Commit graph

9243 commits

Author SHA1 Message Date
Mark Nauwelaerts
857efb4917 qtmux: refactor configuring and sending of moov
Based on patch by Marc-André Lureau <mlureau@flumotion.com>
2010-11-19 19:21:37 +01:00
Mark Nauwelaerts
e361a0b3ce qtmux: refactor extra top-level atom handling
Also check a bit more for possible errors, and free proper items in such case.
2010-11-19 19:21:37 +01:00
Mark Nauwelaerts
261df7636e qtmux: refactor slightly using buffer helper 2010-11-19 19:21:37 +01:00
Mark Nauwelaerts
3fea86fa60 qtmux: fix misinforming comment 2010-11-19 19:21:37 +01:00
Mark Nauwelaerts
1a279a9466 qtmux: delegate mvex handling to atoms
... which keeps qtmux simpler.
2010-11-19 19:21:36 +01:00
Marc-André Lureau
78d2d8598d qtmux: add mvex/trex in header if fragmented
One "trex" is added per "trak". We don't support default values,
but the "trex" box is mandatory.
2010-11-19 19:21:36 +01:00
Marc-André Lureau
84d8b33d1a qtmux: add a couple of fourcc for fragmented mp4 2010-11-19 19:21:36 +01:00
Mark Nauwelaerts
7e4c96359f qtmux: avoid removing temp file when error occurred 2010-11-19 19:21:36 +01:00
Marc-André Lureau
401fb0a09b qtmux: truncate buffer file after each send 2010-11-19 19:21:36 +01:00
Marc-André Lureau
055a1a3b0a qtmux: remove temp file when reset/finalize 2010-11-19 19:21:36 +01:00
Arun Raghavan
8ca66f336e jpegformat: Push tags after setting srcpad caps
This patch defers emission of tag events till caps are set on the source
pad of jpegparse, so that these tags can be seen downstream.

https://bugzilla.gnome.org/show_bug.cgi?id=627211
2010-11-19 14:57:59 -03:00
Alessandro Decina
b005ed2197 mpegtsparse: fix compiler warning 2010-11-19 18:51:32 +01:00
Andoni Morales Alastruey
fb9234de78 mpegtsparse: fix thinko 2010-11-19 18:33:10 +01:00
Andoni Morales Alastruey
b11f737096 mpegtsparse: fix handling of TOT and TDT sections. Fixes #635281.
TDT and TOT sections, with PID=0x14, doesn't extend to several packets
and the section filter is not needed here and shouldn't be used at all
for these tables because the have a different structure.
For example, TDT tables were not parsed for odd hours because this bit
is the 'current_next_indicator' bit for the other sections, and the table
was discarded.
2010-11-19 18:11:50 +01:00
Andoni Morales Alastruey
2611b12970 mpegtsparse: don't calculate the CRC for TOT tables. Fixes #635281.
TOT tables, with table_id=0x73, don't have a CRC, so don't calcute it
2010-11-19 18:10:47 +01:00
Andoni Morales Alastruey
7c43c2fa06 dvbbasebin: Add TDT to the initial pids filter for dvbsrc. Fixes #635200. 2010-11-19 18:04:12 +01:00
Víctor Manuel Jáquez Leal
e5728ebcb1 jpegparse: add gst_jpeg_parse_skip_marker ()
https://bugzilla.gnome.org/show_bug.cgi?id=626618
2010-11-19 16:30:18 +02:00
Víctor Manuel Jáquez Leal
285c32e01f jpegparse: use byte reader accessors
https://bugzilla.gnome.org/show_bug.cgi?id=626618
2010-11-19 16:27:24 +02:00
Víctor Manuel Jáquez Leal
e11fdf67f0 jpegparse: inline gst_jpeg_parse_sof ()
No functional changes (hopefully).

https://bugzilla.gnome.org/show_bug.cgi?id=626618
2010-11-19 16:25:39 +02:00
Víctor Manuel Jáquez Leal
d0dc44db02 jpegparse: fix typo
https://bugzilla.gnome.org/show_bug.cgi?id=626618
2010-11-19 16:25:35 +02:00
Aleksey Lim
d66ef4dea8 camerabin: Do not use audio clock after stopping video capture
Adda provide clock function to camerabin to make it not
provide the audio clock of the record bin when no video
recording is happening

Fixes #613379
2010-11-18 10:58:06 -03:00
Mark Nauwelaerts
68dca2e704 baseparse: use correct offset when adding index entry
... bearing in mind that BUFFER_OFFSET is media specific and may not
reflect the basic offset after having been parsed.
2010-11-17 16:24:13 +01:00
Mark Nauwelaerts
9b6439a36c baseparse: enhancements for timestamp marked framed formats
That is, as such formats allow subclass to extract position from frame,
it is possible to extract duration (if not otherwise provided)
from (near) last frame, and a seek can fairly accurately target the required
position.

