mirrors/gstreamer
mirrors/gstreamer
1
1
Fork 0
mirror of https://gitlab.freedesktop.org/gstreamer/gstreamer.git synced 2024-07-05 14:16:06 +00:00
Code Issues 1 Projects Releases Wiki Activity
gstreamer/ext/ffmpeg/gstffmpegcodecmap.c

1423 lines
39 KiB
C
Raw Normal View History

Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
/* GStreamer
* Copyright (C) <1999> Erik Walthinsen <omega@cse.ogi.edu>
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
* This file:
* Copyright (c) 2002-2003 Ronald Bultje <rbultje@ronald.bitfreak.net>
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public
* License along with this library; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 02111-1307, USA.
*/
fix ffmpeg to work with installd version as well Original commit message from CVS: fix ffmpeg to work with installd version as well
2002-11-26 14:50:05 +00:00
#include "config.h"
Make it compile against latest CVS of FFMPEG: Original commit message from CVS: Make it compile against latest CVS of FFMPEG: * Put gst/gst.h above avcodec.h because it needs some types defined in glib.h * Remove HQ (removed from lavc) property Also add some functions to make SVQ1 decoding work: * Add YUV9 colorspace * Let lavc do its own buffer allocation and copy from there instead of using our own buffer allocation functions. Somehow, that breaks. I don't know why.
2003-10-26 11:30:18 +00:00
#include <gst/gst.h>
fix ffmpeg to work with installd version as well Original commit message from CVS: fix ffmpeg to work with installd version as well
2002-11-26 14:50:05 +00:00
#ifdef HAVE_FFMPEG_UNINSTALLED
#include <avcodec.h>
#else
#include <ffmpeg/avcodec.h>
#endif
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
#include <string.h>
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
#include "gstffmpegcodecmap.h"
/* this macro makes a caps width fixed or unfixed width/height
* properties depending on whether we've got a context.
*
* See below for why we use this.
*/
#define GST_FF_VID_CAPS_NEW(name, mimetype, props...) \
(context != NULL) ? \
GST_CAPS_NEW (name, \
mimetype, \
"width", GST_PROPS_INT (context->width), \
"height", GST_PROPS_INT (context->height),\
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"framerate", GST_PROPS_FLOAT ( \
1.*context->frame_rate/ \
Fix compile problem on gcc-2.95. Fixes #125635. Original commit message from CVS: Fix compile problem on gcc-2.95. Fixes #125635.
2003-10-31 03:43:25 +00:00
context->frame_rate_base) , \
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
##props) \
: \
GST_CAPS_NEW (name, \
mimetype, \
"width", GST_PROPS_INT_RANGE (16, 4096), \
"height", GST_PROPS_INT_RANGE (16, 4096), \
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"framerate", GST_PROPS_FLOAT_RANGE (0, \
Fix compile problem on gcc-2.95. Fixes #125635. Original commit message from CVS: Fix compile problem on gcc-2.95. Fixes #125635.
2003-10-31 03:43:25 +00:00
G_MAXFLOAT) , \
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
##props)
/* same for audio - now with channels/sample rate
*/
#define GST_FF_AUD_CAPS_NEW(name, mimetype, props...) \
(context != NULL) ? \
GST_CAPS_NEW (name, \
mimetype, \
"rate", GST_PROPS_INT (context->sample_rate), \
Fix compile problem on gcc-2.95. Fixes #125635. Original commit message from CVS: Fix compile problem on gcc-2.95. Fixes #125635.
2003-10-31 03:43:25 +00:00
"channels", GST_PROPS_INT (context->channels) , \
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
##props) \
: \
GST_CAPS_NEW (name, \
mimetype, \
##props)
/* Convert a FFMPEG codec ID and optional AVCodecContext
* to a GstCaps. If the context is ommitted, no fixed values
* for video/audio size will be included in the GstCaps
*
* CodecID is primarily meant for compressed data GstCaps!
*/
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
GstCaps *
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
gst_ffmpeg_codecid_to_caps (enum CodecID codec_id,
AVCodecContext *context)
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
{
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
GstCaps *caps = NULL;
More ffmpeg wrapping going on here Original commit message from CVS: More ffmpeg wrapping going on here
2002-11-25 21:37:26 +00:00
switch (codec_id) {
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MPEG1VIDEO:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
/* this caps doesn't need width/height/framerate */
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
caps = GST_CAPS_NEW ("ffmpeg_mpeg1video",
"video/mpeg",
"mpegversion", GST_PROPS_INT (1),
"systemstream", GST_PROPS_BOOLEAN (FALSE)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
case CODEC_ID_H263P:
case CODEC_ID_H263I:
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_H263:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_h263",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-h263"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_RV10:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_rv10",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-pn-realvideo",
"systemstream", GST_PROPS_BOOLEAN (FALSE)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MP2:
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mp2",
random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose Original commit message from CVS: random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose
2003-11-04 22:31:05 +00:00
"audio/mpeg",
"layer", GST_PROPS_INT (2)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MP3LAME:
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mp3",
random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose Original commit message from CVS: random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose
2003-11-04 22:31:05 +00:00
"audio/mpeg",
"layer", GST_PROPS_INT (3)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
case CODEC_ID_VORBIS: /* FIXME? vorbis or ogg? */
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_vorbis",
random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose Original commit message from CVS: random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose
2003-11-04 22:31:05 +00:00
"application/ogg",
NULL
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_AC3:
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_ac3",
random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose Original commit message from CVS: random fixes to ffmpeg: decoding of mp3 now works, call parent_class->dispose
2003-11-04 22:31:05 +00:00
"audio/x-ac3",
NULL
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MJPEG:
Add missing codecs Original commit message from CVS: Add missing codecs
2003-01-07 20:42:56 +00:00
case CODEC_ID_MJPEGB:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
/*case CODEC_ID_LJPEG:*/
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_mjpeg",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-jpeg"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
Add missing codecs Original commit message from CVS: Add missing codecs
2003-01-07 20:42:56 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MPEG4:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_mpeg4",
"video/mpeg",
"mpegversion", GST_PROPS_INT (4),
"systemstream", GST_PROPS_BOOLEAN (FALSE)
);
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_divx",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-divx",
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
"divxversion", GST_PROPS_INT (5)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
));
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_divx",
"video/x-divx",
"divxversion", GST_PROPS_INT (4)
));
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_xvid",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-xvid"
));
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_3ivx",
More compilation fixes from my latest commit Original commit message from CVS: More compilation fixes from my latest commit
2003-07-07 11:32:34 +00:00
"video/x-3ivx"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
));
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
/* weird quasi-codecs for the demuxers only */
case CODEC_ID_RAWVIDEO:
/* we use a shortcut to the raw-video pad function */
return gst_ffmpeg_codectype_to_caps (CODEC_TYPE_VIDEO, context);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MSMPEG4V1:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_msmpeg4v1",
"video/x-msmpeg",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"msmpegversion", GST_PROPS_INT (41)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MSMPEG4V2:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_msmpeg4v2",
"video/x-msmpeg",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"msmpegversion", GST_PROPS_INT (42)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MSMPEG4V3:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_msmpeg4v3",
"video/x-msmpeg",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"msmpegversion", GST_PROPS_INT (43)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_msmpeg4v3_divx3",
"video/x-divx",
"divxversion", GST_PROPS_INT (3)
));
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_WMV1:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_wmv1",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-wmv",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"wmvversion", GST_PROPS_INT (1)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_WMV2:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_wmv2",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-wmv",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"wmvversion", GST_PROPS_INT (2)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_SVQ1:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_svq1",
"video/x-svq",
"svqversion", GST_PROPS_INT (1)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
case CODEC_ID_SVQ3:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_svq3",
"video/x-svq",
"svqversion", GST_PROPS_INT (3)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
case CODEC_ID_DVAUDIO:
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_dvaudio",
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
"audio/x-dv",
NULL);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
case CODEC_ID_DVVIDEO:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_dvvideo",
"video/dv"
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_WMAV1:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_wma1",
"audio/x-wma",
"wmaversion", GST_PROPS_INT (1)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_WMAV2:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_wma2",
"audio/x-wma",
"wmaversion", GST_PROPS_INT (2)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MACE3:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mace3",
"audio/x-mace",
"maceversion", GST_PROPS_INT (3)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_MACE6:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mace6",
"audio/x-mace",
"maceversion", GST_PROPS_INT (6)
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Add missing codecs Original commit message from CVS: Add missing codecs
2003-01-07 20:42:56 +00:00
case CODEC_ID_HUFFYUV:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_huffyuv",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-huffyuv"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
break;
case CODEC_ID_CYUV:
/* .. */
break;
case CODEC_ID_H264:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_h264",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-h264"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
break;
case CODEC_ID_INDEO3:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_indeo3",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-indeo",
"indeoversion", GST_PROPS_INT (3)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
break;
case CODEC_ID_VP3:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_vp3",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-vp3"
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
caps = gst_caps_append(caps,
GST_FF_VID_CAPS_NEW ("ffmpeg_theora",
"video/x-theora"
));
break;
case CODEC_ID_AAC:
Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't l... Original commit message from CVS: Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't list itself after gst-register'ing...
2003-10-26 12:10:01 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mpeg2aac",
"audio/mpeg",
"systemstream", GST_PROPS_BOOLEAN (FALSE),
"mpegversion", GST_PROPS_INT (2)
);
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case CODEC_ID_MPEG4AAC:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mpeg4aac",
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
"audio/mpeg",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"systemstream", GST_PROPS_BOOLEAN (FALSE),
"mpegversion", GST_PROPS_INT (4)
);
break;
case CODEC_ID_ASV1:
/* .. */
break;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
case CODEC_ID_FFV1:
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_ffv1",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-ffv",
"ffvversion", GST_PROPS_INT (1)
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
);
break;
case CODEC_ID_4XM:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_4xmvideo",
More compilation fixes from my latest commit Original commit message from CVS: More compilation fixes from my latest commit
2003-07-07 11:32:34 +00:00
"video/x-4xm"
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
);
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
break;
/* weird quasi-codecs for the demuxers only */
case CODEC_ID_PCM_S16LE:
case CODEC_ID_PCM_S16BE:
case CODEC_ID_PCM_U16LE:
case CODEC_ID_PCM_U16BE:
case CODEC_ID_PCM_S8:
case CODEC_ID_PCM_U8:
do {
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
gint width = 0, depth = 0, endianness = 0;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
gboolean signedness = FALSE; /* blabla */
switch (codec_id) {
case CODEC_ID_PCM_S16LE:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 16; depth = 16;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_LITTLE_ENDIAN;
signedness = TRUE;
break;
case CODEC_ID_PCM_S16BE:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 16; depth = 16;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_BIG_ENDIAN;
signedness = TRUE;
break;
case CODEC_ID_PCM_U16LE:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 16; depth = 16;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_LITTLE_ENDIAN;
signedness = FALSE;
break;
case CODEC_ID_PCM_U16BE:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 16; depth = 16;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_BIG_ENDIAN;
signedness = FALSE;
break;
case CODEC_ID_PCM_S8:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 8; depth = 8;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_BYTE_ORDER;
signedness = TRUE;
break;
case CODEC_ID_PCM_U8:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
width = 8; depth = 8;
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
endianness = G_BYTE_ORDER;
signedness = FALSE;
break;
default:
g_assert(0); /* don't worry, we never get here */
break;
}
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW (
"ffmpeg_pcmaudio",
"audio/x-raw-int",
"width", GST_PROPS_INT (width),
"depth", GST_PROPS_INT (depth),
"endianness", GST_PROPS_INT (endianness),
"signed", GST_PROPS_BOOLEAN (signedness)
);
We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticin... Original commit message from CVS: We also need raw video/audio caps for the demuxer/muxer, plus that caps_to_codecid() didn't exist (thanks for noticing, dave)
2003-06-17 11:44:11 +00:00
} while (0);
break;
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
case CODEC_ID_PCM_MULAW:
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_mulawaudio",
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
"audio/x-mulaw",
NULL);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
break;
case CODEC_ID_PCM_ALAW:
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_alawaudio",
fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ens... Original commit message from CVS: fix GST_FF_AUD_CAPS_NEW to not include channels and rate in unfixed caps and make audio caps use them again. This ensures that audio encoders set these properties when setting fixed caps and audio decoders can advertise fixed caps, too. This should work in all cases. If not, change it back and ffdec_mp3 is broken again ;)
2003-11-05 03:06:10 +00:00
"audio/x-alaw",
NULL);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
break;
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_ADPCM_IMA_QT:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_adpcm_ima_qt",
"audio/x-adpcm",
"layout", GST_PROPS_STRING ("quicktime")
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_ADPCM_IMA_WAV:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_adpcm_ima_wav",
"audio/x-adpcm",
"layout", GST_PROPS_STRING ("wav")
);
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
case CODEC_ID_ADPCM_MS:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_adpcm_ms",
"audio/x-adpcm",
"layout", GST_PROPS_STRING ("microsoft")
);
break;
case CODEC_ID_ADPCM_4XM:
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_adpcm_4xm",
"audio/x-adpcm",
"layout", GST_PROPS_STRING ("4xm")
);
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case CODEC_ID_AMR_NB:
/* .. */
break;
Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't l... Original commit message from CVS: Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't list itself after gst-register'ing...
2003-10-26 12:10:01 +00:00
case CODEC_ID_RA_144:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_realaudio_144",
"audio/x-pn-realaudio",
"bitrate", GST_PROPS_INT (14400)
);
break;
case CODEC_ID_RA_288:
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_realaudio_288",
"audio/x-pn-realaudio",
"bitrate", GST_PROPS_INT (28800)
);
Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't l... Original commit message from CVS: Re-enable realaudio codecs, improve decoder naming and add mpeg2aac decoder... Now to find out why it still doesn't list itself after gst-register'ing...
2003-10-26 12:10:01 +00:00
break;
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
default:
/* .. */
break;
}
if (caps != NULL) {
char *str = g_strdup_printf("The caps that belongs to codec_id=%d",
codec_id);
gst_caps_debug(caps, str);
g_free(str);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
} else {
char *str = g_strdup_printf("No caps found for codec_id=%d",
codec_id);
gst_caps_debug(caps, str);
g_free(str);
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
}
return caps;
}
/* Convert a FFMPEG Pixel Format and optional AVCodecContext
* to a GstCaps. If the context is ommitted, no fixed values
* for video/audio size will be included in the GstCaps
*
* See below for usefullness
*/
static GstCaps *
gst_ffmpeg_pixfmt_to_caps (enum PixelFormat pix_fmt,
AVCodecContext *context)
{
GstCaps *caps = NULL;
