Merge branch 'main' into sbd

docs/source/options/dash_stream_descriptors.rst

@@ -12,5 +12,5 @@ DASH specific stream descriptor fields
 
     Optional semicolon separated list of values for DASH Role element. The
     value should be one of: **caption**, **subtitle**, **main**, **alternate**,
-    **supplementary**, **commentary**, **description** and **dub**. See 
+    **supplementary**, **commentary**, **description**, **dub** and **forced-subtitle** .
-    DASH (ISO/IEC 23009-1) specification for details.
+    See DASH (ISO/IEC 23009-1) specification for details.

docs/source/options/hls_options.rst

@@ -76,6 +76,14 @@ HLS options
     The EXT-X-MEDIA-SEQUENCE documentation can be read here:
     https://tools.ietf.org/html/rfc8216#section-4.3.3.2.
 
+--hls_start_time_offset <seconds>
+
+    Sets EXT-X-START on the media playlists to specify the preferred point
+    at wich the player should start playing.
+    A positive number indicates a time offset from the beginning of the playlist.
+    A negative number indicates a negative time offset from the end of the
+    last media segment in the playlist.
+
 --hls_only=0|1
 
     Optional. Defaults to 0 if not specified. If it is set to 1, indicates the

docs/source/options/stream_descriptors.rst

@@ -73,6 +73,15 @@ These are the available fields:
     CEA allows specifying up to 4 streams within a single video stream. If not
     specified, all subtitles will be merged together.
 
+:forced_subtitle:
+
+    Optional boolean value (0|1). If set to 1 indicates that this stream is a
+    Forced Narrative subtitle that should be displayed when subtitles are otherwise
+    off, for example used to caption short portions of the audio that might be in
+    a foreign language. For DASH this will set role to **forced_subtitle**, for HLS
+    it will set FORCED=YES and AUTOSELECT=YES. Only valid for subtitles.
+
+
 .. include:: /options/drm_stream_descriptors.rst
 .. include:: /options/dash_stream_descriptors.rst
 .. include:: /options/hls_stream_descriptors.rst

include/packager/hls_params.h

@@ -8,6 +8,7 @@
 #define PACKAGER_PUBLIC_HLS_PARAMS_H_
 
 #include <cstdint>
+#include <optional>
 #include <string>
 
 namespace shaka {
@@ -63,6 +64,12 @@ struct HlsParams {
   /// Custom EXT-X-MEDIA-SEQUENCE value to allow continuous media playback
   /// across packager restarts. See #691 for details.
   uint32_t media_sequence_number = 0;
+  /// Sets EXT-X-START on the media playlists to specify the preferred point
+  /// at wich the player should start playing.
+  /// A positive number indicates a time offset from the beginning of the
+  /// playlist. A negative number indicates a negative time offset from the end
+  /// of the last media segment in the playlist.
+  std::optional<double> start_time_offset;
 };
 
 }  // namespace shaka

include/packager/packager.h

@@ -152,6 +152,9 @@ struct StreamDescriptor {
   /// Set to true to indicate that the stream is for hls only.
   bool hls_only = false;
 
+  /// Optional, indicates if this is a Forced Narrative subtitle stream.
+  bool forced_subtitle = false;
+
   /// Optional for DASH output. It defines the Label element in Adaptation Set.
   std::string dash_label;
 };

packager/app/hls_flags.cc

@@ -6,6 +6,8 @@
 
 #include <packager/app/hls_flags.h>
 
+#include <optional>
+
 ABSL_FLAG(std::string,
           hls_master_playlist_output,
           "",
@@ -35,3 +37,12 @@ ABSL_FLAG(int32_t,
           "EXT-X-MEDIA-SEQUENCE value, which allows continuous media "
           "sequence across packager restarts. See #691 for more "
           "information about the reasoning of this and its use cases.");
+ABSL_FLAG(std::optional<double>,
+          hls_start_time_offset,
+          std::nullopt,
+          "Floating-point number. Sets EXT-X-START on the media playlists "
+          "to specify the preferred point at wich the player should start "
+          "playing. A positive number indicates a time offset from the "
+          "beginning of the playlist. A negative number indicates a "
+          "negative time offset from the end of the last media segment "
+          "in the playlist.");

packager/app/hls_flags.h

@@ -15,5 +15,6 @@ ABSL_DECLARE_FLAG(std::string, hls_base_url);
 ABSL_DECLARE_FLAG(std::string, hls_key_uri);
 ABSL_DECLARE_FLAG(std::string, hls_playlist_type);
 ABSL_DECLARE_FLAG(int32_t, hls_media_sequence_number);
+ABSL_DECLARE_FLAG(std::optional<double>, hls_start_time_offset);
 
 #endif  // PACKAGER_APP_HLS_FLAGS_H_

packager/app/packager_main.cc

@@ -120,8 +120,16 @@ const char kUsage[] =
     "    in the format: scheme_id_uri=value.\n"
     "  - dash_roles (roles): Optional semicolon separated list of values for\n"
     "    DASH Role elements. The value should be one of: caption, subtitle,\n"
-    "    main, alternate, supplementary, commentary, description and dub. See\n"
+    "    forced-subtitle, main, alternate, supplementary, commentary, \n"
-    "    DASH (ISO/IEC 23009-1) specification for details.\n";
+    "    description and dub. See DASH\n"
+    "    (ISO/IEC 23009-1) specification for details.\n"
+    "  - forced_subtitle: Optional boolean value (0|1). If set to 1 \n"
+    "    indicates that this stream is a Forced Narrative subtitle that \n"
+    "    should be displayed when subtitles are otherwise off, for example \n"
+    "    used to caption short portions of the audio that might be in a \n"
+    "    foreign language. For DASH this will set role to forced_subtitle, \n"
+    "    for HLS it will set FORCED=YES and AUTOSELECT=YES. \n"
+    "    Only valid for subtitles.";
 
 // Labels for parameters in RawKey key info.
 const char kDrmLabelLabel[] = "label";
@@ -587,6 +595,7 @@ std::optional<PackagingParams> GetPackagingParams() {
   hls_params.default_text_language = absl::GetFlag(FLAGS_default_text_language);
   hls_params.media_sequence_number =
       absl::GetFlag(FLAGS_hls_media_sequence_number);
+  hls_params.start_time_offset = absl::GetFlag(FLAGS_hls_start_time_offset);
 
   TestParams& test_params = packaging_params.test_params;
   test_params.dump_stream_info = absl::GetFlag(FLAGS_dump_stream_info);

packager/app/stream_descriptor.cc

@@ -40,6 +40,7 @@ enum FieldType {
   kDashOnlyField,
   kHlsOnlyField,
   kDashLabelField,
+  kForcedSubtitleField,
 };
 
 struct FieldNameToTypeMapping {
@@ -88,6 +89,7 @@ const FieldNameToTypeMapping kFieldNameTypeMappings[] = {
     {"dash_only", kDashOnlyField},
     {"hls_only", kHlsOnlyField},
     {"dash_label", kDashLabelField},
+    {"forced_subtitle", kForcedSubtitleField},
 };
 
 FieldType GetFieldType(const std::string& field_name) {
@@ -255,12 +257,35 @@ std::optional<StreamDescriptor> ParseStreamDescriptor(
       case kDashLabelField:
         descriptor.dash_label = pair.second;
         break;
+      case kForcedSubtitleField:
+        unsigned forced_subtitle_value;
+        if (!absl::SimpleAtoi(pair.second, &forced_subtitle_value)) {
+          LOG(ERROR) << "Non-numeric option for forced field "
+                        "specified ("
+                     << pair.second << ").";
+          return std::nullopt;
+        }
+        if (forced_subtitle_value > 1) {
+          LOG(ERROR) << "forced should be either 0 or 1.";
+          return std::nullopt;
+        }
+        descriptor.forced_subtitle = forced_subtitle_value > 0;
+        break;
       default:
         LOG(ERROR) << "Unknown field in stream descriptor (\"" << pair.first
                    << "\").";
         return std::nullopt;
     }
   }
+
+  if (descriptor.forced_subtitle) {
+    auto itr = std::find(descriptor.dash_roles.begin(),
+                         descriptor.dash_roles.end(), "forced-subtitle");
+    if (itr == descriptor.dash_roles.end()) {
+      descriptor.dash_roles.push_back("forced-subtitle");
+    }
+  }
+
   return descriptor;
 }
 

packager/app/test/packager_test.py

@@ -310,7 +310,8 @@ class PackagerAppTest(unittest.TestCase):
                  skip_encryption=None,
                  bandwidth=None,
                  split_content_on_ad_cues=False,
-                 test_file=None):
+                 test_file=None,
+                 forced_subtitle=None):
     """Get a stream descriptor as a string.
 
