 OBI), it is mapped to 24 linear channels (some
+        // channels are used for metadata transfer).
+        linearChannelCount =
+            (format.channelCount == 18 || format.channelCount == 21) ? 24 : format.channelCount;
       } else {
         linearChannelCount = format.channelCount;
       }
diff --git a/libraries/extractor/src/main/java/androidx/media3/extractor/Ac4Util.java b/libraries/extractor/src/main/java/androidx/media3/extractor/Ac4Util.java
index 3386f73b41..d895cae984 100644
--- a/libraries/extractor/src/main/java/androidx/media3/extractor/Ac4Util.java
+++ b/libraries/extractor/src/main/java/androidx/media3/extractor/Ac4Util.java
@@ -15,14 +15,22 @@
  */
 package androidx.media3.extractor;
 
+import static java.lang.annotation.ElementType.TYPE_USE;
+import static java.lang.annotation.RetentionPolicy.SOURCE;
+
+import androidx.annotation.IntDef;
 import androidx.annotation.Nullable;
 import androidx.media3.common.C;
 import androidx.media3.common.DrmInitData;
 import androidx.media3.common.Format;
 import androidx.media3.common.MimeTypes;
+import androidx.media3.common.ParserException;
 import androidx.media3.common.util.ParsableBitArray;
 import androidx.media3.common.util.ParsableByteArray;
 import androidx.media3.common.util.UnstableApi;
+import java.lang.annotation.Documented;
+import java.lang.annotation.Retention;
+import java.lang.annotation.Target;
 import java.nio.ByteBuffer;
 
 /** Utility methods for parsing AC-4 frames, which are access units in AC-4 bitstreams. */
@@ -57,6 +65,52 @@ private SyncFrameInfo(
     }
   }
 
+  /**
+   * Types of channel modes for AC-4 audio as per ETSI TS 103 190-2 V1.1.1 (2015-09), Section
+   * 6.3.2.7.2, Table 79.
+   */
+  @Documented
+  @Retention(SOURCE)
+  @Target(TYPE_USE)
+  @IntDef({
+    CHANNEL_MODE_UNKNOWN,
+    CHANNEL_MODE_MONO,
+    CHANNEL_MODE_STEREO,
+    CHANNEL_MODE_3_0,
+    CHANNEL_MODE_5_0,
+    CHANNEL_MODE_5_1,
+    CHANNEL_MODE_7_0_34,
+    CHANNEL_MODE_7_1_34,
+    CHANNEL_MODE_7_0_52,
+    CHANNEL_MODE_7_1_52,
+    CHANNEL_MODE_7_0_322,
+    CHANNEL_MODE_7_1_322,
+    CHANNEL_MODE_7_0_4,
+    CHANNEL_MODE_7_1_4,
+    CHANNEL_MODE_9_0_4,
+    CHANNEL_MODE_9_1_4,
+    CHANNEL_MODE_22_2
+  })
+  private @interface ChannelMode {}
+
+  private static final int CHANNEL_MODE_UNKNOWN = -1;
+  private static final int CHANNEL_MODE_MONO = 0;
+  private static final int CHANNEL_MODE_STEREO = 1;
+  private static final int CHANNEL_MODE_3_0 = 2;
+  private static final int CHANNEL_MODE_5_0 = 3;
+  private static final int CHANNEL_MODE_5_1 = 4;
+  private static final int CHANNEL_MODE_7_0_34 = 5;
+  private static final int CHANNEL_MODE_7_1_34 = 6;
+  private static final int CHANNEL_MODE_7_0_52 = 7;
+  private static final int CHANNEL_MODE_7_1_52 = 8;
+  private static final int CHANNEL_MODE_7_0_322 = 9;
+  private static final int CHANNEL_MODE_7_1_322 = 10;
+  private static final int CHANNEL_MODE_7_0_4 = 11;
+  private static final int CHANNEL_MODE_7_1_4 = 12;
+  private static final int CHANNEL_MODE_9_0_4 = 13;
+  private static final int CHANNEL_MODE_9_1_4 = 14;
+  private static final int CHANNEL_MODE_22_2 = 15;
+
   public static final int AC40_SYNCWORD = 0xAC40;
   public static final int AC41_SYNCWORD = 0xAC41;
 
@@ -104,28 +158,467 @@ private SyncFrameInfo(
 
   /**
    * Returns the AC-4 format given {@code data} containing the AC4SpecificBox according to ETSI TS
-   * 103 190-1 Annex E. The reading position of {@code data} will be modified.
+   * 103 190-1 Annex E.4 (ac4_dsi) and TS 103 190-2 section E.6 (ac4_dsi_v1). The reading position
+   * of {@code data} will be modified.
    *
    * @param data The AC4SpecificBox to parse.
    * @param trackId The track identifier to set on the format.
    * @param language The language to set on the format.
    * @param drmInitData {@link DrmInitData} to be included in the format.
    * @return The AC-4 format parsed from data in the header.
+   * @throws ParserException If an unsupported container feature is encountered while parsing AC-4
+   *     Annex E.
    */
   public static Format parseAc4AnnexEFormat(
-      ParsableByteArray data, String trackId, String language, @Nullable DrmInitData drmInitData) {
-    data.skipBytes(1); // ac4_dsi_version, bitstream_version[0:5]
-    int sampleRate = ((data.readUnsignedByte() & 0x20) >> 5 == 1) ? 48000 : 44100;
+      ParsableByteArray data, String trackId, String language, @Nullable DrmInitData drmInitData)
+      throws ParserException {
+    ParsableBitArray dataBitArray = new ParsableBitArray();
+    dataBitArray.reset(data);
+
+    int dsiSize = dataBitArray.bitsLeft();
+    int ac4DsiVersion = dataBitArray.readBits(3); // ac4_dsi_version
+    if (ac4DsiVersion > 1) {
+      throw ParserException.createForUnsupportedContainerFeature(
+          "Unsupported AC-4 DSI version: " + ac4DsiVersion);
+    }
+
+    int bitstreamVersion = dataBitArray.readBits(7); // bitstream_version
+    int sampleRate = dataBitArray.readBit() ? 48000 : 44100; // fs_index
+    dataBitArray.skipBits(4); // frame_rate_index
+    int numberOfPresentations = dataBitArray.readBits(9); // n_presentations
+
+    if (bitstreamVersion > 1) {
+      if (ac4DsiVersion == 0) {
+        throw ParserException.createForUnsupportedContainerFeature(
+            "Invalid AC-4 DSI version: " + ac4DsiVersion);
+      }
+      if (dataBitArray.readBit()) { // b_program_id
+        dataBitArray.skipBits(16); // short_program_id
+        if (dataBitArray.readBit()) { // b_uuid
+          dataBitArray.skipBits(16 * 8); // program_uuid
+        }
+      }
+    }
+
+    if (ac4DsiVersion == 1) {
+      if (!skipDsiBitrate(dataBitArray)) {
+        throw ParserException.createForUnsupportedContainerFeature("Invalid AC-4 DSI bitrate.");
+      }
+      dataBitArray.byteAlign();
+    }
+
+    Ac4Presentation ac4Presentation = new Ac4Presentation();
+    for (int presentationIdx = 0; presentationIdx < numberOfPresentations; presentationIdx++) {
+      boolean isSingleSubstream = false;
+      boolean isSingleSubstreamGroup = false;
+      int presentationConfig;
+      int presentationVersion;
+      int presentationBytes = 0;
+      int start = 0;
+
+      if (ac4DsiVersion == 0) {
+        isSingleSubstream = dataBitArray.readBit(); // b_single_substream
+        presentationConfig = dataBitArray.readBits(5); // presentation_config
+        presentationVersion = dataBitArray.readBits(5); // presentation_version
+      } else {
+        presentationVersion = dataBitArray.readBits(8); // presentation_version
