feat: default text zero bias (#1330)

A positive value, in milliseconds. It is the threshold used to determine
if we should assume that the text stream actually starts at time zero.
If the first sample comes before default_text_zero_bias_ms, then the
start will be padded as the stream is assumed to start at zero. If the
first sample comes after default_text_zero_bias_ms then the start of the
stream will not be padded as we cannot assume the start time of the
stream.

include/packager/packager.h

@@ -46,6 +46,10 @@ struct PackagingParams {
   /// audio) timestamps to compensate for possible negative timestamps in the
   /// input.
   int32_t transport_stream_timestamp_offset_ms = 0;
+  // the threshold used to determine if we should assume that the text stream
+  // actually starts at time zero
+  int32_t default_text_zero_bias_ms = 0;
+
   /// Chunking (segmentation) related parameters.
   ChunkingParams chunking_params;
 

packager/app/muxer_flags.cc

@@ -62,3 +62,14 @@ ABSL_FLAG(int32_t,
           "input. For example, timestamps from ISO-BMFF after adjusted by "
           "EditList could be negative. In transport streams, timestamps are "
           "not allowed to be less than zero.");
+ABSL_FLAG(
+    int32_t,
+    default_text_zero_bias_ms,
+    0,
+    "A positive value, in milliseconds. It is the threshold used to "
+    "determine if we should assume that the text stream actually starts "
+    "at time zero. If the first sample comes before default_text_zero_bias_ms, "
+    "then the start will be padded as the stream is assumed to start at zero. "
+    "If the first sample comes after default_text_zero_bias_ms then the start "
+    "of the stream will not be padded as we cannot assume the start time of "
+    "the stream.");

packager/app/muxer_flags.h

@@ -21,5 +21,6 @@ ABSL_DECLARE_FLAG(bool, generate_sidx_in_media_segments);
 ABSL_DECLARE_FLAG(std::string, temp_dir);
 ABSL_DECLARE_FLAG(bool, mp4_include_pssh_in_stream);
 ABSL_DECLARE_FLAG(int32_t, transport_stream_timestamp_offset_ms);
+ABSL_DECLARE_FLAG(int32_t, default_text_zero_bias_ms);
 
 #endif  // APP_MUXER_FLAGS_H_

packager/app/packager_main.cc

@@ -461,6 +461,8 @@ std::optional<PackagingParams> GetPackagingParams() {
 
   packaging_params.transport_stream_timestamp_offset_ms =
       absl::GetFlag(FLAGS_transport_stream_timestamp_offset_ms);
+  packaging_params.default_text_zero_bias_ms =
+      absl::GetFlag(FLAGS_default_text_zero_bias_ms);
 
   packaging_params.output_media_info = absl::GetFlag(FLAGS_output_media_info);
 

packager/media/formats/webvtt/text_padder.cc

@@ -40,7 +40,9 @@ Status TextPadder::OnTextSample(std::unique_ptr<StreamData> data) {
   // start at time zero.
   if (max_end_time_ms_ < 0) {
     max_end_time_ms_ =
-        sample.start_time() > zero_start_bias_ms_ ? sample.start_time() : 0;
+        zero_start_bias_ms_ && sample.start_time() > zero_start_bias_ms_
+            ? sample.start_time()
+            : 0;
   }
 
   // Check if there will be a gap between samples if we just dispatch this

packager/packager.cc

@@ -59,8 +59,6 @@ namespace {
 
 const char kMediaInfoSuffix[] = ".media_info";
 
-const int64_t kDefaultTextZeroBiasMs = 10 * 60 * 1000;  // 10 minutes
-
 MuxerListenerFactory::StreamData ToMuxerListenerData(
     const StreamDescriptor& stream) {
   MuxerListenerFactory::StreamData data;
@@ -662,8 +660,8 @@ Status CreateAudioVideoJobs(
 
       std::vector<std::shared_ptr<MediaHandler>> handlers;
       if (is_text) {
-        handlers.emplace_back(
+        handlers.emplace_back(std::make_shared<TextPadder>(
-            std::make_shared<TextPadder>(kDefaultTextZeroBiasMs));
+            packaging_params.default_text_zero_bias_ms));
       }
       if (sync_points) {
         handlers.emplace_back(cue_aligner);