 
@@ -347,8 +348,9 @@ class PackagerAppTest(unittest.TestCase):
           into multiple files, with a total of NumAdCues + 1 files.
       test_file: The input file to use. If the input file is not specified, a
           default file will be used.
-
+      forced_subtitle: If set to true, it marks this as a Forced Narrative
-
+          subtitle, marked in DASH using forced-subtitle role and
+          in HLS using FORCED=YES.
     Returns:
       A string that makes up a single stream descriptor for input to the
       packager.
@@ -402,6 +404,9 @@ class PackagerAppTest(unittest.TestCase):
     if dash_only:
       stream.Append('dash_only', 1)
 
+    if forced_subtitle:
+      stream.Append('forced_subtitle', 1)
+
     if dash_label:
       stream.Append('dash_label', dash_label)
 
@@ -830,7 +835,6 @@ class PackagerFunctionalTest(PackagerAppTest):
         self._GetFlags(output_dash=True, output_hls=True))
     self._CheckTestResults('hls-only-dash-only')
 
-
   def testDashLabel(self):
     streams = [
         self._GetStream('video', dash_label='Main'),
@@ -858,6 +862,21 @@ class PackagerFunctionalTest(PackagerAppTest):
             sbd_key_audio='k5=v5:k6=v6'))
     self._CheckTestResults('sbd_test')
 
+  def testForcedSubtitle(self):
+    streams = [
+      self._GetStream('audio', hls=True),
+      self._GetStream('video', hls=True),
+    ]
+
+    streams += self._GetStreams(
+        ['text'],
+        test_files=['bear-english.vtt'],
+        forced_subtitle=True)
+
+    self.assertPackageSuccess(streams, self._GetFlags(output_dash=True,
+                                                      output_hls=True))
+    self._CheckTestResults('forced-subtitle')
+
   def testAudioVideoWithLanguageOverride(self):
     self.assertPackageSuccess(
         self._GetStreams(['audio', 'video'], language='por', hls=True),
@@ -966,6 +985,13 @@ class PackagerFunctionalTest(PackagerAppTest):
         self._GetFlags(output_dash=True))
     self._CheckTestResults('acc-he')
 
+  def testDtsx(self):
+    self.assertPackageSuccess(
+        self._GetStreams(
+            ['audio'], test_files=['bear-dtsx.mp4']),
+        self._GetFlags(output_dash=True))
+    self._CheckTestResults('dtsx-dash')
+
   def testVideoAudioWebVTT(self):
     audio_video_streams = self._GetStreams(['audio', 'video'])
     text_stream = self._GetStreams(['text'], test_files=['bear-english.vtt'])

packager/app/test/testdata/dtsx-dash/output.mpd

@@ -0,0 +1,15 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>-->
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" minBufferTime="PT2S" type="static" mediaPresentationDuration="PT3.114667S">
+  <Period id="0">
+    <AdaptationSet id="0" contentType="audio" subsegmentAlignment="true">
+      <Representation id="0" bandwidth="227665" codecs="dtsx" mimeType="audio/mp4" audioSamplingRate="48000">
+        <AudioChannelConfiguration schemeIdUri="tag:dts.com,2018:uhd:audio_channel_configuration" value="0000003F"/>
+        <BaseURL>bear-dtsx-audio.mp4</BaseURL>
+        <SegmentBase indexRange="742-821" timescale="48000">
+          <Initialization range="0-741"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

packager/app/test/testdata/forced-subtitle/bear-640x360-audio.m3u8

@@ -0,0 +1,16 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-MAP:URI="bear-640x360-audio.mp4",BYTERANGE="804@0"
+#EXTINF:1.022,
+#EXT-X-BYTERANGE:17028@872
+bear-640x360-audio.mp4
+#EXTINF:0.998,
+#EXT-X-BYTERANGE:16285
+bear-640x360-audio.mp4
+#EXTINF:0.720,
+#EXT-X-BYTERANGE:9558
+bear-640x360-audio.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-640x360-video-iframe.m3u8

@@ -0,0 +1,17 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-I-FRAMES-ONLY
+#EXT-X-MAP:URI="bear-640x360-video.mp4",BYTERANGE="870@0"
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:15581@938
+bear-640x360-video.mp4
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:18221@100251
+bear-640x360-video.mp4
+#EXTINF:0.734,
+#EXT-X-BYTERANGE:19663@222058
+bear-640x360-video.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-640x360-video.m3u8

@@ -0,0 +1,16 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXT-X-MAP:URI="bear-640x360-video.mp4",BYTERANGE="870@0"
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:99313@938
+bear-640x360-video.mp4
+#EXTINF:1.001,
+#EXT-X-BYTERANGE:121807
+bear-640x360-video.mp4
+#EXTINF:0.734,
+#EXT-X-BYTERANGE:79662
+bear-640x360-video.mp4
+#EXT-X-ENDLIST

packager/app/test/testdata/forced-subtitle/bear-english-text.vtt

@@ -0,0 +1,11 @@
+WEBVTT
+
+STYLE
+::cue { color:lime }
+
+00:00:00.000 --> 00:00:00.800 align:center
+Yup, that's a bear, eh.
+
+00:00:01.000 --> 00:00:04.700 align:center
+He 's... um... doing bear-like stuff.
+

packager/app/test/testdata/forced-subtitle/output.m3u8

@@ -0,0 +1,13 @@
+#EXTM3U
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+
+#EXT-X-INDEPENDENT-SEGMENTS
+
+#EXT-X-MEDIA:TYPE=AUDIO,URI="bear-640x360-audio.m3u8",GROUP-ID="default-audio-group",NAME="stream_0",DEFAULT=NO,AUTOSELECT=YES,CHANNELS="2"
+
+#EXT-X-MEDIA:TYPE=SUBTITLES,URI="stream_2.m3u8",GROUP-ID="default-text-group",NAME="stream_2",DEFAULT=NO,AUTOSELECT=YES,FORCED=YES
+
+#EXT-X-STREAM-INF:BANDWIDTH=1106817,AVERAGE-BANDWIDTH=1004632,CODECS="avc1.64001e,mp4a.40.2",RESOLUTION=640x360,FRAME-RATE=29.970,AUDIO="default-audio-group",SUBTITLES="default-text-group",CLOSED-CAPTIONS=NONE
+bear-640x360-video.m3u8
+
+#EXT-X-I-FRAME-STREAM-INF:BANDWIDTH=214292,AVERAGE-BANDWIDTH=156327,CODECS="avc1.64001e",RESOLUTION=640x360,CLOSED-CAPTIONS=NONE,URI="bear-640x360-video-iframe.m3u8"