+        presentationBytes = dataBitArray.readBits(8); // pres_bytes
+        if (presentationBytes == 0xff) {
+          presentationBytes += dataBitArray.readBits(16); // pres_bytes
+        }
+        if (presentationVersion > 2) {
+          dataBitArray.skipBits(presentationBytes * 8);
+          continue;
+        }
+        // record a marker, less the size of the presentation_config
+        start = (dsiSize - dataBitArray.bitsLeft()) / 8;
+        // ac4_presentation_v0_dsi(), ac4_presentation_v1_dsi() and ac4_presentation_v2_dsi()
+        // all start with a presentation_config of 5 bits
+        presentationConfig = dataBitArray.readBits(5); // presentation_config
+        isSingleSubstreamGroup = (presentationConfig == 0x1f);
+      }
+
+      boolean addEmdfSubstreams;
+      if (!(isSingleSubstream || isSingleSubstreamGroup) && presentationConfig == 6) {
+        addEmdfSubstreams = true;
+      } else {
+        ac4Presentation.level = dataBitArray.readBits(3); // mdcompat
+
+        if (dataBitArray.readBit()) { // b_presentation_group_index
+          dataBitArray.skipBits(5); // group_index
+        }
+
+        dataBitArray.skipBits(2); // dsi_frame_rate_multiply_info
+        if (ac4DsiVersion == 1 && (presentationVersion == 1 || presentationVersion == 2)) {
+          dataBitArray.skipBits(2); // dsi_frame_rate_fraction_info
+        }
+        dataBitArray.skipBits(5); // presentation_emdf_version
+        dataBitArray.skipBits(10); // presentation_key_id
+
+        if (ac4DsiVersion == 1) {
+          if (presentationVersion > 0) {
+            ac4Presentation.isChannelCoded = dataBitArray.readBit(); // b_presentation_channel_coded
+          }
+
+          if (ac4Presentation.isChannelCoded) {
+            if (presentationVersion == 1 || presentationVersion == 2) {
+              int channelMode = dataBitArray.readBits(5); // dsi_presentation_ch_mode
+              if (channelMode >= 0 && channelMode <= 15) {
+                ac4Presentation.channelMode = channelMode;
+              }
+
+              if (channelMode >= 11 && channelMode <= 14) {
+                ac4Presentation.hasBackChannels =
+                    dataBitArray.readBit(); // pres_b_4_back_channels_present
+                ac4Presentation.topChannelPairs =
+                    dataBitArray.readBits(2); // pres_top_channel_pairs
+              }
+            }
+            // presentation_channel_mask in ac4_presentation_v0_dsi()
+            dataBitArray.skipBits(24); // presentation_channel_mask_v1
+          }
+
+          if (presentationVersion == 1 || presentationVersion == 2) {
+            if (dataBitArray.readBit()) { // b_presentation_core_differs
+              if (dataBitArray.readBit()) { // b_presentation_core_channel_coded
+                dataBitArray.skipBits(2); // dsi_presentation_channel_mode_core
+              }
+            }
+            if (dataBitArray.readBit()) { // b_presentation_filter
+              // Ignore b_enable_presentation field since this flag occurs in AC-4 elementary stream
+              // TOC and AC-4 decoder doesn't handle it either.