int bpp = 0, depth = 0, endianness = 0;
gulong g_mask = 0, r_mask = 0, b_mask = 0;
guint32 fmt = 0;
switch (pix_fmt) {
case PIX_FMT_YUV420P:
fmt = GST_MAKE_FOURCC ('I','4','2','0');
break;
case PIX_FMT_YUV422:
fmt = GST_MAKE_FOURCC ('Y','U','Y','2');
break;
case PIX_FMT_RGB24:
bpp = depth = 24;
endianness = G_BIG_ENDIAN;
r_mask = 0xff0000; g_mask = 0x00ff00; b_mask = 0x0000ff;
break;
case PIX_FMT_BGR24:
bpp = depth = 24;
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
endianness = G_BIG_ENDIAN;
r_mask = 0x0000ff; g_mask = 0x00ff00; b_mask = 0xff0000;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case PIX_FMT_YUV422P:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
fmt = GST_MAKE_FOURCC ('Y','4','2','B');
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case PIX_FMT_YUV444P:
/* .. */
break;
case PIX_FMT_RGBA32:
bpp = depth = 32;
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
endianness = G_BIG_ENDIAN;
#if (G_BYTE_ORDER == G_BIG_ENDIAN)
r_mask = 0xff000000; g_mask = 0x00ff0000; b_mask = 0x0000ff00;
#else
r_mask = 0x000000ff; g_mask = 0x0000ff00; b_mask = 0x00ff0000;
#endif
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case PIX_FMT_YUV410P:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
fmt = GST_MAKE_FOURCC ('Y','U','V','9');
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case PIX_FMT_YUV411P:
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
fmt = GST_MAKE_FOURCC ('Y','4','1','B');
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
break;
case PIX_FMT_RGB565:
bpp = depth = 16;
endianness = G_BYTE_ORDER;
r_mask = 0xf800; g_mask = 0x07e0; b_mask = 0x001f;
break;
case PIX_FMT_RGB555:
bpp = 16; depth = 15;
endianness = G_BYTE_ORDER;
r_mask = 0x7c00; g_mask = 0x03e0; b_mask = 0x001f;
break;
default:
/* give up ... */
break;
}
if (bpp != 0) {
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_rawvideo",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-raw-rgb",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"bpp", GST_PROPS_INT (bpp),
"depth", GST_PROPS_INT (depth),
"red_mask", GST_PROPS_INT (r_mask),
"green_mask", GST_PROPS_INT (g_mask),
"blue_mask", GST_PROPS_INT (b_mask),
"endianness", GST_PROPS_INT (endianness)
);
} else if (fmt) {
caps = GST_FF_VID_CAPS_NEW ("ffmpeg_rawvideo",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-raw-yuv",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"format", GST_PROPS_FOURCC (fmt)
);
}
if (caps != NULL) {
char *str = g_strdup_printf("The caps that belongs to pix_fmt=%d",
pix_fmt);
gst_caps_debug(caps, str);
g_free(str);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
} else {
char *str = g_strdup_printf("No caps found for pix_fmt=%d",
pix_fmt);
gst_caps_debug(caps, str);
g_free(str);
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
}
return caps;
}
/* Convert a FFMPEG Sample Format and optional AVCodecContext
* to a GstCaps. If the context is ommitted, no fixed values
* for video/audio size will be included in the GstCaps
*
* See below for usefullness
*/
static GstCaps *
gst_ffmpeg_smpfmt_to_caps (enum SampleFormat sample_fmt,
AVCodecContext *context)
{
GstCaps *caps = NULL;
int bpp = 0;
gboolean signedness = FALSE;
switch (sample_fmt) {
case SAMPLE_FMT_S16:
signedness = TRUE;
bpp = 16;
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
default:
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
/* .. */
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
break;
}
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
if (bpp) {
caps = GST_FF_AUD_CAPS_NEW ("ffmpeg_rawaudio",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"audio/x-raw-int",
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
"signed", GST_PROPS_BOOLEAN (signedness),
"endianness", GST_PROPS_INT (G_BYTE_ORDER),
"width", GST_PROPS_INT (bpp),
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"depth", GST_PROPS_INT (bpp)
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
);
}
if (caps != NULL) {
char *str = g_strdup_printf("The caps that belongs to sample_fmt=%d",
sample_fmt);
gst_caps_debug(caps, str);
g_free(str);
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
} else {
char *str = g_strdup_printf("No caps found for sample_fmt=%d",
sample_fmt);
gst_caps_debug(caps, str);
g_free(str);
remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype syst... Original commit message from CVS: * remove gstffmpegdecall. This was a temporary hack to get around some issues with our video/avi-centered mimetype system. Now that we use proper identifiers, we don't need that piece of cruft anymore. * update ffmpeg codec map for ffdec_*, plus make the uncompressed stream caps also be formed by the codec mapper (this is easier to handle imo). * add audio support (untested, though :P). * remove the copying of buffer content and add a get_buffer() handler instead. * add an int to ffmpegenc to make it compile and resolve the symbol. I'll truely fix that later on. * make mimetypes in the ffmpeg codec mapper conform to docs/random/mimetypes * etc. Oh, and let's not forget that it can play back DIVX5 et all nicely. Support for SVQ1/3 is also there, so if someone completes qtdemux, we can watch these, too.
2003-06-07 00:41:32 +00:00
}
return caps;
}
/* Convert a FFMPEG codec Type and optional AVCodecContext
* to a GstCaps. If the context is ommitted, no fixed values
* for video/audio size will be included in the GstCaps
*
* CodecType is primarily meant for uncompressed data GstCaps!
*/
GstCaps *
gst_ffmpeg_codectype_to_caps (enum CodecType codec_type,
AVCodecContext *context)
{
GstCaps *caps = NULL;
switch (codec_type) {
case CODEC_TYPE_VIDEO:
if (context) {
caps = gst_ffmpeg_pixfmt_to_caps (context->pix_fmt, context);
} else {
GstCaps *temp;
enum PixelFormat i;
for (i = 0; i < PIX_FMT_NB; i++) {
temp = gst_ffmpeg_pixfmt_to_caps (i, NULL);
if (temp != NULL) {
caps = gst_caps_append (caps, temp);
}
}
}
break;
case CODEC_TYPE_AUDIO:
if (context) {
caps = gst_ffmpeg_smpfmt_to_caps (context->sample_fmt, context);
} else {
GstCaps *temp;
enum SampleFormat i;
for (i = 0; i <= SAMPLE_FMT_S16; i++) {
temp = gst_ffmpeg_smpfmt_to_caps (i, NULL);
if (temp != NULL) {
caps = gst_caps_append (caps, temp);
}
}
}
break;
default:
/* .. */
break;
}
return caps;
Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. Original commit message from CVS: Release early.. try to wrap more ffmpeg stuff, like muxers and demuxers. This thing needs a fix in ffmpeg (which I will send to the ffmpeg authors eventually)
2002-11-06 23:53:46 +00:00
}
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
/* Convert a GstCaps (audio/raw) to a FFMPEG SampleFmt
* and other audio properties in a AVCodecContext.