packager/app/test/testdata/forced-subtitle/output.mpd

@@ -0,0 +1,29 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!--Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>-->
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 DASH-MPD.xsd" profiles="urn:mpeg:dash:profile:isoff-on-demand:2011" minBufferTime="PT2S" type="static" mediaPresentationDuration="PT2.736067S">
+  <Period id="0">
+    <AdaptationSet id="0" contentType="audio" subsegmentAlignment="true">
+      <Representation id="0" bandwidth="133334" codecs="mp4a.40.2" mimeType="audio/mp4" audioSamplingRate="44100">
+        <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
+        <BaseURL>bear-640x360-audio.mp4</BaseURL>
+        <SegmentBase indexRange="804-871" timescale="44100">
+          <Initialization range="0-803"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet id="1" contentType="video" width="640" height="360" frameRate="30000/1001" subsegmentAlignment="true" par="16:9">
+      <Representation id="1" bandwidth="973483" codecs="avc1.64001e" mimeType="video/mp4" sar="1:1">
+        <BaseURL>bear-640x360-video.mp4</BaseURL>
+        <SegmentBase indexRange="870-937" timescale="30000">
+          <Initialization range="0-869"/>
+        </SegmentBase>
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet id="2" contentType="text" subsegmentAlignment="true">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="forced-subtitle"/>
+      <Representation id="2" bandwidth="317" mimeType="text/vtt">
+        <BaseURL>bear-english-text.vtt</BaseURL>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

packager/app/test/testdata/forced-subtitle/stream_2.m3u8

@@ -0,0 +1,8 @@
+#EXTM3U
+#EXT-X-VERSION:6
+## Generated with https://github.com/shaka-project/shaka-packager version <tag>-<hash>-<test>
+#EXT-X-TARGETDURATION:5
+#EXT-X-PLAYLIST-TYPE:VOD
+#EXTINF:4.700,
+bear-english-text.vtt
+#EXT-X-ENDLIST

packager/hls/base/master_playlist.cc

@@ -318,11 +318,16 @@ void BuildMediaTag(const MediaPlaylist& playlist,
   } else {
     tag.AddString("DEFAULT", "NO");
   }
-
   if (is_autoselect) {
     tag.AddString("AUTOSELECT", "YES");
   }
 
+  if (playlist.stream_type() ==
+          MediaPlaylist::MediaPlaylistStreamType::kSubtitle &&
+      playlist.forced_subtitle()) {
+    tag.AddString("FORCED", "YES");
+  }
+
   const std::vector<std::string>& characteristics = playlist.characteristics();
   if (!characteristics.empty()) {
     tag.AddQuotedString("CHARACTERISTICS", absl::StrJoin(characteristics, ","));
@@ -401,6 +406,12 @@ void BuildMediaTags(
         }
       }
 
+      if (playlist->stream_type() ==
+              MediaPlaylist::MediaPlaylistStreamType::kSubtitle &&
+          playlist->forced_subtitle()) {
+        is_autoselect = true;
+      }
+
       BuildMediaTag(*playlist, group_id, is_default, is_autoselect, base_url,
                     out);
     }

packager/hls/base/media_playlist.cc

@@ -10,6 +10,7 @@
 #include <cinttypes>
 #include <cmath>
 #include <memory>
+#include <optional>
 
 #include <absl/log/check.h>
 #include <absl/log/log.h>
@@ -109,7 +110,8 @@ std::string CreatePlaylistHeader(
     HlsPlaylistType type,
     MediaPlaylist::MediaPlaylistStreamType stream_type,
     uint32_t media_sequence_number,
-    int discontinuity_sequence_number) {
+    int discontinuity_sequence_number,
+    std::optional<double> start_time_offset) {
   const std::string version = GetPackagerVersion();
   std::string version_line;
   if (!version.empty()) {
@@ -151,6 +153,10 @@ std::string CreatePlaylistHeader(
       MediaPlaylist::MediaPlaylistStreamType::kVideoIFramesOnly) {
     absl::StrAppendFormat(&header, "#EXT-X-I-FRAMES-ONLY\n");
   }
+  if (start_time_offset.has_value()) {
+    absl::StrAppendFormat(&header, "#EXT-X-START:TIME-OFFSET=%f\n",
+                          start_time_offset.value());
+  }
 
   // Put EXT-X-MAP at the end since the rest of the playlist is about the
   // segment and key info.
@@ -373,6 +379,10 @@ void MediaPlaylist::SetCharacteristicsForTesting(
   characteristics_ = characteristics;
 }
 
+void MediaPlaylist::SetForcedSubtitleForTesting(const bool forced_subtitle) {
+  forced_subtitle_ = forced_subtitle;
+}
+
 bool MediaPlaylist::SetMediaInfo(const MediaInfo& media_info) {
   const int32_t time_scale = GetTimeScale(media_info);
   if (time_scale == 0) {
@@ -400,6 +410,8 @@ bool MediaPlaylist::SetMediaInfo(const MediaInfo& media_info) {
       std::vector<std::string>(media_info_.hls_characteristics().begin(),
                                media_info_.hls_characteristics().end());
 
+  forced_subtitle_ = media_info_.forced_subtitle();
+
   return true;
 }
 
@@ -479,7 +491,8 @@ bool MediaPlaylist::WriteToFile(const std::filesystem::path& file_path) {
 
   std::string content = CreatePlaylistHeader(
       media_info_, target_duration_, hls_params_.playlist_type, stream_type_,
-      media_sequence_number_, discontinuity_sequence_number_);
+      media_sequence_number_, discontinuity_sequence_number_,
+      hls_params_.start_time_offset);
 
   for (const auto& entry : entries_)
     absl::StrAppendFormat(&content, "%s\n", entry->ToString().c_str());
@@ -513,8 +526,8 @@ void MediaPlaylist::SetTargetDuration(int32_t target_duration) {
   if (target_duration_set_) {
     if (target_duration_ == target_duration)
       return;
-    VLOG(1) << "Updating target duration from " << target_duration << " to "
+    VLOG(1) << "Updating target duration from " << target_duration_ << " to "
-            << target_duration_;
+            << target_duration;
   }
   target_duration_ = target_duration;
   target_duration_set_ = true;

packager/hls/base/media_playlist.h

@@ -90,6 +90,9 @@ class MediaPlaylist {
   /// For testing only.
   void SetLanguageForTesting(const std::string& language);
 
+  /// For testing only.
+  void SetForcedSubtitleForTesting(const bool forced_subtitle);
+
   /// For testing only.
   void SetCharacteristicsForTesting(
       const std::vector<std::string>& characteristics);
@@ -223,6 +226,8 @@ class MediaPlaylist {
     return characteristics_;
   }
 
+  bool forced_subtitle() const { return forced_subtitle_; }
+
   bool is_dvs() const {
     // HLS Authoring Specification for Apple Devices
     // https://developer.apple.com/documentation/http_live_streaming/hls_authoring_specification_for_apple_devices#overview
@@ -262,6 +267,7 @@ class MediaPlaylist {
   std::string codec_;
   std::string language_;
   std::vector<std::string> characteristics_;
+  bool forced_subtitle_ = false;
   uint32_t media_sequence_number_ = 0;
   bool inserted_discontinuity_tag_ = false;
   int discontinuity_sequence_number_ = 0;

packager/hls/base/media_playlist_unittest.cc

@@ -51,6 +51,10 @@ class MediaPlaylistTest : public ::testing::Test {
         default_group_id_("default_group_id") {
     hls_params_.playlist_type = type;
     hls_params_.time_shift_buffer_depth = kTimeShiftBufferDepth;
+
+    // NOTE: hls_params_ is passed by and stored by reference in MediaPlaylist,
+    // so changed made to it through mutable_hls_params() after this point
+    // still affect what the playlist see in its own hls_params_ later.
     media_playlist_.reset(new MediaPlaylist(hls_params_, default_file_name_,
                                             default_name_, default_group_id_));
   }
@@ -658,6 +662,90 @@ TEST_F(MediaPlaylistMultiSegmentTest, MultipleEncryptionInfo) {
   ASSERT_FILE_STREQ(kMemoryFilePath, kExpectedOutput);
 }
 
