Commit 4ac9b64f authored by Michael Smith's avatar Michael Smith
Browse files

gst/audiorate/gstaudiorate.c: Fix audiorate, so that it accurately sets offsets and timestamps.

Original commit message from CVS:
* gst/audiorate/gstaudiorate.c: (gst_audio_rate_reset),
(gst_audio_rate_sink_event), (gst_audio_rate_chain):
Fix audiorate, so that it accurately sets offsets and timestamps.
Doesn't change the fundamental algorithmic decisions; so should be
safe.

* tests/check/Makefile.am:
Enable audiorate test now that it passes.
parent 8795eafa
2006-11-16 Michael Smith <msmith@fluendo.com>
* gst/audiorate/gstaudiorate.c: (gst_audio_rate_reset),
(gst_audio_rate_sink_event), (gst_audio_rate_chain):
Fix audiorate, so that it accurately sets offsets and timestamps.
Doesn't change the fundamental algorithmic decisions; so should be
safe.
* tests/check/Makefile.am:
Enable audiorate test now that it passes.
2006-11-09 Stefan Kost <ensonic@users.sf.net>
 
* sys/xvimage/xvimagesink.c: (gst_xvimagesink_change_state):
......@@ -17,6 +17,8 @@
* Boston, MA 02111-1307, USA.
*/
#include <stdlib.h>
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
......@@ -58,6 +60,7 @@ struct _GstAudioRate
/* audio state */
guint64 next_offset;
guint64 next_ts;
gboolean discont;
......@@ -207,6 +210,7 @@ static void
gst_audio_rate_reset (GstAudioRate * audiorate)
{
audiorate->next_offset = -1;
audiorate->next_ts = -1;
audiorate->discont = TRUE;
gst_segment_init (&audiorate->sink_segment, GST_FORMAT_UNDEFINED);
gst_segment_init (&audiorate->src_segment, GST_FORMAT_TIME);
......@@ -327,6 +331,7 @@ gst_audio_rate_sink_event (GstPad * pad, GstEvent * event)
* sample offset. We mark the offsets as invalid so that the _chain
* function will perform this calculation. */
audiorate->next_offset = -1;
audiorate->next_ts = -1;
}
/* we accept all formats */
......@@ -498,10 +503,21 @@ gst_audio_rate_chain (GstPad * pad, GstBuffer * buf)
GST_DEBUG_OBJECT (audiorate, "resync to offset %" G_GINT64_FORMAT, pos);
audiorate->next_offset = pos;
audiorate->next_ts = gst_util_uint64_scale_int (audiorate->next_offset,
GST_SECOND, audiorate->rate);
}
audiorate->in++;
static guint64 nextts = 0;
#define LLABS(a) ((gint64)(a) < 0 ? -(a):(a))
if (nextts != GST_BUFFER_TIMESTAMP (buf) && LLABS (GST_BUFFER_TIMESTAMP (buf) - nextts) > 21000) /* 21 us, ~1 sample */
GST_DEBUG_OBJECT (audiorate, "Expected %lld, got %lld! --> %lld", nextts,
GST_BUFFER_TIMESTAMP (buf),
LLABS (GST_BUFFER_TIMESTAMP (buf) - nextts));
nextts = GST_BUFFER_TIMESTAMP (buf) + GST_BUFFER_DURATION (buf);
in_time = GST_BUFFER_TIMESTAMP (buf);
in_size = GST_BUFFER_SIZE (buf);
in_samples = in_size / audiorate->bytes_per_sample;
......@@ -538,14 +554,21 @@ gst_audio_rate_chain (GstPad * pad, GstBuffer * buf)
/* FIXME, 0 might not be the silence byte for the negotiated format. */
memset (GST_BUFFER_DATA (fill), 0, fillsize);
GST_LOG_OBJECT (audiorate, "inserting %lld samples", fillsamples);
GST_DEBUG_OBJECT (audiorate, "inserting %lld samples", fillsamples);
GST_BUFFER_DURATION (fill) = in_duration * fillsize / in_size;
GST_BUFFER_TIMESTAMP (fill) = in_time - GST_BUFFER_DURATION (fill);
GST_BUFFER_OFFSET (fill) = audiorate->next_offset;
GST_BUFFER_OFFSET_END (fill) = in_offset;
/* we created this buffer to filla gap */
audiorate->next_offset += fillsamples;
GST_BUFFER_OFFSET_END (fill) = audiorate->next_offset;
/* Use next timestamp, then calculate following timestamp based on in_offset
* to get duration. Neccesary complexity to get 'perfect' streams */
GST_BUFFER_TIMESTAMP (fill) = audiorate->next_ts;
audiorate->next_ts = gst_util_uint64_scale_int (in_offset,
GST_SECOND, audiorate->rate);
GST_BUFFER_DURATION (fill) = audiorate->next_ts -
GST_BUFFER_TIMESTAMP (fill);
/* we created this buffer to fill a gap */
GST_BUFFER_FLAG_SET (fill, GST_BUFFER_FLAG_GAP);
/* set discont if it's pending, this is mostly done for the first buffer and
* after a flushing seek */
......@@ -570,7 +593,7 @@ gst_audio_rate_chain (GstPad * pad, GstBuffer * buf)
audiorate->drop += drop;
GST_LOG_OBJECT (audiorate, "dropping %lld samples", drop);
GST_DEBUG_OBJECT (audiorate, "dropping %lld samples", drop);
/* we can drop the buffer completely */
gst_buffer_unref (buf);
......@@ -590,13 +613,6 @@ gst_audio_rate_chain (GstPad * pad, GstBuffer * buf)
leftsize = in_size - truncsize;
trunc = gst_buffer_create_sub (buf, truncsize, leftsize);
GST_BUFFER_DURATION (trunc) = in_duration * leftsize / in_size;
GST_BUFFER_TIMESTAMP (trunc) =
in_time + in_duration - GST_BUFFER_DURATION (trunc);
GST_BUFFER_OFFSET (trunc) = audiorate->next_offset;
GST_BUFFER_OFFSET_END (trunc) = in_offset_end;
GST_LOG_OBJECT (audiorate, "truncating %lld samples", truncsamples);
gst_buffer_unref (buf);
buf = trunc;
......@@ -606,6 +622,17 @@ gst_audio_rate_chain (GstPad * pad, GstBuffer * buf)
audiorate->drop += truncsamples;
}
}
/* Now calculate parameters for whichever buffer (either the original
* or truncated one) we're pushing. */
GST_BUFFER_OFFSET (buf) = audiorate->next_offset;
GST_BUFFER_OFFSET_END (buf) = in_offset_end;
GST_BUFFER_TIMESTAMP (buf) = audiorate->next_ts;
audiorate->next_ts = gst_util_uint64_scale_int (in_offset_end,
GST_SECOND, audiorate->rate);
GST_BUFFER_DURATION (buf) = audiorate->next_ts - GST_BUFFER_TIMESTAMP (buf);
if (audiorate->discont) {
/* we need to output a discont buffer, do so now */
GST_DEBUG_OBJECT (audiorate, "marking DISCONT on output buffer");
......
......@@ -52,6 +52,7 @@ check_PROGRAMS = \
$(check_theora) \
elements/adder \
elements/audioconvert \
elements/audiorate \
elements/audioresample \
elements/audiotestsrc \
elements/gdpdepay \
......@@ -81,7 +82,6 @@ VALGRIND_TO_FIX = \
# these tests don't even pass
noinst_PROGRAMS = \
elements/audiorate \
elements/ffmpegcolorspace
AM_CFLAGS = $(GST_CFLAGS) $(GST_CHECK_CFLAGS)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment