| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
| |
simplify job initialization sequence, clean up code, and document
dependencies in the sequence better.
Make hb_add set job->sequence_id. It is no longer necessary for the
frontend to do this. If the frontend needs the sequence_id, it is
returned by hb_add().
Clean up use of interjob. do_job() now uses sequence_id to detect when
a new sequence of related jobs is running and automatically clears
interjob.
|
| |
|
|
|
|
|
| |
libavcodec/mpeg12dec.c expects input buffers to be zero padded to 32 bit
alignment. If not zero padded, it can get caught in an infinite loop.
|
|
|
|
|
| |
AV_NOPTS_VALUE was getting modified by subtracting an offset. This lead
to garbage start values reaching the muxer and causing hangs.
|
|
|
|
|
| |
libav just deprecated AVPicture and all av_picture_*/avpicture_*
functions.
|
|
|
|
| |
Per BradleyS suggestion
|
|
|
|
|
|
| |
In both hb_set_anamorphic_size and hb_get_preview, check that image
dimensions are > 32x32 and < 10240x10240. This should help to prevent
crashes with unusual dimensions settings.
|
|
|
|
|
|
| |
Try to avoid failures to initialize sws context by setting minimum
dimensions. And if initialization does fail, exit gracefully instead of
crashing in sws_scale.
|
|
|
|
|
|
| |
When "keep aspect" is unset in custom anamorphic, it was making very
uintuitive changes to PAR. This simplifies the code and makes the
behavior more sane.
|
|
|
|
|
|
|
|
|
|
| |
Fixes https://forum.handbrake.fr/viewtopic.php?f=12&t=33345
Some players expect each packet to start on an mp3 frame header. Our
mp3lame encoder did not ensure this and resulted in failure to play
audio on these players.
libav already has the necessary code to parse headers and
accumulate a full frame of data, so use it.
|
|
|
|
|
|
| |
windows gui stores all presets in now.
This should allow the CLI to read GUI presets now.
|
| |
|
| |
|
| |
|
|\
| |
| | |
grayscale: make it a real filter
|
| |
| |
| |
| |
| | |
It only worked properly with the x264 encoder. Now it works with all
encoders.
|
| | |
|
| | |
|
|/
|
|
|
|
|
| |
When an encode is cancelled, cached mux buffers are not currently
written to the output file. Since we don't start writing output till
50MB of data has been cached, this can result in no data being written
when an encode is cancelled early.
|
|
|
|
| |
This is no longer a problem with new versions of iOS and QuickTime
|
|
|
|
|
|
| |
This was originally done because > 2 lines is not supported in mp4 timed
text. But this line combining is now performed only when necessary in
muxavformat.
|
| |
|
|
|
|
|
| |
This information is useful to the frontends in creation of preset
backups.
|
|
|
|
|
| |
I.e. use a dropdown to select between the filters instead of a radio
button.
|
|
|
|
| |
prevent sleep on OS X.
|
|
|
|
| |
We need to rescan when the number of previews changes.
|
|
|
|
| |
var needed
|
| |
|
|
|
|
|
| |
This brings together several independent implementations of a simple
buffer list manager.
|
| |
|
|
|
|
|
| |
The video frame subtitles are meant to be rendered to may not match the
actual video we are encoding. So scale to match as needed.
|
| |
|
|
|
|
| |
It was borken in 293083e5
|
| |
|
| |
|
|
|
|
| |
gcc didn't like the way variables were being initialized.
|
|
|
|
|
|
|
|
| |
Allows for arbitrary frame rates between 1 and 1000 fps.
Adds min/max frame rates to CLI help.
Removes hardcoded instances of the internal clock rate in favor of hb_video_framerate_get_limits().
Unfortunately, much of the codebase generally refers to clock rate as frame rate, so a little extra care is still necessary going forward.
|
| |
|
| |
|
|
|
|
|
|
| |
Migrate from svn to git. Since our build system automatically generates
version info from the svn repo, this needs to change when we move the
repo to git.
|
|
|
|
|
|
|
|
| |
This will result in no video in the output file, but in such cases, this
might actually be desirable. Scaling to such small dimensions is often
used simply to re-encode audio.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7412 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
Make the nlmeans default (when no string supplied) match "medium"
preset. And make the CLI default the medium preset.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7402 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
It was not continuing if it failed to find the preset in the first
folder
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7401 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7400 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7397 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
|
| |
In non-anamorphic and custom-anamorphic, if keep-aspect is not set,
don't adjust dimensions to fix aspect when applying max width or height.
If the user is telling us to distort the image, do it.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7362 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
This avoids issues with OpenCL buffer mapping.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7343 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
OpenCL buffer mapping was not disabled during cleanup.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7342 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
|
|
|
| |
While it worked fine w/HSW, it led to some false negatives on IVB hardware.
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7334 b64f7644-9d1e-0410-96f1-a4d463321fa5
|
|
|
|
| |
git-svn-id: svn://svn.handbrake.fr/HandBrake/trunk@7332 b64f7644-9d1e-0410-96f1-a4d463321fa5
|