+TEST_F(MediaPlaylistSingleSegmentTest, StartTimeEmpty) {
+  const std::string kExpectedOutput =
+      "#EXTM3U\n"
+      "#EXT-X-VERSION:6\n"
+      "## Generated with https://github.com/shaka-project/shaka-packager "
+      "version test\n"
+      "#EXT-X-TARGETDURATION:0\n"
+      "#EXT-X-PLAYLIST-TYPE:VOD\n"
+      "#EXT-X-ENDLIST\n";
+
+  // Because this is std::nullopt, the tag isn't in the playlist at all.
+  mutable_hls_params()->start_time_offset = std::nullopt;
+
+  ASSERT_TRUE(media_playlist_->SetMediaInfo(valid_video_media_info_));
+
+  const char kMemoryFilePath[] = "memory://media.m3u8";
+  EXPECT_TRUE(media_playlist_->WriteToFile(kMemoryFilePath));
+
+  ASSERT_FILE_STREQ(kMemoryFilePath, kExpectedOutput);
+}
+
+TEST_F(MediaPlaylistSingleSegmentTest, StartTimeZero) {
+  const std::string kExpectedOutput =
+      "#EXTM3U\n"
+      "#EXT-X-VERSION:6\n"
+      "## Generated with https://github.com/shaka-project/shaka-packager "
+      "version test\n"
+      "#EXT-X-TARGETDURATION:0\n"
+      "#EXT-X-PLAYLIST-TYPE:VOD\n"
+      "#EXT-X-START:TIME-OFFSET=0.000000\n"
+      "#EXT-X-ENDLIST\n";
+
+  mutable_hls_params()->start_time_offset = 0;
+
+  ASSERT_TRUE(media_playlist_->SetMediaInfo(valid_video_media_info_));
+
+  const char kMemoryFilePath[] = "memory://media.m3u8";
+  EXPECT_TRUE(media_playlist_->WriteToFile(kMemoryFilePath));
+
+  ASSERT_FILE_STREQ(kMemoryFilePath, kExpectedOutput);
+}
+
+TEST_F(MediaPlaylistSingleSegmentTest, StartTimePositive) {
+  const std::string kExpectedOutput =
+      "#EXTM3U\n"
+      "#EXT-X-VERSION:6\n"
+      "## Generated with https://github.com/shaka-project/shaka-packager "
+      "version test\n"
+      "#EXT-X-TARGETDURATION:0\n"
+      "#EXT-X-PLAYLIST-TYPE:VOD\n"
+      "#EXT-X-START:TIME-OFFSET=20.000000\n"
+      "#EXT-X-ENDLIST\n";
+
+  mutable_hls_params()->start_time_offset = 20;
+
+  ASSERT_TRUE(media_playlist_->SetMediaInfo(valid_video_media_info_));
+
+  const char kMemoryFilePath[] = "memory://media.m3u8";
+  EXPECT_TRUE(media_playlist_->WriteToFile(kMemoryFilePath));
+
+  ASSERT_FILE_STREQ(kMemoryFilePath, kExpectedOutput);
+}
+
+TEST_F(MediaPlaylistSingleSegmentTest, StartTimeNegative) {
+  const std::string kExpectedOutput =
+      "#EXTM3U\n"
+      "#EXT-X-VERSION:6\n"
+      "## Generated with https://github.com/shaka-project/shaka-packager "
+      "version test\n"
+      "#EXT-X-TARGETDURATION:0\n"
+      "#EXT-X-PLAYLIST-TYPE:VOD\n"
+      "#EXT-X-START:TIME-OFFSET=-3.141590\n"
+      "#EXT-X-ENDLIST\n";
+
+  mutable_hls_params()->start_time_offset = -3.14159;
+
+  ASSERT_TRUE(media_playlist_->SetMediaInfo(valid_video_media_info_));
+
+  const char kMemoryFilePath[] = "memory://media.m3u8";
+  EXPECT_TRUE(media_playlist_->WriteToFile(kMemoryFilePath));
+
+  ASSERT_FILE_STREQ(kMemoryFilePath, kExpectedOutput);
+}
+
 class LiveMediaPlaylistTest : public MediaPlaylistMultiSegmentTest {
  protected:
   LiveMediaPlaylistTest()

packager/media/base/audio_stream_info.cc

@@ -150,6 +150,8 @@ std::string AudioStreamInfo::GetCodecString(Codec codec,
       return "dts-";
     case kCodecDTSP:
       return "dts+";
+    case kCodecDTSX:
+      return "dtsx";
     case kCodecEAC3:
       return "ec-3";
     case kCodecAC4:

packager/media/base/fourccs.h

@@ -58,6 +58,7 @@ enum FourCC : uint32_t {
   FOURCC_dtsl = 0x6474736c,
   FOURCC_dtsm = 0x6474732d,  // "dts-"
   FOURCC_dtsp = 0x6474732b,  // "dts+"
+  FOURCC_dtsx = 0x64747378,  // "dtsx"
   FOURCC_dvcC = 0x64766343,
   FOURCC_dvh1 = 0x64766831,
   FOURCC_dvhe = 0x64766865,
@@ -151,8 +152,9 @@ enum FourCC : uint32_t {
   FOURCC_trex = 0x74726578,
   FOURCC_trun = 0x7472756e,
   FOURCC_udta = 0x75647461,
-  FOURCC_url = 0x75726c20,  // "url "
+  FOURCC_udts = 0x75647473,  // "udts"
-  FOURCC_urn = 0x75726e20,  // "urn "
+  FOURCC_url = 0x75726c20,   // "url "
+  FOURCC_urn = 0x75726e20,   // "urn "
   FOURCC_uuid = 0x75756964,
   FOURCC_vide = 0x76696465,
   FOURCC_vlab = 0x766c6162,

packager/media/base/stream_info.h

@@ -49,6 +49,7 @@ enum Codec {
   kCodecDTSL,
   kCodecDTSM,
   kCodecDTSP,
+  kCodecDTSX,
   kCodecEAC3,
   kCodecFlac,
   kCodecOpus,

packager/media/codecs/CMakeLists.txt

@@ -12,6 +12,7 @@ add_library(media_codecs STATIC
     avc_decoder_configuration_record.cc
     decoder_configuration_record.cc
     dovi_decoder_configuration_record.cc
+    dts_audio_specific_config.cc
     ec3_audio_util.cc
     ac4_audio_util.cc
     es_descriptor.cc

packager/media/codecs/dts_audio_specific_config.cc

@@ -0,0 +1,28 @@
+// Copyright (c) 2023 Google Inc. All rights reserved.
+//
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include <packager/media/codecs/dts_audio_specific_config.h>
+
+#include <packager/media/base/bit_reader.h>
+#include <packager/media/base/rcheck.h>
+
+namespace shaka {
+namespace media {
+
+bool GetDTSXChannelMask(const std::vector<uint8_t>& udts, uint32_t& mask) {
+  // udts is the DTS-UHD Specific Box: ETSI TS 103 491 V1.2.1 Table B-2
+  // DecoderProfileCode(6 bits)
+  // FrameDurationCode(2 bits)
+  // MaxPayloadCode(3 bits)
+  // NumPresentationsCode(5 bits)
+  // ChannelMask (32 bits)
+  BitReader bit_reader(udts.data(), udts.size());
+  RCHECK(bit_reader.SkipBits(16));
+  RCHECK(bit_reader.ReadBits(32, &mask));
+  return true;
+}
+
+}  // namespace media
+}  // namespace shaka

packager/media/codecs/dts_audio_specific_config.h

@@ -0,0 +1,24 @@
+// Copyright (c) 2023 Google Inc. All rights reserved.
+//
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef PACKAGER_MEDIA_CODECS_DTS_AUDIO_SPECIFIC_CONFIG_H_
+#define PACKAGER_MEDIA_CODECS_DTS_AUDIO_SPECIFIC_CONFIG_H_
+
+#include <stddef.h>
+#include <stdint.h>
+
+#include <vector>
+
+namespace shaka {
+namespace media {
+
+class BitReader;
+
+bool GetDTSXChannelMask(const std::vector<uint8_t>& udts, uint32_t& mask);
+
+}  // namespace media
+}  // namespace shaka
+
+#endif  // PACKAGER_MEDIA_CODECS_DTS_AUDIO_SPECIFIC_CONFIG_H_

