Fix analysis of audio when the subject is later in the playlist than

[dcpomatic.git] / src / lib / audio_buffers.h

diff --git a/src/lib/audio_buffers.h b/src/lib/audio_buffers.h
index bcf5d5928ee1330758573cf4d8a7417f46d76e56..122716d7765aef8b76eaba643ad9440cb74b6332 100644 (file)
--- a/src/lib/audio_buffers.h
+++ b/src/lib/audio_buffers.h
@@ -28,11 +28,15 @@
 
 /** @class AudioBuffers
  *  @brief A class to hold multi-channel audio data in float format.
+ *
+ *  The use of int32_t for frame counts in this class is due to the
+ *  round-up to the next power-of-2 code in ::ensure_size; if that
+ *  were changed the frame count could use any integer type.
  */
 class AudioBuffers
 {
 public:
-       AudioBuffers (int channels, int frames);
+       AudioBuffers (int channels, int32_t frames);
        AudioBuffers (AudioBuffers const &);
        AudioBuffers (boost::shared_ptr<const AudioBuffers>);
        ~AudioBuffers ();
@@ -42,7 +46,7 @@ public:
        boost::shared_ptr<AudioBuffers> clone () const;
        boost::shared_ptr<AudioBuffers> channel (int) const;
 
-       void ensure_size (int);
+       void ensure_size (int32_t);
 
        float** data () const {
                return _data;
@@ -58,30 +62,30 @@ public:
                return _frames;
        }
 
-       void set_frames (int f);
+       void set_frames (int32_t f);
 
        void make_silent ();
        void make_silent (int c);
-       void make_silent (int from, int frames);
+       void make_silent (int32_t from, int32_t frames);
 
        void apply_gain (float);
 
-       void copy_from (AudioBuffers const * from, int frames_to_copy, int read_offset, int write_offset);
+       void copy_from (AudioBuffers const * from, int32_t frames_to_copy, int32_t read_offset, int32_t write_offset);
        void copy_channel_from (AudioBuffers const * from, int from_channel, int to_channel);
-       void move (int from, int to, int frames);
+       void move (int32_t from, int32_t to, int32_t frames);
        void accumulate_channel (AudioBuffers const * from, int from_channel, int to_channel, float gain = 1);
-       void accumulate_frames (AudioBuffers const *, int read_offset, int write_offset, int frames);
+       void accumulate_frames (AudioBuffers const *, int32_t read_offset, int32_t write_offset, int32_t frames);
 
 private:
-       void allocate (int, int);
+       void allocate (int channels, int32_t frames);
        void deallocate ();
 
        /** Number of channels */
        int _channels;
        /** Number of frames (where a frame is one sample across all channels) */
-       int _frames;
+       int32_t _frames;
        /** Number of frames that _data can hold */
-       int _allocated_frames;
+       int32_t _allocated_frames;
        /** Audio data (so that, e.g. _data[2][6] is channel 2, sample 6) */
        float** _data;
 };