Fixes #631389.
2010-11-17 15:36:41 +01:00
Mark Nauwelaerts
08cab00b8b baseparse: refactor frame scanning peformed by _loop 2010-11-17 15:36:38 +01:00
Mark Nauwelaerts
587106f1cf baseparse: slightly optimize sending of pending newsegment events 2010-11-17 15:36:36 +01:00
Mark Nauwelaerts
f82f330537 baseparse: minor fixes and enhancements
Arrange for upstream as well as downstream flushing when seeking.
Also determine upstream size as well as seekability.  Adjust some comments
to reality and employ debug statement in proper order.
2010-11-17 15:36:23 +01:00
Mark Nauwelaerts
b65da8e775 aacparse: minor cleanups 2010-11-17 15:36:20 +01:00
Mark Nauwelaerts
0fdc68cd46 aacparse: fix regression in ADIF src caps setting 2010-11-17 15:36:18 +01:00
Mark Nauwelaerts
36cbbd94c6 flacparse: parse seektable
Fixes #631389 (partially).
2010-11-17 15:36:02 +01:00
Mark Nauwelaerts
551f9707fa flacparse: minor refactor and enable default baseparse segment clipping 2010-11-17 15:35:59 +01:00
Ole André Vadla Ravnås
abdb30c567 celvideosrc: fix nasty deadlock
We cannot call any CMBufferQueue functions while holding the lock that
our callback also depends on. So now we make use of CMBufferQueue's
trigger API in order to get notified when the queue has data.
2010-11-12 00:44:42 +01:00
Ole André Vadla Ravnås
de76e9fdb6 cmapi: cover CMBufferQueue's trigger API 2010-11-12 00:44:42 +01:00
Ole André Vadla Ravnås
e4db4257d5 celvideosrc: unlock should also notify
Otherwise we depend on being woken up by the CoreMedia queue getting
its next buffer.
2010-11-12 00:44:42 +01:00
Ole André Vadla Ravnås
ebd2a6d6e9 applemedia: don't push synchronously from callback
The codec that called us might be holding locks to shared resources, so
we should never push downstream from within its buffer callback.

Note that a GstBufferList is not used here because we need to preserve
the buffer metadata held by our GstBuffer subclasses.
2010-11-11 00:22:31 +01:00
Ole André Vadla Ravnås
6b63954f63 vtenc: bump H.264 level from 1.3 to 3.0
This should be turned into a property.
2010-11-11 00:22:31 +01:00
Ole André Vadla Ravnås
6e909dc474 celvideosrc: add basic device selection support 2010-11-11 00:22:31 +01:00
Ole André Vadla Ravnås
2382f10b98 cvapi: add wrapper for IO surface access
To be used in a future video sink for zero-copy rendering.
2010-11-11 00:22:31 +01:00
Ole André Vadla Ravnås
420943972c celapi: add capture port type constants 2010-11-11 00:22:31 +01:00
Mart Raudsepp
c364f2b46b assrender: Remove unused src_stride variable from blit functions 2010-11-09 23:36:59 +00:00
Thiago Santos
a64e0ff930 win32: Adding photography interface .def file
Adds check-exports to -bad to have photography interface
.def file created

Fixes #578629
2010-11-09 18:32:07 -03:00
Mark Nauwelaerts
896c4b71af mpegaudioparse: fix silly leak in _reset 2010-11-09 20:52:10 +01:00
Ole André Vadla Ravnås
d10e029f51 applemedia: biplanar is actually NV12, not I420
D'oh!
2010-11-08 23:58:25 +01:00
Ole André Vadla Ravnås
d865feca82 vtenc: remove keyframe enforcement workaround
Was only needed for old iOS where the H.264 encoder didn't support
kVTEncodeFrameOptionKey_ForceKeyFrame.
2010-11-08 23:58:25 +01:00
Ole André Vadla Ravnås
abba3a4126 cvapi: add a few extra bits of wrapping 2010-11-08 23:58:25 +01:00
Ole André Vadla Ravnås
b4654f537f applemedia: lock CVPixelBuffer read-only
As per Apple's docs, this may improve performance by avoiding redundant
invalidations of existing caches of the buffer contents.
2010-11-08 23:58:24 +01:00
Lasse Laukkanen
6867ec5627 examples: camerabin: Remove gstring leak
Simplify gst-camerabin-test by removing GString and replacing
with gchar for the filename handling. Also removes a leak of
the GString data.
2010-11-08 11:07:17 -03:00
Stefan Kost
fedcff79d7 rsvgdec: avoid reffing the element in chain 2010-11-08 16:02:07 +02:00
Stefan Kost
9ddeba2407 rsvgdec: add some minimal logging to track what it is doing 2010-11-08 16:02:07 +02:00
Ole André Vadla Ravnås
6428d306ea celvideosrc: put planar first in template caps 2010-11-08 14:02:50 +01:00
Ole André Vadla Ravnås
c2672f71cd applemedia: improve vtenc/vtdec performance
Profiling of H.264 encode and decode revealed that conversions
between packed and planar were happening behind the scenes.

Hence we now choose I420 instead of YUY2.
2010-11-08 14:02:50 +01:00