*
* For usefullness, see below
*/
static void
gst_ffmpeg_caps_to_smpfmt (GstCaps *caps,
AVCodecContext *context)
{
if (gst_caps_has_property_typed (caps, "width",
GST_PROPS_INT_TYPE) &&
gst_caps_has_property_typed (caps, "depth",
GST_PROPS_INT_TYPE) &&
gst_caps_has_property_typed (caps, "signed",
GST_PROPS_BOOLEAN_TYPE) &&
gst_caps_has_property_typed (caps, "endianness",
GST_PROPS_INT_TYPE)) {
gint depth = 0, width = 0, endianness = 0;
gboolean signedness = FALSE;
gst_caps_get (caps,
"width", &width,
"depth", &depth,
"endianness", &endianness,
"signed", &signedness,
NULL);
if (width == 16 && depth == 16 &&
endianness == G_BYTE_ORDER && signedness == TRUE) {
context->sample_fmt = SAMPLE_FMT_S16;
}
}
if (gst_caps_has_property_typed (caps, "channels",
GST_PROPS_INT_TYPE)) {
Fix stupid typo Original commit message from CVS: Fix stupid typo
2003-10-22 10:17:15 +00:00
gst_caps_get_int (caps, "channels", &context->channels);
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
}
if (gst_caps_has_property_typed (caps, "rate",
GST_PROPS_INT_TYPE)) {
gst_caps_get_int (caps, "rate", &context->sample_rate);
}
}
/* Convert a GstCaps (video/raw) to a FFMPEG PixFmt
* and other video properties in a AVCodecContext.
*
* For usefullness, see below
*/
static void
gst_ffmpeg_caps_to_pixfmt (GstCaps *caps,
AVCodecContext *context)
{
if (gst_caps_has_property_typed (caps, "width",
GST_PROPS_INT_TYPE) &&
gst_caps_has_property_typed (caps, "height",
GST_PROPS_INT_TYPE)) {
gst_caps_get (caps,
"width", &context->width,
"height", &context->height,
NULL);
}
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
if (gst_caps_has_property_typed (caps, "framerate",
GST_PROPS_FLOAT_TYPE)) {
gfloat fps;
gst_caps_get_float (caps, "framerate", &fps);
context->frame_rate = fps * 1000;
context->frame_rate_base = 1000;
}
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
if (gst_caps_has_property_typed (caps, "format",
GST_PROPS_FOURCC_TYPE)) {
guint32 fourcc;
gst_caps_get_fourcc_int (caps, "format", &fourcc);
switch (fourcc) {
case GST_MAKE_FOURCC ('Y','U','Y','2'):
context->pix_fmt = PIX_FMT_YUV422;
break;
case GST_MAKE_FOURCC ('I','4','2','0'):
context->pix_fmt = PIX_FMT_YUV420P;
break;
Make it compile against latest CVS of FFMPEG: Original commit message from CVS: Make it compile against latest CVS of FFMPEG: * Put gst/gst.h above avcodec.h because it needs some types defined in glib.h * Remove HQ (removed from lavc) property Also add some functions to make SVQ1 decoding work: * Add YUV9 colorspace * Let lavc do its own buffer allocation and copy from there instead of using our own buffer allocation functions. Somehow, that breaks. I don't know why.
2003-10-26 11:30:18 +00:00
case GST_MAKE_FOURCC ('Y','4','1','B'):
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
context->pix_fmt = PIX_FMT_YUV411P;
break;
Make it compile against latest CVS of FFMPEG: Original commit message from CVS: Make it compile against latest CVS of FFMPEG: * Put gst/gst.h above avcodec.h because it needs some types defined in glib.h * Remove HQ (removed from lavc) property Also add some functions to make SVQ1 decoding work: * Add YUV9 colorspace * Let lavc do its own buffer allocation and copy from there instead of using our own buffer allocation functions. Somehow, that breaks. I don't know why.
2003-10-26 11:30:18 +00:00
case GST_MAKE_FOURCC ('Y','U','V','9'):
context->pix_fmt = PIX_FMT_YUV410P;
break;
#if 0
case FIXME:
context->pix_fmt = PIX_FMT_YUV444P;
break;
#endif
Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions Original commit message from CVS: Add a caps_to_codecid() function to read properties from a caps function for use in _connect() functions
2003-06-07 20:45:56 +00:00
case GST_MAKE_FOURCC ('R','G','B',' '):
if (gst_caps_has_property_typed (caps, "depth",
GST_PROPS_INT_TYPE) &&
gst_caps_has_property_typed (caps, "endianness",
GST_PROPS_INT_TYPE)) {
gint depth = 0, endianness = 0;
gst_caps_get_int (caps, "depth", &depth);
gst_caps_get_int (caps, "endianness", &endianness);
switch (depth) {
case 32:
if (endianness == G_BYTE_ORDER)
context->pix_fmt = PIX_FMT_RGBA32;
break;
case 24:
switch (endianness) {
case G_LITTLE_ENDIAN:
context->pix_fmt = PIX_FMT_BGR24;
break;
case G_BIG_ENDIAN:
context->pix_fmt = PIX_FMT_RGB24;
break;
default:
/* nothing */
break;
}
break;
case 16:
if (endianness == G_BYTE_ORDER)
context->pix_fmt = PIX_FMT_RGB565;
break;
case 15:
if (endianness == G_BYTE_ORDER)
context->pix_fmt = PIX_FMT_RGB555;
break;
default:
/* nothing */
break;
}
}
break;
default:
/* nothing */
break;
}
}
}
/* Convert a GstCaps and a FFMPEG codec Type to a
* AVCodecContext. If the context is ommitted, no fixed values
* for video/audio size will be included in the context
*
* CodecType is primarily meant for uncompressed data GstCaps!