+              dataBitArray.skipBit(); // b_enable_presentation
+              int filterBytes = dataBitArray.readBits(8); // n_filter_bytes
+              for (int i = 0; i < filterBytes; i++) {
+                dataBitArray.skipBits(8); // filter_data
+              }
+            }
+          }
+        }
+
+        if (isSingleSubstream || isSingleSubstreamGroup) {
+          if (presentationVersion == 0) {
+            parseDsiSubstream(dataBitArray, ac4Presentation);
+          } else {
+            parseDsiSubstreamGroup(dataBitArray, ac4Presentation);
+          }
+        } else {
+          // b_hsf_ext for ac4DsiVersion 0 OR b_multi_pid for ac4DsiVersion 1
+          dataBitArray.skipBit();
+          switch (presentationConfig) {
+            case 0:
+            case 1:
+            case 2:
+              if (presentationVersion == 0) {
+                for (int substreamId = 0; substreamId < 2; substreamId++) {
+                  parseDsiSubstream(dataBitArray, ac4Presentation);
+                }
+              } else {
+                for (int substreamGroupId = 0; substreamGroupId < 2; substreamGroupId++) {
+                  parseDsiSubstreamGroup(dataBitArray, ac4Presentation);
+                }
+              }
+              break;
+            case 3:
+            case 4:
+              if (presentationVersion == 0) {
+                for (int substreamId = 0; substreamId < 3; substreamId++) {
+                  parseDsiSubstream(dataBitArray, ac4Presentation);
+                }
+              } else {
+                for (int substreamGroupId = 0; substreamGroupId < 3; substreamGroupId++) {
+                  parseDsiSubstreamGroup(dataBitArray, ac4Presentation);
+                }
+              }
+              break;
+            case 5:
+              if (presentationVersion == 0) {
+                parseDsiSubstream(dataBitArray, ac4Presentation);
+              } else {
+                int nSubstreamGroupsMinus2 = dataBitArray.readBits(3);
+                for (int substreamGroupId = 0;
+                    substreamGroupId < nSubstreamGroupsMinus2 + 2;
+                    substreamGroupId++) {
+                  parseDsiSubstreamGroup(dataBitArray, ac4Presentation);
+                }
+              }
+              break;
+            default:
+              int nSkipBytes = dataBitArray.readBits(7); // n_skip_bytes
+              for (int j = 0; j < nSkipBytes; j++) {
+                dataBitArray.skipBits(8);
+              }
+              break;
+          }
+        }
+        dataBitArray.skipBit(); // b_pre_virtualized
+        addEmdfSubstreams = dataBitArray.readBit(); // b_add_emdf_substreams
+      }
+      if (addEmdfSubstreams) {
+        int nAddEmdfSubstreams = dataBitArray.readBits(7); // n_add_emdf_substreams
+        for (int j = 0; j < nAddEmdfSubstreams; j++) {
+          dataBitArray.skipBits(5 + 10); // substream_emdf_version and substream_key_id
+        }
+      }
+
+      if (presentationVersion > 0) {
+        if (dataBitArray.readBit()) { // b_presentation_bitrate_info
+          if (!skipDsiBitrate(dataBitArray)) {
+            throw ParserException.createForUnsupportedContainerFeature("Can't parse bitrate DSI.");
+          }
+        }
+
+        if (dataBitArray.readBit()) { // b_alternative
+          dataBitArray.byteAlign();
+          int nameLen = dataBitArray.readBits(16); // name_len
+          dataBitArray.skipBytes(nameLen); // presentation_name
+
+          int nTargets = dataBitArray.readBits(5); // n_targets
+          for (int i = 0; i < nTargets; i++) {
+            dataBitArray.skipBits(3); // target_md_compat
+            dataBitArray.skipBits(8); // target_device_category
+          }
+        }
+      }
+
+      dataBitArray.byteAlign();
+
+      if (ac4DsiVersion == 1) {
+        int end = (dsiSize - dataBitArray.bitsLeft()) / 8;
+        int presentationBytesRead = end - start;
+        if (presentationBytes < presentationBytesRead) {
+          throw ParserException.createForUnsupportedContainerFeature(
+              "pres_bytes is smaller than presentation bytes read.");
+        }
+        int skipBytes = presentationBytes - presentationBytesRead;
+        dataBitArray.skipBytes(skipBytes);
+      }
+      // We should know this or something is probably wrong with the bitstream (or we don't support
+      // it)
+      if (ac4Presentation.isChannelCoded && ac4Presentation.channelMode == CHANNEL_MODE_UNKNOWN) {
+        throw ParserException.createForUnsupportedContainerFeature(
+            "Can't determine channel mode of presentation " + presentationIdx);
+      }
+      break; // Successfully parsed the first presentation with presentation version 0, 1 or 2.
