Rework audio to deinterleave straight away and pass data
[dcpomatic.git] / src / lib / j2k_wav_encoder.cc
index e5d120ad66795d89d725a6101f093be135ae0945..58f8a101f7440c0be6312846114c9ebeed474011 100644 (file)
@@ -49,8 +49,6 @@ J2KWAVEncoder::J2KWAVEncoder (shared_ptr<const FilmState> s, shared_ptr<const Op
 #ifdef HAVE_SWRESAMPLE   
        , _swr_context (0)
 #endif   
-       , _deinterleave_buffer_size (8192)
-       , _deinterleave_buffer (0)
        , _process_end (false)
 {
        /* Create sound output files with .tmp suffixes; we will rename
@@ -68,15 +66,11 @@ J2KWAVEncoder::J2KWAVEncoder (shared_ptr<const FilmState> s, shared_ptr<const Op
                }
                _sound_files.push_back (f);
        }
-
-       /* Create buffer for deinterleaving audio */
-       _deinterleave_buffer = new uint8_t[_deinterleave_buffer_size];
 }
 
 J2KWAVEncoder::~J2KWAVEncoder ()
 {
        terminate_worker_threads ();
-       delete[] _deinterleave_buffer;
        close_sound_files ();
 }
 
@@ -230,14 +224,15 @@ J2KWAVEncoder::process_begin (int64_t audio_channel_layout, AVSampleFormat audio
                stringstream s;
                s << "Will resample audio from " << _fs->audio_sample_rate() << " to " << _fs->target_sample_rate();
                _log->log (s.str ());
-               
+
+               /* We will be using planar float data when we call the resampler */
                _swr_context = swr_alloc_set_opts (
                        0,
                        audio_channel_layout,
-                       audio_sample_format,
+                       AV_SAMPLE_FMT_FLTP,
                        _fs->target_sample_rate(),
                        audio_channel_layout,
-                       audio_sample_format,
+                       AV_SAMPLE_FMT_FLTP,
                        _fs->audio_sample_rate(),
                        0, 0
                        );
@@ -308,14 +303,13 @@ J2KWAVEncoder::process_end ()
 #if HAVE_SWRESAMPLE    
        if (_swr_context) {
 
+               float* out[_fs->audio_channels()];
+               for (int i = 0; i < _fs->audio_channels(); ++i) {
+                       out[i] = new float[256];
+               }
+                       
                while (1) {
-                       uint8_t buffer[256 * _fs->bytes_per_sample() * _fs->audio_channels()];
-                       uint8_t* out[2] = {
-                               buffer,
-                               0
-                       };
-
-                       int const frames = swr_convert (_swr_context, out, 256, 0, 0);
+                       int const frames = swr_convert (_swr_context, (uint8_t **) out, 256, 0, 0);
 
                        if (frames < 0) {
                                throw EncodeError ("could not run sample-rate converter");
@@ -325,7 +319,11 @@ J2KWAVEncoder::process_end ()
                                break;
                        }
 
-                       write_audio (buffer, frames * _fs->bytes_per_sample() * _fs->audio_channels());
+                       write_audio (out, frames);
+               }
+
+               for (int i = 0; i < _fs->audio_channels(); ++i) {
+                       delete[] out[i];
                }
 
                swr_free (&_swr_context);
@@ -344,97 +342,50 @@ J2KWAVEncoder::process_end ()
 }
 
 void
-J2KWAVEncoder::process_audio (uint8_t* data, int size)
+J2KWAVEncoder::process_audio (float** data, int frames)
 {
-       /* This is a buffer we might use if we are sample-rate converting;
-          it will need freeing if so.
-       */
-       uint8_t* out_buffer = 0;
+       float* resampled[_fs->audio_channels()];
        
-       /* Maybe sample-rate convert */
 #if HAVE_SWRESAMPLE    
+       /* Maybe sample-rate convert */
        if (_swr_context) {
 
-               uint8_t const * in[2] = {
-                       data,
-                       0
-               };
+               /* Compute the resampled frames count and add 32 for luck */
+               int const resampled_frames = ceil (frames * _fs->target_sample_rate() / _fs->audio_sample_rate()) + 32;
 