packager/media/codecs/dts_audio_specific_config_unittest.cc

@@ -0,0 +1,37 @@
+// Copyright 2023 Google Inc. All rights reserved.
+//
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#include <gtest/gtest.h>
+
+#include "packager/media/codecs/dts_audio_specific_config.h"
+
+namespace shaka {
+namespace media {
+
+TEST(DTSAudioSpecificConfigTest, BasicProfileTest) {
+  uint8_t buffer[] = {0x01, 0x20, 0x00, 0x00, 0x0, 0x3F, 0x80, 0x00};
+  std::vector<uint8_t> data(std::begin(buffer), std::end(buffer));
+  uint32_t mask;
+  EXPECT_TRUE(GetDTSXChannelMask(data, mask));
+  EXPECT_EQ(0x3F, mask);
+}
+
+TEST(DTSAudioSpecificConfigTest, ChannelMaskBytes) {
+  uint8_t buffer[] = {0x01, 0x20, 0x12, 0x34, 0x56, 0x78, 0x80, 0x00};
+  std::vector<uint8_t> data(std::begin(buffer), std::end(buffer));
+  uint32_t mask;
+  EXPECT_TRUE(GetDTSXChannelMask(data, mask));
+  EXPECT_EQ(0x12345678, mask);
+}
+
+TEST(DTSAudioSpecificConfigTest, Truncated) {
+  uint8_t buffer[] = {0x01, 0x20, 0x00, 0x00, 0x00};
+  std::vector<uint8_t> data(std::begin(buffer), std::end(buffer));
+  uint32_t mask;
+  EXPECT_FALSE(GetDTSXChannelMask(data, mask));
+}
+
+}  // namespace media
+}  // namespace shaka

packager/media/codecs/h265_byte_to_unit_stream_converter.cc

@@ -74,21 +74,24 @@ bool H265ByteToUnitStreamConverter::GetDecoderConfigurationRecord(
   buffer.AppendInt(static_cast<uint8_t>(kUnitStreamNaluLengthSize - 1));
   buffer.AppendInt(static_cast<uint8_t>(3) /* numOfArrays */);
 
+  // More parameter set NALUs may follow when strip_parameter_set_nalus is
+  // disabled.
+  const uint8_t array_completeness = strip_parameter_set_nalus() ? 0x80 : 0;
+
   // VPS
-  const uint8_t kArrayCompleteness = 0x80;
+  buffer.AppendInt(static_cast<uint8_t>(array_completeness | Nalu::H265_VPS));
-  buffer.AppendInt(static_cast<uint8_t>(kArrayCompleteness | Nalu::H265_VPS));
   buffer.AppendInt(static_cast<uint16_t>(1) /* numNalus */);
   buffer.AppendInt(static_cast<uint16_t>(last_vps_.size()));
   buffer.AppendVector(last_vps_);
 
   // SPS
-  buffer.AppendInt(static_cast<uint8_t>(kArrayCompleteness | Nalu::H265_SPS));
+  buffer.AppendInt(static_cast<uint8_t>(array_completeness | Nalu::H265_SPS));
   buffer.AppendInt(static_cast<uint16_t>(1) /* numNalus */);
   buffer.AppendInt(static_cast<uint16_t>(last_sps_.size()));
   buffer.AppendVector(last_sps_);
 
   // PPS
-  buffer.AppendInt(static_cast<uint8_t>(kArrayCompleteness | Nalu::H265_PPS));
+  buffer.AppendInt(static_cast<uint8_t>(array_completeness | Nalu::H265_PPS));
   buffer.AppendInt(static_cast<uint16_t>(1) /* numNalus */);
   buffer.AppendInt(static_cast<uint16_t>(last_pps_.size()));
   buffer.AppendVector(last_pps_);

packager/media/event/hls_notify_muxer_listener.cc

@@ -26,6 +26,7 @@ HlsNotifyMuxerListener::HlsNotifyMuxerListener(
     const std::string& ext_x_media_name,
     const std::string& ext_x_media_group_id,
     const std::vector<std::string>& characteristics,
+    bool forced_subtitle,
     hls::HlsNotifier* hls_notifier,
     std::optional<uint32_t> index)
     : playlist_name_(playlist_name),
@@ -33,6 +34,7 @@ HlsNotifyMuxerListener::HlsNotifyMuxerListener(
       ext_x_media_name_(ext_x_media_name),
       ext_x_media_group_id_(ext_x_media_group_id),
       characteristics_(characteristics),
+      forced_subtitle_(forced_subtitle),
       hls_notifier_(hls_notifier),
       index_(index) {
   DCHECK(hls_notifier);
@@ -103,6 +105,9 @@ void HlsNotifyMuxerListener::OnMediaStart(const MuxerOptions& muxer_options,
     for (const std::string& characteristic : characteristics_)
       media_info->add_hls_characteristics(characteristic);
   }
+  if (forced_subtitle_) {
+    media_info->set_forced_subtitle(forced_subtitle_);
+  }
   if (index_.has_value())
     media_info->set_index(index_.value());
 

packager/media/event/hls_notify_muxer_listener.h

@@ -39,12 +39,16 @@ class HlsNotifyMuxerListener : public MuxerListener {
   /// @param characteristics is the characteristics for this playlist. This is
   ///        the value of CHARACTERISTICS attribute for EXT-X-MEDIA. This may be
   ///        empty.
+  /// @param forced is the HLS FORCED SUBTITLE setting for this playlist. This
+  ///        is the value of FORCED attribute for EXT-X-MEDIA. This may be
+  ///        empty.
   /// @param hls_notifier used by this listener. Ownership does not transfer.
   HlsNotifyMuxerListener(const std::string& playlist_name,
                          bool iframes_only,
                          const std::string& ext_x_media_name,
                          const std::string& ext_x_media_group_id,
                          const std::vector<std::string>& characteristics,
+                         bool forced,
                          hls::HlsNotifier* hls_notifier,
                          std::optional<uint32_t> index);
   ~HlsNotifyMuxerListener() override;
@@ -86,6 +90,7 @@ class HlsNotifyMuxerListener : public MuxerListener {
   const std::string ext_x_media_name_;
   const std::string ext_x_media_group_id_;
   const std::vector<std::string> characteristics_;
+  const bool forced_subtitle_;
   hls::HlsNotifier* const hls_notifier_;
   std::optional<uint32_t> stream_id_;
   std::optional<uint32_t> index_;

packager/media/event/hls_notify_muxer_listener_unittest.cc

@@ -99,6 +99,7 @@ const char kDefaultName[] = "DEFAULTNAME";
 const char kDefaultGroupId[] = "DEFAULTGROUPID";
 const char kCharactersticA[] = "public.accessibility.transcribes-spoken-dialog";
 const char kCharactersticB[] = "public.easy-to-read";
+const bool kForced = false;
 
 MATCHER_P(HasEncryptionScheme, expected_scheme, "") {
   *result_listener << "it has_protected_content: "
@@ -121,6 +122,7 @@ class HlsNotifyMuxerListenerTest : public ::testing::Test {
                   kDefaultName,
                   kDefaultGroupId,
                   std::vector<std::string>{kCharactersticA, kCharactersticB},
+                  kForced,
                   &mock_notifier_,
                   0) {}
 