+    }
+
+    int channelCount;
+    if (ac4Presentation.isChannelCoded) {
+      channelCount =
+          getAdjustedChannelCount(
+              ac4Presentation.channelMode,
+              ac4Presentation.hasBackChannels,
+              ac4Presentation.topChannelPairs);
+    } else {
+      channelCount = ac4Presentation.numOfUmxObjects;
+      // TODO: There is a bug in ETSI TS 103 190-2 V1.2.1 (2018-02), E.11.11
+      // For AC-4 level 4 stream, the intention is to set 19 to n_umx_objects_minus1 but it is
+      // equal to 15 based on current specification. Dolby has filed a bug report to ETSI.
+      // The following sentence should be deleted after ETSI specification error is fixed.
+      if (ac4Presentation.level == 4) {
+        channelCount = channelCount == 16 ? 21 : channelCount;
+      }
+    }
+
+    if (channelCount <= 0) {
+      throw ParserException.createForUnsupportedContainerFeature(
+          "Can't determine channel count of presentation.");
+    }
+
     return new Format.Builder()
         .setId(trackId)
         .setSampleMimeType(MimeTypes.AUDIO_AC4)
-        .setChannelCount(CHANNEL_COUNT_2)
+        .setChannelCount(channelCount)
         .setSampleRate(sampleRate)
         .setDrmInitData(drmInitData)
         .setLanguage(language)
         .build();
   }
 
+  /**
+   * Parses the AC-4 DSI substream according to TS 103 190-1 v1.2.1 section E.5 and TS 103 190-2
+   * v1.1.1 section E.9. Modifies the reading position of {@code data} to be just after the AC-4 DSI
+   * substream field.
+   *
+   * @param data The {@link ParsableBitArray} containing the AC-4 DSI, positioned at the start of
+   *     the substream field.
+   * @param ac4Presentation A structure to store parsed AC-4 presentation information.
+   * @throws ParserException If an unsupported container feature is encountered while parsing the
+   *     AC-4 DSI substream.
+   */
+  private static void parseDsiSubstream(ParsableBitArray data, Ac4Presentation ac4Presentation)
+      throws ParserException {
+    int channelMode = data.readBits(5); // channel_mode
+    data.skipBits(2); // dsi_sf_multiplier
+
+    if (data.readBit()) { // b_bitrate_indicator
+      data.skipBits(5); // b_bitrate_indicator
+    }
+    if (channelMode >= 7 && channelMode <= 10) {
+      data.skipBit(); // add_ch_base
+    }
+
+    if (data.readBit()) { // b_content_type
+      int contentClassifier = data.readBits(3); // content_classifier
+      // For streams based on TS 103 190 part 1 the presentation level channel_mode doesn't exist
+      // and so we use the channel_mode from either the CM or M&E substream (they are mutually
+      // exclusive).
+      if (ac4Presentation.channelMode == CHANNEL_MODE_UNKNOWN
+          && (channelMode >= 0 && channelMode <= 15)
+          && (contentClassifier == 0 || contentClassifier == 1)) {
+        ac4Presentation.channelMode = channelMode;
+      }
+
+      if (data.readBit()) { // b_language_indicator
+        skipDsiLanguage(data);
+      }
+    }
+  }
+
+  /**
+   * Parses the AC-4 DSI substream group information according to ETSI TS 103 190-2 v1.1.1 section
+   * E.11. Modifies the reading position of {@code data} to be just after the AC-4 DSI substream
+   * group field.
+   *
+   * @param data The {@link ParsableBitArray} containing the AC-4 DSI, positioned at the start of
+   *     the substream group field.
+   * @param ac4Presentation A structure to store parsed AC-4 presentation information.
+   * @throws ParserException If an unsupported container feature is encountered while parsing the
+   *     AC-4 DSI substream group.