*/
void
gst_ffmpeg_caps_to_codectype (enum CodecType type,
GstCaps *caps,
AVCodecContext *context)
{
if (context == NULL)
return;
switch (type) {
case CODEC_TYPE_VIDEO:
gst_ffmpeg_caps_to_pixfmt (caps, context);
break;
case CODEC_TYPE_AUDIO:
gst_ffmpeg_caps_to_smpfmt (caps, context);
break;
default:
/* unknown */
break;
}
}
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
/* _formatid_to_caps () is meant for muxers/demuxers, it
* transforms a name (ffmpeg way of ID'ing these, why don't
* they have unique numerical IDs?) to the corresponding
* caps belonging to that mux-format
*
* Note: we don't need any additional info because the caps
* isn't supposed to contain any useful info besides the
* media type anyway
*/
GstCaps *
gst_ffmpeg_formatid_to_caps (const gchar *format_name)
{
GstCaps *caps = NULL;
if (!strcmp (format_name, "mpeg")) {
caps = GST_CAPS_NEW ("ffmpeg_mpeg",
"video/mpeg",
"systemstream", GST_PROPS_BOOLEAN (TRUE)
);
} else if (!strcmp (format_name, "mpegts")) {
caps = GST_CAPS_NEW ("ffmpeg_mpegts",
"video/mpegts",
"systemstream", GST_PROPS_BOOLEAN (TRUE)
);
} else if (!strcmp (format_name, "rm")) {
caps = GST_CAPS_NEW ("ffmpeg_rm",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"audio/x-pn-realvideo",
"systemstream", GST_PROPS_BOOLEAN (TRUE)
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
);
} else if (!strcmp (format_name, "asf")) {
caps = GST_CAPS_NEW ("ffmpeg_asf",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-asf",
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
NULL
);
} else if (!strcmp (format_name, "avi")) {
caps = GST_CAPS_NEW ("ffmpeg_avi",
"video/avi",
NULL
);
} else if (!strcmp (format_name, "wav")) {
caps = GST_CAPS_NEW ("ffmpeg_wav",
"video/x-wav",
NULL
);
} else if (!strcmp (format_name, "swf")) {
caps = GST_CAPS_NEW ("ffmpeg_swf",
"application/x-shockwave-flash",
NULL
);
} else if (!strcmp (format_name, "au")) {
caps = GST_CAPS_NEW ("ffmpeg_au",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"audio/x-au",
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
NULL
);
} else if (!strcmp (format_name, "mov")) {
caps = GST_CAPS_NEW ("ffmpeg_quicktime",
"video/quicktime",
NULL
);
} else if (!strcmp (format_name, "dv")) {
caps = GST_CAPS_NEW ("ffmpeg_dv",
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
"video/x-dv",
MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format ... Original commit message from CVS: MPEG4AAC is MPEG-4 audio, not video, and implement a _formatid_to_caps() function which tracks a GstCaps to a format name (muxing format etc.), useful for setting caps on demuxers
2003-06-08 13:30:10 +00:00
"systemstream", GST_PROPS_BOOLEAN (TRUE)
);
} else if (!strcmp (format_name, "4xm")) {
caps = GST_CAPS_NEW ("ffmpeg_4xm",
"video/x-4xm",
NULL
);
} else {
/* unknown! */
}
return caps;
}
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
/* Convert a GstCaps to a FFMPEG codec ID. Size et all
* are omitted, that can be queried by the user itself,
* we're not eating the GstCaps or anything
* A pointer to an allocated context is also needed for
* optional extra info (not used yet, though)
*/
enum CodecID
gst_ffmpeg_caps_to_codecid (GstCaps *caps,
AVCodecContext *context)
{
enum CodecID id = CODEC_ID_NONE;
const gchar *mimetype;
gboolean video = FALSE, audio = FALSE; /* we want to be sure! */
g_return_val_if_fail (caps != NULL, CODEC_ID_NONE);
mimetype = gst_caps_get_mime(caps);
if (!strcmp (mimetype, "video/x-raw-rgb") ||
!strcmp (mimetype, "video/x-raw-yuv")) {
id = CODEC_ID_RAWVIDEO;
if (context != NULL) {
gint depth = 0, endianness = 0;
guint32 fmt_fcc = 0;
enum PixelFormat pix_fmt = -1;
More compilation fixes from my latest commit Original commit message from CVS: More compilation fixes from my latest commit
2003-07-07 11:32:34 +00:00
if (gst_caps_has_property (caps, "format"))