@@ -459,6 +461,7 @@ class HlsNotifyMuxerListenerKeyFrameTest : public TestWithParam<bool> {
                   kDefaultName,
                   kDefaultGroupId,
                   std::vector<std::string>(),  // no characteristics.
+                  kForced,
                   &mock_notifier_,
                   0) {}
 

packager/media/event/muxer_listener_factory.cc

@@ -62,6 +62,7 @@ std::list<std::unique_ptr<MuxerListener>> CreateHlsListenersInternal(
   const std::string& group_id = stream.hls_group_id;
   const std::string& iframe_playlist_name = stream.hls_iframe_playlist_name;
   const std::vector<std::string>& characteristics = stream.hls_characteristics;
+  const bool forced_subtitle = stream.forced_subtitle;
 
   if (name.empty()) {
     name = absl::StrFormat("stream_%d", stream_index);
@@ -73,13 +74,13 @@ std::list<std::unique_ptr<MuxerListener>> CreateHlsListenersInternal(
 
   const bool kIFramesOnly = true;
   std::list<std::unique_ptr<MuxerListener>> listeners;
-  listeners.emplace_back(
+  listeners.emplace_back(new HlsNotifyMuxerListener(
-      new HlsNotifyMuxerListener(playlist_name, !kIFramesOnly, name, group_id,
+      playlist_name, !kIFramesOnly, name, group_id, characteristics,
-                                 characteristics, notifier, stream.index));
+      forced_subtitle, notifier, stream.index));
   if (!iframe_playlist_name.empty()) {
     listeners.emplace_back(new HlsNotifyMuxerListener(
         iframe_playlist_name, kIFramesOnly, name, group_id,
-        std::vector<std::string>(), notifier, stream.index));
+        std::vector<std::string>(), forced_subtitle, notifier, stream.index));
   }
   return listeners;
 }

packager/media/event/muxer_listener_factory.h

@@ -47,6 +47,7 @@ class MuxerListenerFactory {
     std::string hls_playlist_name;
     std::string hls_iframe_playlist_name;
     std::vector<std::string> hls_characteristics;
+    bool forced_subtitle = false;
     bool hls_only = false;
 
     // DASH specific values needed to write DASH mpd. Will only be used if an

packager/media/event/muxer_listener_internal.cc

@@ -20,6 +20,7 @@
 #include <packager/media/base/text_stream_info.h>
 #include <packager/media/base/video_stream_info.h>
 #include <packager/media/codecs/ac4_audio_util.h>
+#include <packager/media/codecs/dts_audio_specific_config.h>
 #include <packager/media/codecs/ec3_audio_util.h>
 #include <packager/mpd/base/media_info.pb.h>
 #include <packager/utils/bytes_to_string_view.h>
@@ -165,6 +166,16 @@ void AddAudioInfo(const AudioStreamInfo* audio_stream_info,
     codec_data->set_ac4_ims_flag(ac4_ims_flag);
     codec_data->set_ac4_cbi_flag(ac4_cbi_flag);
   }
+
+  if (audio_stream_info->codec() == kCodecDTSX) {
+    auto* codec_data = audio_info->mutable_codec_specific_data();
+    uint32_t channel_mask;
+    if (!GetDTSXChannelMask(codec_config, channel_mask)) {
+      LOG(ERROR) << "Failed to parse DTSX channel mask.";
+      return;
+    }
+    codec_data->set_channel_mask(channel_mask);
+  }
 }
 
 void AddTextInfo(const TextStreamInfo& text_stream_info,

packager/media/event/muxer_listener_internal_unittest.cc

@@ -72,6 +72,24 @@ TEST_F(MuxerListenerInternalVideoStreamTest, TransferCharacteristics) {
   EXPECT_EQ(18u, media_info.video_info().transfer_characteristics());
 }
 
+class MuxerListenerInternalAudioStreamTest : public MuxerListenerInternalTest {
+};
+
+// AddAudioInfo function should parse the channel mask
+TEST_F(MuxerListenerInternalAudioStreamTest, DTSX) {
+  MediaInfo media_info;
+  std::shared_ptr<AudioStreamInfo> audio_info = CreateAudioStreamInfo(
+      GetAudioStreamInfoParams(kCodecDTSX, "dtsx",
+                               {0x01, 0x20, 0x00, 0x00, 0x0, 0x3F, 0x80,
+                                0x00}));  // Channel mask = 3F
+  ASSERT_TRUE(GenerateMediaInfo(MuxerOptions(), *audio_info,
+                                kReferenceTimeScale,
+                                MuxerListener::kContainerMp4, &media_info));
+  MediaInfo_AudioInfo* info = media_info.mutable_audio_info();
+  auto* codec_data = info->mutable_codec_specific_data();
+  EXPECT_EQ(0x3F, codec_data->channel_mask());
+}
+
 }  // namespace internal
 }  // namespace media
 }  // namespace shaka

packager/media/event/muxer_listener_test_helper.cc

@@ -104,5 +104,53 @@ std::vector<ProtectionSystemSpecificInfo> GetDefaultKeySystemInfo() {
             std::end(kExpectedDefaultPsshBox) - 1}}};
 }
 
+AudioStreamInfoParameters::AudioStreamInfoParameters() {}
+AudioStreamInfoParameters::~AudioStreamInfoParameters() {}
+
+std::shared_ptr<AudioStreamInfo> CreateAudioStreamInfo(
+    const AudioStreamInfoParameters& param) {
+  return std::make_shared<AudioStreamInfo>(
+      param.track_id, param.time_scale, param.duration, param.codec,
+      param.codec_string, param.codec_config.data(), param.codec_config.size(),
+      param.sample_bits, param.num_channels, param.sampling_frequency,
+      param.seek_preroll_ns, param.codec_delay_ns, param.max_bitrate,
+      param.avg_bitrate, param.language, param.is_encrypted);
+}
+
+AudioStreamInfoParameters GetAudioStreamInfoParams(
+    Codec codec,
+    const char* codec_string,
+    const std::vector<uint8_t>& codec_config) {
+  const int kTrackId = 0;
+  const int32_t kTimeScale = 10;
+  const int64_t kAudioStreamDuration = 200;
+  const char* kLanuageUndefined = "und";
+  const uint8_t kSampleBits = 16;
+  const uint8_t kNumChannels = 6;
+  const uint32_t kSamplingFrequency = 48000;
+  const uint64_t kSeekPrerollNs = 0;
+  const uint64_t kCodecDelayNs = 0;
+  const uint32_t kMaxBitrate = 0;
+  const uint32_t kAvgBitrate = 0;
+  const bool kEncryptedFlag = false;
+  AudioStreamInfoParameters params;
+  params.track_id = kTrackId;
+  params.time_scale = kTimeScale;
+  params.duration = kAudioStreamDuration;
+  params.codec = codec;
+  params.codec_string = codec_string;
+  params.language = kLanuageUndefined;
+  params.sample_bits = kSampleBits;
+  params.num_channels = kNumChannels;
+  params.sampling_frequency = kSamplingFrequency;
+  params.seek_preroll_ns = kSeekPrerollNs;
+  params.codec_delay_ns = kCodecDelayNs;
+  params.max_bitrate = kMaxBitrate;
+  params.avg_bitrate = kAvgBitrate;
+  params.codec_config = codec_config;
+  params.is_encrypted = kEncryptedFlag;
+  return params;
+}
+
 }  // namespace media
 }  // namespace shaka

packager/media/event/muxer_listener_test_helper.h

@@ -10,6 +10,7 @@
 #include <cstdint>
 #include <vector>
 
+#include <packager/media/base/audio_stream_info.h>
 #include <packager/media/base/key_source.h>
 #include <packager/media/base/muxer_options.h>
 #include <packager/media/base/stream_info.h>
@@ -95,6 +96,29 @@ struct VideoStreamInfoParameters {
   bool is_encrypted;
 };
 
