aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMax Kellermann <max@duempel.org>2012-03-01 01:56:23 +0100
committerMax Kellermann <max@duempel.org>2012-03-01 02:05:40 +0100
commit9c36e71081841e2203ddb23ff9c9bef6ce65a0fe (patch)
tree7f3e4d4047d1a5cd3e6ee52c3ada4c3a0e65d0f1
parentc9c57af5f7995946007a029c1d21f7bc6834f1d0 (diff)
downloadmpd-9c36e71081841e2203ddb23ff9c9bef6ce65a0fe.tar.gz
mpd-9c36e71081841e2203ddb23ff9c9bef6ce65a0fe.tar.xz
mpd-9c36e71081841e2203ddb23ff9c9bef6ce65a0fe.zip
decoder/dsdiff: don't convert to PCM
Move the responsibility for the conversion to the PCM library. This will allow passing the verbatim DSD samples to an output plugin.
-rw-r--r--Makefile.am1
-rw-r--r--src/decoder/dsdiff_decoder_plugin.c47
2 files changed, 9 insertions, 39 deletions
diff --git a/Makefile.am b/Makefile.am
index c7089ae13..07782fdcb 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -481,7 +481,6 @@ libdecoder_plugins_a_SOURCES = \
src/decoder/pcm_decoder_plugin.c \
src/decoder/dsdiff_decoder_plugin.c \
src/decoder/dsdiff_decoder_plugin.h \
- src/dsd2pcm/dsd2pcm.c src/dsd2pcm/dsd2pcm.h \
src/decoder_buffer.c \
src/decoder_plugin.c \
src/decoder_list.c
diff --git a/src/decoder/dsdiff_decoder_plugin.c b/src/decoder/dsdiff_decoder_plugin.c
index 391a5bf78..751eff4d6 100644
--- a/src/decoder/dsdiff_decoder_plugin.c
+++ b/src/decoder/dsdiff_decoder_plugin.c
@@ -28,7 +28,6 @@
#include "dsdiff_decoder_plugin.h"
#include "decoder_api.h"
#include "audio_check.h"
-#include "dsd2pcm/dsd2pcm.h"
#include <unistd.h>
#include <stdio.h> /* for SEEK_SET, SEEK_CUR */
@@ -55,12 +54,14 @@ struct dsdiff_metadata {
unsigned sample_rate, channels;
};
-static bool lsbitfirst;
+static enum sample_format dsd_sample_format;
static bool
dsdiff_init(const struct config_param *param)
{
- lsbitfirst = config_get_block_bool(param, "lsbitfirst", false);
+ dsd_sample_format = config_get_block_bool(param, "lsbitfirst", false)
+ ? SAMPLE_FORMAT_DSD_LSBFIRST
+ : SAMPLE_FORMAT_DSD;
return true;
}
@@ -306,7 +307,7 @@ dsdiff_read_metadata(struct decoder *decoder, struct input_stream *is,
static bool
dsdiff_decode_chunk(struct decoder *decoder, struct input_stream *is,
unsigned channels,
- dsd2pcm_ctx **dsd2pcm, uint64_t chunk_size)
+ uint64_t chunk_size)
{
uint8_t buffer[8192];
const size_t sample_size = sizeof(buffer[0]);
@@ -314,18 +315,15 @@ dsdiff_decode_chunk(struct decoder *decoder, struct input_stream *is,
const unsigned buffer_frames = sizeof(buffer) / frame_size;
const unsigned buffer_samples = buffer_frames * frame_size;
const size_t buffer_size = buffer_samples * sample_size;
- float f_buffer[G_N_ELEMENTS(buffer)];
while (chunk_size > 0) {
/* see how much aligned data from the remaining chunk
fits into the local buffer */
unsigned now_frames = buffer_frames;
size_t now_size = buffer_size;
- unsigned now_samples = buffer_samples;
if (chunk_size < (uint64_t)now_size) {
now_frames = (unsigned)chunk_size / frame_size;
now_size = now_frames * frame_size;
- now_samples = now_frames * channels;
}
size_t nbytes = decoder_read(decoder, is, buffer, now_size);
@@ -334,20 +332,8 @@ dsdiff_decode_chunk(struct decoder *decoder, struct input_stream *is,
chunk_size -= nbytes;
- /* invoke the dsp2pcm library, once for each
- channel */
-
- for (unsigned c = 0; c < channels; ++c)
- dsd2pcm_translate(dsd2pcm[c], now_frames,
- buffer + c, channels,
- lsbitfirst, f_buffer + c, channels);
-
- /* convert to integer and submit to the decoder API */
-
enum decoder_command cmd =
- decoder_data(decoder, is, f_buffer,
- now_samples * sizeof(f_buffer[0]),
- 0);
+ decoder_data(decoder, is, buffer, nbytes, 0);
switch (cmd) {
case DECODE_COMMAND_NONE:
break;
@@ -381,25 +367,13 @@ dsdiff_stream_decode(struct decoder *decoder, struct input_stream *is)
GError *error = NULL;
struct audio_format audio_format;
if (!audio_format_init_checked(&audio_format, metadata.sample_rate / 8,
- SAMPLE_FORMAT_FLOAT,
+ dsd_sample_format,
metadata.channels, &error)) {
g_warning("%s", error->message);
g_error_free(error);
return;
}
- /* initialize the dsd2pcm library */
-
- dsd2pcm_ctx *dsd2pcm[MAX_CHANNELS];
- for (unsigned i = 0; i < metadata.channels; ++i) {
- dsd2pcm[i] = dsd2pcm_init();
- if (dsd2pcm[i] == NULL) {
- for (unsigned j = 0; j < i; ++j)
- dsd2pcm_destroy(dsd2pcm[j]);
- return;
- }
- }
-
/* success: file was recognized */
decoder_initialized(decoder, &audio_format, false, -1);
@@ -413,7 +387,7 @@ dsdiff_stream_decode(struct decoder *decoder, struct input_stream *is)
if (dsdiff_id_equals(&chunk_header.id, "DSD ")) {
if (!dsdiff_decode_chunk(decoder, is,
metadata.channels,
- dsd2pcm, chunk_size))
+ chunk_size))
break;
} else {
/* ignore other chunks */
@@ -428,9 +402,6 @@ dsdiff_stream_decode(struct decoder *decoder, struct input_stream *is)
if (!dsdiff_read_chunk_header(decoder, is, &chunk_header))
break;
}
-
- for (unsigned i = 0; i < metadata.channels; ++i)
- dsd2pcm_destroy(dsd2pcm[i]);
}
static bool
@@ -449,7 +420,7 @@ dsdiff_scan_stream(struct input_stream *is,
struct audio_format audio_format;
if (!audio_format_init_checked(&audio_format, metadata.sample_rate / 8,
- SAMPLE_FORMAT_S24_P32,
+ dsd_sample_format,
metadata.channels, NULL))
/* refuse to parse files which we cannot play anyway */
return false;