Improving WebM video duration estimation.

media/formats/webm/webm_cluster_parser.h

 // Copyright 2014 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef MEDIA_FORMATS_WEBM_WEBM_CLUSTER_PARSER_H_
 #define MEDIA_FORMATS_WEBM_WEBM_CLUSTER_PARSER_H_
 
 #include <deque>
 #include <map>
 #include <set>
 #include <string>
 
 #include "base/memory/scoped_ptr.h"
 #include "media/base/audio_decoder_config.h"
 #include "media/base/media_export.h"
 #include "media/base/media_log.h"
 #include "media/base/stream_parser.h"
 #include "media/base/stream_parser_buffer.h"
 #include "media/formats/webm/webm_parser.h"
 #include "media/formats/webm/webm_tracks_parser.h"
 
 namespace media {
 
 class MEDIA_EXPORT WebMClusterParser : public WebMParserClient {
  public:
   typedef StreamParser::TrackId TrackId;
   typedef std::deque<scoped_refptr<StreamParserBuffer> > BufferQueue;
   typedef std::map<TrackId, const BufferQueue> TextBufferQueueMap;
 
-  // Arbitrarily-chosen numbers to estimate the duration of a buffer if none is
-  // set and there is not enough information to get a better estimate.
+  // Numbers chosen to estimate the duration of a buffer if none is set and
+  // there is not enough information to get a better estimate.
   enum {
-    kDefaultAudioBufferDurationInMs = 23,  // Common 1k samples @44.1kHz
-    kDefaultVideoBufferDurationInMs = 42  // Low 24fps to reduce stalls
+    // Common 1k samples @44.1kHz
+    kDefaultAudioBufferDurationInMs = 23,
+    // Chosen to represent 16fps duration, which will prevent MSE stalls in

[wolenetz, 2015/04/15 02:55:23]

nit: insert blank line to break the two comment+item entries apart visually?

[chcunningham, 2015/04/16 18:04:16]

Done.

+    // videos with frame-rates as low as 8fps.
+    kDefaultVideoBufferDurationInMs = 63
   };
 
   // Opus packets encode the duration and other parameters in the 5 most
   // significant bits of the first byte. The index in this array corresponds
   // to the duration of each frame of the packet in microseconds. See
   // https://tools.ietf.org/html/rfc6716#page-14
   static const uint16_t kOpusFrameDurationsMu[];
 
  private:
   // Helper class that manages per-track state.
@@ skipping 20 matching lines @@
     const BufferQueue& ready_buffers() const { return ready_buffers_; }
 
     // If |last_added_buffer_missing_duration_| is set, updates its duration
     // relative to |buffer|'s timestamp, and adds it to |buffers_| and unsets
     // |last_added_buffer_missing_duration_|. Then, if |buffer| is missing
     // duration, saves |buffer| into |last_added_buffer_missing_duration_|, or
     // otherwise adds |buffer| to |buffers_|.
     bool AddBuffer(const scoped_refptr<StreamParserBuffer>& buffer);
 
     // If |last_added_buffer_missing_duration_| is set, updates its duration to
-    // be non-kNoTimestamp() value of |estimated_next_frame_duration_| or an
-    // arbitrary default, then adds it to |buffers_| and unsets
+    // be non-kNoTimestamp() value of |estimated_next_frame_duration_| or a
+    // hard-coded default, then adds it to |buffers_| and unsets
     // |last_added_buffer_missing_duration_|. (This method helps stream parser
     // emit all buffers in a media segment before signaling end of segment.)
     void ApplyDurationEstimateIfNeeded();
 
     // Clears |ready_buffers_| (use ExtractReadyBuffers() to fill it again).
     // Leaves as-is |buffers_| and any possibly held-aside buffer that is
     // missing duration.
     void ClearReadyBuffers();
 
     // Clears all buffer state, including any possibly held-aside buffer that
@@ skipping 13 matching lines @@
     // Helper that sanity-checks |buffer| duration, updates
     // |estimated_next_frame_duration_|, and adds |buffer| to |buffers_|.
     // Returns false if |buffer| failed sanity check and therefore was not added
     // to |buffers_|. Returns true otherwise.
     bool QueueBuffer(const scoped_refptr<StreamParserBuffer>& buffer);
 
     // Helper that calculates the buffer duration to use in
     // ApplyDurationEstimateIfNeeded().
     base::TimeDelta GetDurationEstimate();
 
+    // Counts the number of estimated durations used in this track. Used to
+    // prevent log spam for MEDIA_LOG()s about estimated duration.
+    int num_duration_estimates_;
+
     int track_num_;
     bool is_video_;
 
     // Parsed track buffers, each with duration and in (decode) timestamp order,
     // that have not yet been extracted into |ready_buffers_|. Note that up to
     // one additional buffer missing duration may be tracked by
     // |last_added_buffer_missing_duration_|.
     BufferQueue buffers_;
     scoped_refptr<StreamParserBuffer> last_added_buffer_missing_duration_;
 
     // Buffers in (decode) timestamp order that were previously parsed into and
     // extracted from |buffers_|. Buffers are moved from |buffers_| to
     // |ready_buffers_| by ExtractReadyBuffers() if they are below a specified
     // upper bound timestamp. Track users can therefore extract only those
     // parsed buffers which are "ready" for emission (all before some maximum
     // timestamp).
     BufferQueue ready_buffers_;
 
     // If kNoTimestamp(), then |estimated_next_frame_duration_| will be used.
     base::TimeDelta default_duration_;
 
     // If kNoTimestamp(), then a default value will be used. This estimate is
-    // the maximum duration seen or derived so far for this track, and is valid
-    // only if |default_duration_| is kNoTimestamp().
+    // the maximum (for video), or minimum (for audio) duration seen so far for
+    // this track, and is used only if |default_duration_| is kNoTimestamp().
+    // TODO(chcunningham): Use maximum for audio too, adding checks to disable
+    // splicing when these estimates are observed in SourceBufferStream.
     base::TimeDelta estimated_next_frame_duration_;
 
     LogCB log_cb_;
   };
 
   typedef std::map<int, Track> TextTrackMap;
 
  public:
   WebMClusterParser(int64 timecode_scale,
                     int audio_track_num,
@@ skipping 162 matching lines @@
   DecodeTimestamp ready_buffer_upper_bound_;
 
   LogCB log_cb_;
 
   DISALLOW_IMPLICIT_CONSTRUCTORS(WebMClusterParser);
 };
 
 }  // namespace media
 
 #endif  // MEDIA_FORMATS_WEBM_WEBM_CLUSTER_PARSER_H_