+// Struct that gets passed for to CreateAudioStreamInfo() to create a
+// StreamInfo instance. Useful for generating multiple AudioStreamInfo with
+// slightly different parameters.
+struct AudioStreamInfoParameters {
+  AudioStreamInfoParameters();
+  ~AudioStreamInfoParameters();
+  int track_id;
+  int32_t time_scale;
+  int64_t duration;
+  Codec codec;
+  std::string codec_string;
+  std::vector<uint8_t> codec_config;
+  uint8_t sample_bits;
+  uint8_t num_channels;
+  uint32_t sampling_frequency;
+  uint64_t seek_preroll_ns;
+  uint64_t codec_delay_ns;
+  uint32_t max_bitrate;
+  uint32_t avg_bitrate;
+  std::string language;
+  bool is_encrypted;
+};
+
 struct OnNewSegmentParameters {
   std::string file_name;
   int64_t start_time;
@@ -115,6 +139,16 @@ std::shared_ptr<VideoStreamInfo> CreateVideoStreamInfo(
 // Returns the "default" VideoStreamInfoParameters for testing.
 VideoStreamInfoParameters GetDefaultVideoStreamInfoParams();
 
+// Creates StreamInfo instance from AudioStreamInfoParameters.
+std::shared_ptr<AudioStreamInfo> CreateAudioStreamInfo(
+    const AudioStreamInfoParameters& param);
+
+// Returns the "default" configuration for testing given codec and parameters.
+AudioStreamInfoParameters GetAudioStreamInfoParams(
+    Codec codec,
+    const char* codec_string,
+    const std::vector<uint8_t>& codec_config);
+
 // Returns the "default" values for OnMediaEnd().
 OnMediaEndParameters GetDefaultOnMediaEndParams();
 

packager/media/formats/mp4/box_definitions.cc

@@ -1811,6 +1811,27 @@ size_t DTSSpecific::ComputeSizeInternal() {
          sizeof(kDdtsExtraData);
 }
 
+UDTSSpecific::UDTSSpecific() = default;
+UDTSSpecific::~UDTSSpecific() = default;
+
+FourCC UDTSSpecific::BoxType() const {
+  return FOURCC_udts;
+}
+
+bool UDTSSpecific::ReadWriteInternal(BoxBuffer* buffer) {
+  RCHECK(ReadWriteHeaderInternal(buffer) &&
+         buffer->ReadWriteVector(
+             &data, buffer->Reading() ? buffer->BytesLeft() : data.size()));
+  return true;
+}
+
+size_t UDTSSpecific::ComputeSizeInternal() {
+  // This box is optional. Skip it if not initialized.
+  if (data.empty())
+    return 0;
+  return HeaderSize() + data.size();
+}
+
 AC3Specific::AC3Specific() = default;
 AC3Specific::~AC3Specific() = default;
 
@@ -1983,6 +2004,7 @@ bool AudioSampleEntry::ReadWriteInternal(BoxBuffer* buffer) {
 
   RCHECK(buffer->TryReadWriteChild(&esds));
   RCHECK(buffer->TryReadWriteChild(&ddts));
+  RCHECK(buffer->TryReadWriteChild(&udts));
   RCHECK(buffer->TryReadWriteChild(&dac3));
   RCHECK(buffer->TryReadWriteChild(&dec3));
   RCHECK(buffer->TryReadWriteChild(&dac4));
@@ -2014,7 +2036,7 @@ size_t AudioSampleEntry::ComputeSizeInternal() {
          sizeof(samplesize) + sizeof(samplerate) + sinf.ComputeSize() +
          esds.ComputeSize() + ddts.ComputeSize() + dac3.ComputeSize() +
          dec3.ComputeSize() + dops.ComputeSize() + dfla.ComputeSize() +
-         dac4.ComputeSize() + mhac.ComputeSize() +
+         dac4.ComputeSize() + mhac.ComputeSize() + udts.ComputeSize() +
          // Reserved and predefined bytes.
          6 + 8 +  // 6 + 8 bytes reserved.
          4;       // 4 bytes predefined.

packager/media/formats/mp4/box_definitions.h

@@ -334,6 +334,12 @@ struct DTSSpecific : Box {
   std::vector<uint8_t> extra_data;
 };
 
+struct UDTSSpecific : Box {
+  DECLARE_BOX_METHODS(UDTSSpecific);
+
+  std::vector<uint8_t> data;
+};
+
 struct AC3Specific : Box {
   DECLARE_BOX_METHODS(AC3Specific);
 
@@ -396,6 +402,7 @@ struct AudioSampleEntry : Box {
 
   ElementaryStreamDescriptor esds;
   DTSSpecific ddts;
+  UDTSSpecific udts;
   AC3Specific dac3;
   EC3Specific dec3;
   AC4Specific dac4;

packager/media/formats/mp4/box_definitions_unittest.cc

@@ -417,6 +417,16 @@ class BoxDefinitionsTestGeneral : public testing::Test {
     ddts->pcm_sample_depth = 24;
   }
 
+  void Fill(UDTSSpecific* udts) {
+    const uint8_t kUdtsData[] = {0x01, 0x20, 0x00, 0x00, 0x0, 0x3F, 0x80, 0x00};
+    udts->data.assign(kUdtsData, kUdtsData + std::size(kUdtsData));
+  }
+
+  void Modify(UDTSSpecific* udts) {
+    const uint8_t kUdtsData[] = {0x01, 0x20, 0x01, 0x80, 0xA, 0x3F, 0x80, 0x00};
+    udts->data.assign(kUdtsData, kUdtsData + std::size(kUdtsData));
+  }
+
   void Fill(AC3Specific* dac3) {
     const uint8_t kAc3Data[] = {0x50, 0x11, 0x60};
     dac3->data.assign(kAc3Data, kAc3Data + std::size(kAc3Data));
@@ -1217,6 +1227,20 @@ TEST_F(BoxDefinitionsTest, DTSSampleEntry) {
   ASSERT_EQ(entry, entry_readback);
 }
 
+TEST_F(BoxDefinitionsTest, UDTSSampleEntry) {
+  AudioSampleEntry entry;
+  entry.format = FOURCC_dtsx;
+  entry.data_reference_index = 2;
+  entry.channelcount = 6;
+  entry.samplesize = 16;
+  entry.samplerate = 48000;
+  Fill(&entry.udts);
+  entry.Write(this->buffer_.get());
+  AudioSampleEntry entry_readback;
+  ASSERT_TRUE(ReadBack(&entry_readback));
+  ASSERT_EQ(entry, entry_readback);
+}
+
 TEST_F(BoxDefinitionsTest, AC3SampleEntry) {
   AudioSampleEntry entry;
   entry.format = FOURCC_ac_3;

packager/media/formats/mp4/mp4_media_parser.cc

@@ -90,6 +90,8 @@ Codec FourCCToCodec(FourCC fourcc) {
       return kCodecDTSL;
     case FOURCC_dtse:
       return kCodecDTSE;
+    case FOURCC_dtsx:
+      return kCodecDTSX;
     case FOURCC_dtsp:
       return kCodecDTSP;
     case FOURCC_dtsm:
@@ -491,6 +493,9 @@ bool MP4MediaParser::ParseMoov(BoxReader* reader) {
           max_bitrate = entry.ddts.max_bitrate;
           avg_bitrate = entry.ddts.avg_bitrate;
           break;
+        case FOURCC_dtsx:
+          codec_config = entry.udts.data;
+          break;
         case FOURCC_ac_3:
           codec_config = entry.dac3.data;
           num_channels = static_cast<uint8_t>(GetAc3NumChannels(codec_config));

