All audio content should resample if the output frame rate and content

frame rates differ; make Sndfile sources use the video-frame-rate-based
calculation for output frame rate (like FFmpeg sources do).  Also, fix
small problems when flushing Resamplers used to get fed back to
Resamplers again.

1 files changed, 26 insertions, 0 deletions

diff --git a/src/lib/audio_content.cc b/src/lib/audio_content.cc
index dbca6652c..79912f1ae 100644
--- a/src/lib/audio_content.cc
+++ b/src/lib/audio_content.cc
@@ -29,6 +29,7 @@
 #include "i18n.h"
 
 using std::string;
+using std::cout;
 using std::vector;
 using boost::shared_ptr;
 using boost::dynamic_pointer_cast;
@@ -148,3 +149,28 @@ AudioContent::technical_summary () const
 {
 	return String::compose ("audio: channels %1, length %2, raw rate %3, out rate %4", audio_channels(), audio_length(), content_audio_frame_rate(), output_audio_frame_rate());
 }
+
+int
+AudioContent::output_audio_frame_rate () const
+{
+	shared_ptr<const Film> film = _film.lock ();
+	assert (film);
+	
+	/* Resample to a DCI-approved sample rate */
+	double t = dcp_audio_frame_rate (content_audio_frame_rate ());
+
+	FrameRateConversion frc (video_frame_rate(), film->video_frame_rate());
+
+	/* Compensate if the DCP is being run at a different frame rate
+	   to the source; that is, if the video is run such that it will
+	   look different in the DCP compared to the source (slower or faster).
+	   skip/repeat doesn't come into effect here.
+	*/
+
+	if (frc.change_speed) {
+		t *= video_frame_rate() * frc.factor() / film->video_frame_rate();
+	}
+
+	return rint (t);
+}
+