using std::list;
using boost::shared_ptr;
-Matcher::Matcher (Log* log, int sample_rate, float frames_per_second)
+Matcher::Matcher (shared_ptr<Log> log, int sample_rate, float frames_per_second)
: Processor (log)
, _sample_rate (sample_rate)
, _frames_per_second (frames_per_second)
, _video_frames (0)
, _audio_frames (0)
+ , _had_first_video (false)
+ , _had_first_audio (false)
{
}
void
-Matcher::process_video (boost::shared_ptr<Image> image, bool same, boost::shared_ptr<Subtitle> sub, double t)
+Matcher::process_video (shared_ptr<const Image> image, bool same, boost::shared_ptr<Subtitle> sub, double t)
{
_pixel_format = image->pixel_format ();
_size = image->size ();
- _log->log(String::compose("Matcher video @ %1 (same=%2)", t, same));
+ _log->log(String::compose("Matcher video @ %1 [audio=%2, video=%3, pending_audio=%4]", t, _audio_frames, _video_frames, _pending_audio.size()));
if (!_first_input) {
_first_input = t;
}
- if (!_pending_audio.empty() && _video_frames == 0) {
+ bool const this_is_first_video = !_had_first_video;
+ _had_first_video = true;
+
+ if (this_is_first_video && _had_first_audio) {
/* First video since we got audio */
fix_start (t);
}
-
+
/* Video before audio is fine, since we can make up an arbitrary difference
with audio samples (contrasting with video which is quantised to frames)
*/
}
void
-Matcher::process_audio (boost::shared_ptr<AudioBuffers> b, double t)
+Matcher::process_audio (shared_ptr<const AudioBuffers> b, double t)
{
_channels = b->channels ();
-
+
+ _log->log (String::compose (
+ "Matcher audio (%1 frames) @ %2 [video=%3, audio=%4, pending_audio=%5]",
+ b->frames(), t, _video_frames, _audio_frames, _pending_audio.size()
+ )
+ );
+
if (!_first_input) {
_first_input = t;
}
+
+ bool const this_is_first_audio = _had_first_audio;
+ _had_first_audio = true;
- if (_video_frames == 0) {
+ if (!_had_first_video) {
/* No video yet; we must postpone these data until we have some */
_pending_audio.push_back (AudioRecord (b, t));
- } else if (_video_frames > 0 && _audio_frames == 0 && _pending_audio.empty()) {
+ } else if (this_is_first_audio && !_had_first_video) {
/* First audio since we got video */
_pending_audio.push_back (AudioRecord (b, t));
fix_start (_first_input.get ());
void
Matcher::match (double extra_video_needed)
{
+ _log->log (String::compose ("Match %1", extra_video_needed));
+
if (extra_video_needed > 0) {
/* Emit black video frames */
/* Emit silence */
int64_t to_do = -extra_video_needed * _sample_rate;
- _log->log (String::compose (N_("Emitted %1 frames of silence"), to_do));
+ _log->log (String::compose (N_("Emitting %1 frames of silence"), to_do));
/* Do things in half second blocks as I think there may be limits
to what FFmpeg (and in particular the resampler) can cope with.
Matcher::repeat_last_video ()
{
if (!_last_image) {
- _last_image.reset (new SimpleImage (_pixel_format.get(), _size.get(), true));
- _last_image->make_black ();
+ shared_ptr<Image> im (new SimpleImage (_pixel_format.get(), _size.get(), true));
+ im->make_black ();
+ _last_image = im;
}
Video (_last_image, true, _last_subtitle);