+   */
+  private static void parseDsiSubstreamGroup(ParsableBitArray data, Ac4Presentation ac4Presentation)
+      throws ParserException {
+    data.skipBits(2); // b_substreams_present(1), b_hsf_ext(1)
+    boolean channelCoded = data.readBit(); // b_channel_coded
+    int numberOfSubstreams = data.readBits(8); // n_substreams
+
+    for (int i = 0; i < numberOfSubstreams; i++) {
+      data.skipBits(2); // dsi_sf_multiplier
+      if (data.readBit()) { // b_substream_bitrate_indicator
+        data.skipBits(5); // substream_bitrate_indicator
+      }
+      if (channelCoded) {
+        data.skipBits(24); // dsi_substream_channel_mask
+      } else {
+        if (data.readBit()) { // b_ajoc
+          if (!data.readBit()) { // b_static_dmx
+            data.skipBits(4); // n_dmx_objects_minus1
+          }
+          ac4Presentation.numOfUmxObjects = data.readBits(6) + 1; // n_umx_objects_minus1
+        }
+        data.skipBits(4); // objects_assignment_mask
+      }
+    }
+
+    if (data.readBit()) { // b_content_type
+      data.skipBits(3); // content_classifier
+
+      if (data.readBit()) { // b_language_indicator
+        skipDsiLanguage(data);
+      }
+    }
+  }
+
+  /**
+   * Skips the language information fields in an AC-4 DSI bit stream according to TS 103 190-1
+   * section 4.3.3.8.7. Modifies the reading position of {@code data} to be just after the language
+   * fields.
+   *
+   * @param data The {@link ParsableBitArray} containing the AC-4 DSI, positioned at the start of
+   *     the language tag field.
+   * @throws ParserException If the language tag length is invalid.
+   */
+  private static void skipDsiLanguage(ParsableBitArray data) throws ParserException {
+    int languageTagBytesNumber = data.readBits(6); // n_language_tag_bytes
+    if (languageTagBytesNumber < 2 || languageTagBytesNumber > 42) {
+      throw ParserException.createForUnsupportedContainerFeature(
+          String.format(
+              "Invalid language tag bytes number: %d. Must be between 2 and 42.",
+              languageTagBytesNumber));
+    }
+    // Can't use readBytes() since it is not byte-aligned here.
+    data.skipBits(languageTagBytesNumber * C.BITS_PER_BYTE);
+  }
+
+  /**
+   * Skips the bitrate information fields in an AC-4 DSI bit stream. The reading position of {@code
+   * data} will be modified to be just after the bitrate fields, if sufficient bits remain.
+   *
+   * @param data The {@link ParsableBitArray} containing the AC-4 DSI, positioned at the start of
+   *     the bitrate information.
+   * @return {@code true} if the bitrate fields were successfully skipped or {@code false} if there
+   *     were insufficient bits remaining in {@code data}.
+   */
+  private static boolean skipDsiBitrate(ParsableBitArray data) {
+    int totalBitsToSkip = 2 + 32 + 32; // bit_rate_mode, bit_rate, bit_rate_precision
+    if (data.bitsLeft() < totalBitsToSkip) {
+      return false;
+    }
+    data.skipBits(totalBitsToSkip);
+    return true;
+  }
+
+  private static int getAdjustedChannelCount(
+      @ChannelMode int channelMode, boolean hasBackChannels, int topChannelPairs) {
+    int channelCount = getChannelCountFromChannelMode(channelMode);
+    if (channelMode == CHANNEL_MODE_7_0_4
+        || channelMode == CHANNEL_MODE_7_1_4
+        || channelMode == CHANNEL_MODE_9_0_4
+        || channelMode == CHANNEL_MODE_9_1_4) {
+
+      if (!hasBackChannels) {
+        channelCount -= 2;
+      }
+
+      switch (topChannelPairs) {
+        case 0:
+          channelCount -= 4;
+          break;
+        case 1:
+          channelCount -= 2;
+          break;
+        default:
+          break;
+      }
+    }
+    return channelCount;
+  }
+
+  private static int getChannelCountFromChannelMode(@ChannelMode int channelMode) {
+    switch (channelMode) {
+      case CHANNEL_MODE_MONO:
+        return 1;
+      case CHANNEL_MODE_STEREO:
+        return 2;
+      case CHANNEL_MODE_3_0:
+        return 3;
+      case CHANNEL_MODE_5_0:
+        return 5;
+      case CHANNEL_MODE_5_1:
+        return 6;
+      case CHANNEL_MODE_7_0_34:
+      case CHANNEL_MODE_7_0_52:
+      case CHANNEL_MODE_7_0_322:
+        return 7;
+      case CHANNEL_MODE_7_1_34:
+      case CHANNEL_MODE_7_1_52:
+      case CHANNEL_MODE_7_1_322:
+        return 8;
+      case CHANNEL_MODE_7_0_4:
+        return 11;
+      case CHANNEL_MODE_7_1_4:
+        return 12;
+      case CHANNEL_MODE_9_0_4:
+        return 13;
+      case CHANNEL_MODE_9_1_4:
