| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
movflags were being overwritten, silly error on my part.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6386 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
| |
PGS subtitles are terminated by an explicit "end of sub" subtitle in
the stream. The duration is not used (or known in advance). So every
PGS was getting the default 10s duration assigned. This made the
overall duration of the stream appear longer than it really is.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6384 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
much of this work was borrowed from ccextractor which is GPL v2
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6383 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6376 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
the style was not getting flushed if there were no style changes
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6375 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
mergeSubtitle was originally written for pure text subs. But now the
subtitle buffers have extra ssa formatting that needs different treatment.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6374 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
We don't need the empty subtitle buffers.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6373 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
scr_offset was not accounted for in stop time
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6372 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The GUI will precompute both width and height and specify both on the
command line. But then the CLI recomputes width based on height which
can result in the width being changed.
So, in non-anamorphic mode, disable keep_aspect so that width and heigh
are both kept. And in other anamorphic modes, use custom ana when computing
dimensions so that both width and height are kept (i.e. recompute par).
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6369 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6366 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6363 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
Allocate extra byte so extradata pointer is not NULL.
Fix logging of malloc failures so we can see what failed.
Fix test for existence of extradata, check length NOT bytes. bytes is never NULL
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6327 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
If the job aborts right away, an uninitialized buffer is closed and crash!
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6326 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
...and enable x265 by default
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6309 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
It appears some BDs have invalid chapter durations, see
https://forum.handbrake.fr/viewtopic.php?f=12&t=30479
This compares chapter start time deltas and title duration to chapter
duration to make best guess corrections.
Note that chapter events generated by libbluray are based on the chapter
start times. So the start time should be trusted more than the duration.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6292 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
mp4 audio track titles were not being applied by libavformat
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6270 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
| |
A couple common mp4 taggers (mp3Tag and iTunes) fail to tag files
that have the Nero chapter atom (chpl). This atom was added when
we switched to libavformat for muxing. I haven't found any players
yet that rely on this for chapters. If we ever find any, we can
consider adding an option somewhere to enable chpl.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6269 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6248 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6247 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6246 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This new function has a couple advantages over the old one (which we
should phase out).
It does not require hb_job_t as a parameter, instead it uses hb_ui_geometry_t
which is a smaller and simpler struct. The entire job struct is overkill as
input to this function.
It returns an hb_image_t that fully describes the returned image instead of
just a uint8_t array. The caller does not have to make assumptions about image
size, line stide, or pixel format since hb_image_t specifies these things.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6242 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Move nlmeans preset&tune parsing from cli to libhb
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6239 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Solves: https://forum.handbrake.fr/viewtopic.php?f=10&t=28074&sid=68d8c5bfffd7081625bcd56fb7c4dad1
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6236 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Addresses https://forum.handbrake.fr/viewtopic.php?f=12&t=30307&p=140045#p139847
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6235 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
Fix bug where non-anamorphic mode ignores modulus
Change default modulus to 2
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6234 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
In some cases, initial data when in p-to-p mode causes libav decoder
initialization to fail. This only happens when multi-threaded encoding
is enabled.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6229 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
NO_IDR and SCAN_COMPLETE values overlapped
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6226 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
Fixes a typo (1/27 instead of 1/127) that caused edgeboost to match on nearly everything, and tweaks this number slightly.
Adds a post-processing step to handle marked pixels having too few marked neighboring pixels as false positives.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6224 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6223 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
bd_seek and the new m2ts_filter in libbluray have a bad interaction.
So use bd_seek_time instead.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6221 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
Non-local means averages multiple patches of similar pixels together, preserving similarities and attenuating variance (usually noise). This is typically more effective than lowpass and more faithfully restores the appearance of structure and detail found in the original source, especially in the high frequency range. Parameters for origin patch weight tuning and pre-filtering further improve on the original algorithm.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6216 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
| |
macui: Simplifies code in Controller.m and fixes a few bugs
macui: Fixes "Windows 8 Phone" preset aspect ratio problem
macui: round up to next mod boundary when limiting to title dimensions - crop
libhb: Eliminates hb_fix_aspect and hb_set_size functions.
libhb: Makes hb_validate_size a passthru to hb_set_anamorphic_size
cli: eliminage use of hb_validate_size
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6215 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
| |
We were packaging multiple subtitle packets with the same pts and passing
down to rendersub which would then throw away all but the last sub with
a given pts. So put all AVSubtitleRects into the same hb_buffer_t.
A patch to libav is also required to fix the multi-object problem.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6211 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
simplifies the code and makes it work better
also updates preview window size calculations
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6209 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
... and probably other formats that use a palette.
The palette is stored in AVPacket side data which we did not read. So
read the side data and stash it in hb_buffer_t so that it can be used
later by the decoder.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6207 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6206 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
after last change, CCs were getting positioned too far into the middle of
the image.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6204 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6203 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
| |
First problem was pre-existing, but hidden. Interleaving was severely
broken because we relied on buf->s.stop to determin interleave points.
stop is not guaranteed to be set.
Other problems with the new bit vector code caused failure to complete
muxing.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6187 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Miscellaneous bugfixes and performance improvements.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6185 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Fixes this bug report https://forum.handbrake.fr/viewtopic.php?f=12&t=30032
Split MPEG demuxer into TS demuxer and PS demuxer to allow using different
PCR/SCR delta tolerance in each scenario.
We were using a PCR/SCR delta tolerance that was a bit of a compromise
between the max allowed by the specification for TS vs PS streams. But
the spec for TS PCR is 100ms and the spec for PS SCR is 700ms, so we
risked false detection of discontinuities in PS. In PS streams that pack
multiple frames into one PES packet, the detection of a discontinuity
causes recalculation of an scr offset which jumps around wildly.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6183 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
The limit was due to using an uint32_t as a bit vector. This solves
the limitation by introducing an hb_bitvec_t struct for bit vectors
of unlimited size.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6181 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
Checking whether we should attach a chapter to
the output buffer is not enough, we actually
have to attach it too.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6176 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
If the CC was in the crop area at the top of the image, it's position
was not moved properly.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6172 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Thanks to Matthew Harvey for this patch
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6165 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
| |
Add support for font color to tx3g.
Allow more than one style flag at time in tx3g.
Add positioning support to CC subs
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6163 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6160 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, the DTS provided by MSDK was used
if it was available and the frame rate was constant.
Our own DTS is indentical to MSDK's (when the
frame rate is constant), but MSDK will sometimes
provide 1 or 2 bogus DTS over the duration of a
file, causing the encode to abort (lavf is picky).
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6158 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6157 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
unit bitstreams (H.264, HEVC) to and from
Annex B format.
Only used by QSV for now, but could
always come in handy down the road.
May also fix an issue introduced
by the recent QSV cleanup patchset.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@6156 b64f7644-9d1e-0410-96f1-a4d463321fa5
|