-               /* Here's samples per channel */
-               int const samples = size / _fs->bytes_per_sample();
-               
-               /* And here's frames (where 1 frame is a collection of samples, 1 for each channel,
-                  so for 5.1 a frame would be 6 samples)
-               */
-               int const frames = samples / _fs->audio_channels();
-
-               /* Compute the resampled frame count and add 32 for luck */
-               int const out_buffer_size_frames = ceil (frames * _fs->target_sample_rate() / _fs->audio_sample_rate()) + 32;
-               int const out_buffer_size_bytes = out_buffer_size_frames * _fs->audio_channels() * _fs->bytes_per_sample();
-               out_buffer = new uint8_t[out_buffer_size_bytes];
-
-               uint8_t* out[2] = {
-                       out_buffer, 
-                       0
-               };
+               /* Make a buffer to put the result in */
+               for (int i = 0; i < _fs->audio_channels(); ++i) {
+                       resampled[i] = new float[resampled_frames];
+               }
 
                /* Resample audio */
-               int out_frames = swr_convert (_swr_context, out, out_buffer_size_frames, in, frames);
+               int out_frames = swr_convert (_swr_context, (uint8_t **) resampled, resampled_frames, (uint8_t const **) data, frames);
                if (out_frames < 0) {
                        throw EncodeError ("could not run sample-rate converter");
                }
 
                /* And point our variables at the resampled audio */
-               data = out_buffer;
-               size = out_frames * _fs->audio_channels() * _fs->bytes_per_sample();
+               data = resampled;
+               frames = resampled_frames;
        }
 #endif
 
-       write_audio (data, size);
+       write_audio (data, frames);
 
-       /* Delete the sample-rate conversion buffer, if it exists */
-       delete[] out_buffer;
+#if HAVE_SWRESAMPLE
+       if (_swr_context) {
+               for (int i = 0; i < _fs->audio_channels(); ++i) {
+                       delete[] resampled[i];
+               }
+       }
+#endif 
 }
 
 void
-J2KWAVEncoder::write_audio (uint8_t* data, int size)
+J2KWAVEncoder::write_audio (float** data, int frames)
 {
-       /* XXX: we are assuming that the _deinterleave_buffer_size is a multiple
-          of the sample size and that size is a multiple of _fs->audio_channels * sample_size.
-       */
-
-       assert ((size % (_fs->audio_channels() * _fs->bytes_per_sample())) == 0);
-       assert ((_deinterleave_buffer_size % _fs->bytes_per_sample()) == 0);
-       
-       /* XXX: this code is very tricksy and it must be possible to make it simpler ... */
-       
-       /* Number of bytes left to read this time */
-       int remaining = size;
-       /* Our position in the output buffers, in bytes */
-       int position = 0;
-       while (remaining > 0) {
-               /* How many bytes of the deinterleaved data to do this time */
-               int this_time = min (remaining / _fs->audio_channels(), _deinterleave_buffer_size);
-               for (int i = 0; i < _fs->audio_channels(); ++i) {
-                       for (int j = 0; j < this_time; j += _fs->bytes_per_sample()) {
-                               for (int k = 0; k < _fs->bytes_per_sample(); ++k) {
-                                       int const to = j + k;
-                                       int const from = position + (i * _fs->bytes_per_sample()) + (j * _fs->audio_channels()) + k;
-                                       _deinterleave_buffer[to] = data[from];
-                               }
-                       }
-                       
-                       switch (_fs->audio_sample_format()) {
-                       case AV_SAMPLE_FMT_S16:
-                               sf_write_short (_sound_files[i], (const short *) _deinterleave_buffer, this_time / _fs->bytes_per_sample());
-                               break;
-                       default:
-                               throw EncodeError ("unknown audio sample format");
-                       }
-               }
-               
-               position += this_time;
-               remaining -= this_time * _fs->audio_channels();
+       for (int i = 0; i < _fs->audio_channels(); ++i) {
+               sf_write_float (_sound_files[i], data[i], frames);
        }
 }