New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as descri... Original commit message from CVS: New mimetypes gone into effect today - this commit changes all old mimetypes over to the new mimetypes spec as described in the previous commit's document. Note: some plugins will break, some pipelines will break, expect HEAD to be broken or at least not 100% working for a few days, but don't forget to report bugs
2003-07-06 20:49:15 +00:00
gst_caps_get_fourcc_int (caps, "format", &fmt_fcc);
else
fmt_fcc = GST_MAKE_FOURCC ('R','G','B',' ');
switch (fmt_fcc) {
case GST_MAKE_FOURCC ('R','G','B',' '):
gst_caps_get_int (caps, "endianness", &endianness);
gst_caps_get_int (caps, "depth", &depth);
switch (depth) {
case 15:
if (endianness == G_BYTE_ORDER) {
pix_fmt = PIX_FMT_RGB555;
}
break;
case 16:
if (endianness == G_BYTE_ORDER) {
pix_fmt = PIX_FMT_RGB565;
}
break;
case 24:
if (endianness == G_BIG_ENDIAN) {
pix_fmt = PIX_FMT_RGB24;
} else {
pix_fmt = PIX_FMT_BGR24;
}
break;
case 32:
if (endianness == G_BIG_ENDIAN) {
pix_fmt = PIX_FMT_RGBA32;
}
break;
default:
/* ... */
break;
}
break;
case GST_MAKE_FOURCC ('Y','U','Y','2'):
pix_fmt = PIX_FMT_YUV422;
break;
case GST_MAKE_FOURCC ('I','4','2','0'):
pix_fmt = PIX_FMT_YUV420P;
break;
case GST_MAKE_FOURCC ('Y','4','1','B'):
pix_fmt = PIX_FMT_YUV411P;
break;
case GST_MAKE_FOURCC ('Y','4','2','B'):
pix_fmt = PIX_FMT_YUV422P;
break;
case GST_MAKE_FOURCC ('Y','U','V','9'):
pix_fmt = PIX_FMT_YUV410P;
break;
default:
/* ... */
break;
}
/* only set if actually recognized! */
if (pix_fmt != -1) {
video = TRUE;
context->pix_fmt = pix_fmt;
} else {
id = CODEC_ID_NONE;
}
}
} else if (!strcmp (mimetype, "audio/x-raw-int")) {
gint depth = 0, width = 0, endianness = 0;
gboolean signedness = FALSE; /* bla default value */
if (gst_caps_has_property(caps, "signed")) {
gst_caps_get_int(caps, "endianness", &endianness);
gst_caps_get_boolean(caps, "signed", &signedness);
gst_caps_get_int(caps, "width", &width);
gst_caps_get_int(caps, "depth", &depth);
if (context) {
context->sample_rate = 0;
context->channels = 0;
gst_caps_get_int(caps, "channels", &context->channels);
gst_caps_get_int(caps, "rate", &context->sample_rate);
}
if (depth == width) {
switch (depth) {
case 8:
if (signedness) {
id = CODEC_ID_PCM_S8;
} else {
id = CODEC_ID_PCM_U8;
}
break;
case 16:
switch (endianness) {
case G_BIG_ENDIAN:
if (signedness) {
id = CODEC_ID_PCM_S16BE;
} else {
id = CODEC_ID_PCM_U16BE;
}
break;
case G_LITTLE_ENDIAN:
if (signedness) {
id = CODEC_ID_PCM_S16LE;
} else {
id = CODEC_ID_PCM_U16LE;
}
break;
}
break;
}
if (id != CODEC_ID_NONE) {
audio = TRUE;
}
}
}
} else if (!strcmp(mimetype, "audio/x-mulaw")) {
id = CODEC_ID_PCM_MULAW;
audio = TRUE;
} else if (!strcmp(mimetype, "audio/x-alaw")) {
id = CODEC_ID_PCM_ALAW;
audio = TRUE;
} else if (!strcmp(mimetype, "video/x-dv")) {
id = CODEC_ID_DVVIDEO;
video = TRUE;
} else if (!strcmp(mimetype, "audio/x-dv")) { /* ??? */
id = CODEC_ID_DVAUDIO;
audio = TRUE;
} else if (!strcmp(mimetype, "video/x-h263")) {
id = CODEC_ID_H263; /* or H263[IP] */
video = TRUE;
} else if (!strcmp(mimetype, "video/mpeg")) {
gboolean sys_strm = TRUE;
gint mpegversion = 0;
if (gst_caps_has_property(caps, "systemstream")) {
gst_caps_get_boolean(caps, "systemstream", &sys_strm);
}
if (!sys_strm && gst_caps_has_property(caps, "mpegversion")) {
gst_caps_get_int(caps, "mpegversion", &mpegversion);
switch (mpegversion) {
case 1:
id = CODEC_ID_MPEG1VIDEO;
break;
case 4:
id = CODEC_ID_MPEG4;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp(mimetype, "video/x-jpeg")) {
id = CODEC_ID_MJPEG; /* A... B... */
video = TRUE;
} else if (!strcmp(mimetype, "video/x-wmv")) {
if (gst_caps_has_property (caps, "wmvversion")) {
gint wmvversion = 0;
gst_caps_get_int (caps, "wmvversion", &wmvversion);
switch (wmvversion) {
case 1:
id = CODEC_ID_WMV1;
break;
case 2:
id = CODEC_ID_WMV2;
break;
default:
/* ... */
break;
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
}
} else if (!strcmp(mimetype, "application/ogg")) {
id = CODEC_ID_VORBIS;
} else if (!strcmp(mimetype, "audio/mpeg")) {
if (gst_caps_has_property (caps, "layer")) {
gint layer = 0;
gst_caps_get_int (caps, "layer", &layer);
switch (layer) {