+        return 14;
+      case CHANNEL_MODE_22_2:
+        return 24;
+      default:
+        return -1;
+    }
+  }
+
   /**
    * Returns AC-4 format information given {@code data} containing a syncframe. The reading position
    * of {@code data} will be modified.
@@ -255,5 +748,24 @@ private static int readVariableBits(ParsableBitArray data, int bitsPerRead) {
     return value;
   }
 
+  /** Holds AC-4 presentation information. */
+  private static final class Ac4Presentation {
+    public boolean isChannelCoded;
+    public @ChannelMode int channelMode;
+    public int numOfUmxObjects;
+    public boolean hasBackChannels;
+    public int topChannelPairs;
+    public int level;
+
+    private Ac4Presentation() {
+      isChannelCoded = true;
+      channelMode = CHANNEL_MODE_UNKNOWN;
+      numOfUmxObjects = -1;
+      hasBackChannels = true;
+      topChannelPairs = 2;
+      level = 0;
+    }
+  }
+
   private Ac4Util() {}
 }
diff --git a/libraries/extractor/src/test/java/androidx/media3/extractor/mp4/Mp4ExtractorParameterizedTest.java b/libraries/extractor/src/test/java/androidx/media3/extractor/mp4/Mp4ExtractorParameterizedTest.java
index 528a62d337..dbb1e82f81 100644
--- a/libraries/extractor/src/test/java/androidx/media3/extractor/mp4/Mp4ExtractorParameterizedTest.java
+++ b/libraries/extractor/src/test/java/androidx/media3/extractor/mp4/Mp4ExtractorParameterizedTest.java
@@ -107,6 +107,11 @@ public void mp4SampleWithAc4Track() throws Exception {
     assertExtractorBehavior("media/mp4/sample_ac4.mp4");
   }
 
+  @Test
+  public void mp4SampleWithAc4Level4Track() throws Exception {
+    assertExtractorBehavior("media/mp4/sample_ac4_level4.mp4");
+  }
+
   @Test
   public void mp4SampleWithEac3Track() throws Exception {
     assertExtractorBehavior("media/mp4/sample_eac3.mp4");
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.0.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.0.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.0.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.1.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.1.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.1.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.2.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.2.dump
new file mode 100644
index 0000000000..5f9737eed8
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.2.dump
@@ -0,0 +1,60 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 81350
+  sample count = 10
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 1:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 2:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 3:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 4:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 5:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 6:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 7:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 8:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 9:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.3.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.3.dump
new file mode 100644
index 0000000000..5f9737eed8
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.3.dump
@@ -0,0 +1,60 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 81350
+  sample count = 10
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 1:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 2:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 3:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 4:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 5:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 6:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 7:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 8:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 9:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.0.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.0.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.0.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.1.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.1.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.1.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.2.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.2.dump
new file mode 100644
index 0000000000..5f9737eed8
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.2.dump
@@ -0,0 +1,60 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 81350
+  sample count = 10
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 1:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 2:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 3:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 4:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 5:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 6:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 7:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 8:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 9:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.3.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.3.dump
new file mode 100644
index 0000000000..5f9737eed8
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.3.dump
@@ -0,0 +1,60 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 81350
+  sample count = 10
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 1:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 2:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 3:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 4:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 5:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 6:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 7:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 8:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 9:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.unknown_length.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.unknown_length.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.reading_within_gop_sample_dependencies.unknown_length.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.unknown_length.dump b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.unknown_length.dump
new file mode 100644
index 0000000000..90cd9d402c
--- /dev/null
+++ b/libraries/test_data/src/test/assets/extractordumps/mp4/sample_ac4_level4.mp4.unknown_length.dump
@@ -0,0 +1,100 @@
+seekMap:
+  isSeekable = true
+  duration = 853333
+  getPosition(0) = [[timeUs=0, position=665]]
+  getPosition(1) = [[timeUs=1, position=665]]
+  getPosition(426666) = [[timeUs=426666, position=81945]]
+  getPosition(853333) = [[timeUs=853333, position=81945]]
+numberOfTracks = 1
+track 0:
+  total output bytes = 162700
+  sample count = 20
+  format 0:
+    id = 1
+    sampleMimeType = audio/ac4
+    maxInputSize = 8158
+    channelCount = 21
+    sampleRate = 48000
+    language = und
+    metadata = entries=[Mp4Timestamp: creation time=3785281997, modification time=3785281997, timescale=600]
+  sample 0:
+    time = 0
+    flags = 1
+    data = length 8135, hash B524F88E
+  sample 1:
+    time = 42666
+    flags = 0
+    data = length 8135, hash FB80C2FB
+  sample 2:
+    time = 85333
+    flags = 0
+    data = length 8135, hash 907C0C31
+  sample 3:
+    time = 128000
+    flags = 0
+    data = length 8135, hash FDFBD32B
+  sample 4:
+    time = 170666
+    flags = 0
+    data = length 8135, hash 6CAF0549
+  sample 5:
+    time = 213333
+    flags = 0
+    data = length 8135, hash F5CA1C9A
+  sample 6:
+    time = 256000
+    flags = 0
+    data = length 8135, hash B1B5160D
+  sample 7:
+    time = 298666
+    flags = 0
+    data = length 8135, hash 9E923B3F
+  sample 8:
+    time = 341333
+    flags = 0
+    data = length 8135, hash B1C0BB1F
+  sample 9:
+    time = 384000
+    flags = 0
+    data = length 8135, hash 56F65A03
+  sample 10:
+    time = 426666
+    flags = 1
+    data = length 8135, hash D07FA9A1
+  sample 11:
+    time = 469333
+    flags = 0
+    data = length 8135, hash EF26FDDE
+  sample 12:
+    time = 512000
+    flags = 0
+    data = length 8135, hash 8946EEEB
+  sample 13:
+    time = 554666
+    flags = 0
+    data = length 8135, hash AC2E4C99
+  sample 14:
+    time = 597333
+    flags = 0
+    data = length 8135, hash B63A1D8
+  sample 15:
+    time = 640000
+    flags = 0
+    data = length 8135, hash 23119F0F
+  sample 16:
+    time = 682666
+    flags = 0
+    data = length 8135, hash 507972CA
+  sample 17:
+    time = 725333
+    flags = 0
+    data = length 8135, hash E574BC00
+  sample 18:
+    time = 768000
+    flags = 0
+    data = length 8135, hash 52F482FA
+  sample 19:
+    time = 810666
+    flags = 536870912
+    data = length 8135, hash C1A7B518
+tracksEnded = true
diff --git a/libraries/test_data/src/test/assets/media/mp4/sample_ac4_level4.mp4 b/libraries/test_data/src/test/assets/media/mp4/sample_ac4_level4.mp4
new file mode 100644
index 0000000000..7e110d301e
Binary files /dev/null and b/libraries/test_data/src/test/assets/media/mp4/sample_ac4_level4.mp4 differ