packager/media/formats/mp4/mp4_muxer.cc

@@ -86,6 +86,8 @@ FourCC CodecToFourCC(Codec codec, H26xStreamFormat h26x_stream_format) {
       return FOURCC_dtse;
     case kCodecDTSM:
       return FOURCC_dtsm;
+    case kCodecDTSX:
+      return FOURCC_dtsx;
     case kCodecEAC3:
       return FOURCC_ec_3;
     case kCodecAC4:
@@ -498,6 +500,9 @@ bool MP4Muxer::GenerateAudioTrak(const AudioStreamInfo* audio_info,
       audio.ddts.sampling_frequency = audio_info->sampling_frequency();
       audio.ddts.pcm_sample_depth = audio_info->sample_bits();
       break;
+    case kCodecDTSX:
+      audio.udts.data = audio_info->codec_config();
+      break;
     case kCodecAC3:
       audio.dac3.data = audio_info->codec_config();
       break;

packager/media/test/data/README

@@ -118,3 +118,11 @@ sintel-1024x436.mp4 - First 6 seconds of Sintel stream.
 // http://download.opencontent.netflix.com/?prefix=TechblogAssets/Sparks/Sparks_DolbyVision_P3D65_PQ_5994fps_4096x2160_LtRt_IMF_20170214/
 sparks_dovi_5.mp4  - Dolby Vision profile 5
 sparks_dovi_8.mp4  - Dolby Vision profile 8
+
+// DTS:X Profile 2 5.1 channel audio
+bear-dtsx.mp4:
+  Generated using ffmpeg and proprietary encoder:
+    ffmpeg -i bear.aiff -i bear.aiff -i bear.aiff \
+        -filter_complex "[0:a][1:a][2:a]concat=n=3:v=0:a=1[out]" \
+        -ar 48000 -map "[out]" -y bear.wav
+    dtsx-ott-enc -w -o bear-dtsx.mp4 -c 5.1 bear.wav bear.wav bear.wav

packager/mpd/base/adaptation_set.cc

@@ -59,6 +59,8 @@ std::string RoleToText(AdaptationSet::Role role) {
       return "commentary";
     case AdaptationSet::kRoleDub:
       return "dub";
+    case AdaptationSet::kRoleForcedSubtitle:
+      return "forced-subtitle";
     case AdaptationSet::kRoleDescription:
       return "description";
     default:

packager/mpd/base/adaptation_set.h

@@ -43,6 +43,7 @@ class AdaptationSet {
     kRoleSupplementary,
     kRoleCommentary,
     kRoleDub,
+    kRoleForcedSubtitle,
     kRoleDescription
   };
 

packager/mpd/base/media_info.proto

@@ -213,6 +213,11 @@ message MediaInfo {
   // Equal to the target segment duration times the reference time scale.
   optional uint64 segment_duration = 25;
 
+  // Marks stream as a Forced Narrative subtitle stream, indicated using
+  // forced-subtitle role in DASH
+  // and FORCED=YES in HLS
+  optional bool forced_subtitle = 26 [default = false];
+
   // stream index for consistent ordering of streams
   optional uint32 index = 28;
 

packager/mpd/base/period.cc

@@ -59,6 +59,8 @@ AdaptationSet::Role RoleFromString(const std::string& role_str) {
     return AdaptationSet::Role::kRoleCommentary;
   if (role_str == "dub")
     return AdaptationSet::Role::kRoleDub;
+  if (role_str == "forced-subtitle")
+    return AdaptationSet::Role::kRoleForcedSubtitle;
   if (role_str == "description")
     return AdaptationSet::Role::kRoleDescription;
   return AdaptationSet::Role::kRoleUnknown;

packager/mpd/base/xml/xml_node.cc

@@ -49,6 +49,9 @@ typedef MediaInfo::VideoInfo VideoInfo;
 namespace {
 const char kEC3Codec[] = "ec-3";
 const char kAC4Codec[] = "ac-4";
+const char kDTSCCodec[] = "dtsc";
+const char kDTSECodec[] = "dtse";
+const char kDTSXCodec[] = "dtsx";
 
 std::string RangeToString(const Range& range) {
   return absl::StrFormat("%u-%u", range.begin(), range.end());
@@ -680,6 +683,18 @@ bool RepresentationXmlNode::AddAudioChannelInfo(const AudioInfo& audio_info) {
                            "1");
     }
     return ret;
+  } else if (audio_info.codec() == kDTSCCodec ||
+             audio_info.codec() == kDTSECodec) {
+    audio_channel_config_value =
+        absl::StrFormat("%u", audio_info.num_channels());
+    audio_channel_config_scheme =
+        "tag:dts.com,2014:dash:audio_channel_configuration:2012";
+  } else if (audio_info.codec() == kDTSXCodec) {
+    const auto& codec_data = audio_info.codec_specific_data();
+    audio_channel_config_value =
+        absl::StrFormat("%08X", codec_data.channel_mask());
+    audio_channel_config_scheme =
+        "tag:dts.com,2018:uhd:audio_channel_configuration";
   } else {
     audio_channel_config_value =
         absl::StrFormat("%u", audio_info.num_channels());

packager/mpd/base/xml/xml_node_unittest.cc

@@ -766,5 +766,61 @@ TEST_F(LowLatencySegmentTest, LowLatencySegmentTemplate) {
                    "</Representation>"));
 }
 
+TEST(XmlNodeTest, AddDTSCAudioInfo) {
+  MediaInfo::AudioInfo audio_info;
+  audio_info.set_codec("dtsc");
+  audio_info.set_sampling_frequency(48000);
+  audio_info.set_num_channels(6);
+
+  RepresentationXmlNode representation;
+  ASSERT_TRUE(representation.AddAudioInfo(audio_info));
+  EXPECT_THAT(
+      representation,
+      XmlNodeEqual(
+          "<Representation audioSamplingRate=\"48000\">\n"
+          "  <AudioChannelConfiguration\n"
+          "   schemeIdUri=\n"
+          "    \"tag:dts.com,2014:dash:audio_channel_configuration:2012\"\n"
+          "   value=\"6\"/>\n"
+          "</Representation>\n"));
+}
+
+TEST(XmlNodeTest, AddDTSEAudioInfo) {
+  MediaInfo::AudioInfo audio_info;
+  audio_info.set_codec("dtse");
+  audio_info.set_sampling_frequency(48000);
+  audio_info.set_num_channels(6);
+
+  RepresentationXmlNode representation;
+  ASSERT_TRUE(representation.AddAudioInfo(audio_info));
+  EXPECT_THAT(
+      representation,
+      XmlNodeEqual(
+          "<Representation audioSamplingRate=\"48000\">\n"
+          "  <AudioChannelConfiguration\n"
+          "   schemeIdUri=\n"
+          "    \"tag:dts.com,2014:dash:audio_channel_configuration:2012\"\n"
+          "   value=\"6\"/>\n"
+          "</Representation>\n"));
+}
+
+TEST(XmlNodeTest, AddDTSXAudioInfo) {
+  MediaInfo::AudioInfo audio_info;
+  audio_info.set_codec("dtsx");
+  audio_info.set_sampling_frequency(48000);
+  audio_info.mutable_codec_specific_data()->set_channel_mask(0x3F);
+
+  RepresentationXmlNode representation;
+  ASSERT_TRUE(representation.AddAudioInfo(audio_info));
+  EXPECT_THAT(
+      representation,
+      XmlNodeEqual("<Representation audioSamplingRate=\"48000\">\n"
+                   "  <AudioChannelConfiguration\n"
+                   "   schemeIdUri=\n"
+                   "    \"tag:dts.com,2018:uhd:audio_channel_configuration\"\n"
+                   "   value=\"0000003F\"/>\n"
+                   "</Representation>\n"));
+}
+
 }  // namespace xml
 }  // namespace shaka

packager/packager.cc

@@ -69,6 +69,7 @@ MuxerListenerFactory::StreamData ToMuxerListenerData(
   data.hls_playlist_name = stream.hls_playlist_name;
   data.hls_iframe_playlist_name = stream.hls_iframe_playlist_name;
   data.hls_characteristics = stream.hls_characteristics;
+  data.forced_subtitle = stream.forced_subtitle;
   data.hls_only = stream.hls_only;
 
   data.dash_accessiblities = stream.dash_accessiblities;