case 1:
case 2:
id = CODEC_ID_MP2;
break;
case 3:
id = CODEC_ID_MP3LAME;
break;
default:
/* ... */
break;
}
} else if (gst_caps_has_property (caps, "mpegversion")) {
gint mpegversion = 0;
gst_caps_get_int (caps, "mpegversion", &mpegversion);
if (mpegversion == 4) {
id = CODEC_ID_MPEG4AAC;
}
}
if (id != CODEC_ID_NONE) {
audio = TRUE;
}
} else if (!strcmp(mimetype, "audio/x-wma")) {
if (gst_caps_has_property (caps, "wmaversion")) {
gint wmaversion = 0;
gst_caps_get_int (caps, "wmaversion", &wmaversion);
switch (wmaversion) {
case 1:
id = CODEC_ID_WMAV1;
break;
case 2:
id = CODEC_ID_WMAV2;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
audio = TRUE;
}
} else if (!strcmp(mimetype, "audio/x-ac3")) {
id = CODEC_ID_AC3;
} else if (!strcmp(mimetype, "video/x-msmpeg")) {
if (gst_caps_has_property (caps, "msmpegversion")) {
gint msmpegversion = 0;
gst_caps_get_int (caps, "msmpegversion", &msmpegversion);
switch (msmpegversion) {
case 41:
id = CODEC_ID_MSMPEG4V1;
break;
case 42:
id = CODEC_ID_MSMPEG4V2;
break;
case 43:
id = CODEC_ID_MSMPEG4V3;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp(mimetype, "video/x-svq")) {
if (gst_caps_has_property (caps, "svqversion")) {
gint svqversion = 0;
gst_caps_get_int (caps, "svqversion", &svqversion);
switch (svqversion) {
case 1:
id = CODEC_ID_SVQ1;
break;
case 3:
id = CODEC_ID_SVQ3;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp (mimetype, "video/x-huffyuv")) {
id = CODEC_ID_HUFFYUV;
video = TRUE;
} else if (!strcmp (mimetype, "audio/x-mace")) {
if (gst_caps_has_property (caps, "maceversion")) {
gint maceversion;
gst_caps_get_int (caps, "maceversion", &maceversion);
switch (maceversion) {
case 3:
id = CODEC_ID_MACE3;
break;
case 6:
id = CODEC_ID_MACE6;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
audio = TRUE;
}
} else if (!strcmp (mimetype, "video/x-theora") ||
!strcmp (mimetype, "video/x-vp3")) {
id = CODEC_ID_VP3;
video = TRUE;
} else if (!strcmp (mimetype, "video/x-indeo")) {
if (gst_caps_has_property (caps, "indeoversion")) {
gint indeoversion = 0;
gst_caps_get_int (caps, "indeoversion", &indeoversion);
switch (indeoversion) {
case 3:
id = CODEC_ID_INDEO3;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp (mimetype, "video/x-divx")) {
if (gst_caps_has_property (caps, "divxversion")) {
gint divxversion = 0;
gst_caps_get_int (caps, "divxversion", &divxversion);
switch (divxversion) {
case 3:
id = CODEC_ID_MSMPEG4V3;
break;
case 4:
case 5:
id = CODEC_ID_MPEG4;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp (mimetype, "video/x-3ivx") ||
!strcmp (mimetype, "video/x-divx")) {
id = CODEC_ID_MPEG4;
video = TRUE;
} else if (!strcmp (mimetype, "video/x-ffv")) {
if (gst_caps_has_property (caps, "ffvversion")) {
gint ffvversion = 0;
gst_caps_get_int (caps, "ffvversion", &ffvversion);
switch (ffvversion) {
case 1:
id = CODEC_ID_FFV1;
break;
default:
/* ... */
break;
}
}
if (id != CODEC_ID_NONE) {
video = TRUE;
}
} else if (!strcmp (mimetype, "x-adpcm")) {
if (gst_caps_has_property (caps, "layout")) {
const gchar *layout = "";
gst_caps_get_string (caps, "layout", &layout);
if (!strcmp (layout, "quicktime")) {
id = CODEC_ID_ADPCM_IMA_QT;
} else if (!strcmp (layout, "microsoft")) {
id = CODEC_ID_ADPCM_MS;
} else if (!strcmp (layout, "wav")) {
id = CODEC_ID_ADPCM_IMA_WAV;
} else if (!strcmp (layout, "4xm")) {
id = CODEC_ID_ADPCM_4XM;
}
}
if (id != CODEC_ID_NONE) {
audio = TRUE;
}
} else if (!strcmp (mimetype, "video/x-4xm")) {
id = CODEC_ID_4XM;
video = TRUE;
}
/* TODO: realvideo/audio (well, we can't write them anyway) */
if (context != NULL) {
if (video == TRUE) {
gst_ffmpeg_caps_to_pixfmt (caps, context);
context->codec_type = CODEC_TYPE_VIDEO;
} else if (audio == TRUE) {
gst_ffmpeg_caps_to_smpfmt (caps, context);
context->codec_type = CODEC_TYPE_AUDIO;
}
context->codec_id = id;
}
if (id != CODEC_ID_NONE) {
char *str = g_strdup_printf("The id=%d belongs to this caps", id);
gst_caps_debug(caps, str);
g_free(str);
}
return id;
}
Reference in a new issue Copy permalink