jstebbins [Fri, 4 Jun 2010 16:27:15 +0000 (16:27 +0000)]
MacGui: Add flexible dylib path manipulation so that we can handle external dylibs more easily
At startup, add any extra dylib paths to DYLD_FALLBACK_LIBRARY_PATH. This is the last path
searched by the system for dylibs and we add our paths to the end of it's list, so this
will never override any system libs or paths the user has explicitly set.
Since applications read the environment once at startup, these changes don't take effect
until we restart with execv. In order to avoid a possible exec bomb, we add a parameter to
the argv list to prevent any further exec's.
Note that this causes a minor glitch when running under gdb. The execv triggers a trap. You
just have to 'continue'.
dynaflash [Wed, 2 Jun 2010 16:49:18 +0000 (16:49 +0000)]
MacGui: Set the hud control panels origin based on the size of the actual preview window anchored at the bottom instead of setting it at the middle of the window.
- Should not block the actual preview as much.
- Move the origin setting code from -SetTitle into -displayPreview so that it is recalculated any time a new preview is generated, which includes changing encode size.
jstebbins [Sun, 30 May 2010 17:26:03 +0000 (17:26 +0000)]
Improve left/right dplii surround separation
This analysis and measurement shows that a difference of 4.8db between
Ls and Rs in Lt and Rt provides better separation than 3db
http://forum.doom9.org/showthread.php?t=57988
sr55 [Sun, 30 May 2010 13:02:23 +0000 (13:02 +0000)]
WinGui:
- Added resharper 5 file for the 2008 project.
- Unlocked VobSub in Mp4 to allow multiple tracks via passthru. Also fixed a bug with srt import handling.
jstebbins [Sat, 29 May 2010 22:16:48 +0000 (22:16 +0000)]
LinGui: tweak the ui to accommodate universal subtitle input
don't show burned or forced fields for non-vobsubs
show subtitle type in the subtitle track combo box
jstebbins [Wed, 26 May 2010 20:39:52 +0000 (20:39 +0000)]
fix a problem with point-to-point hanging when searching for start point
in reader, the timestamps were not being correctly adjusted for scr offset
before comparing to start time. This could cause an early start in reader.
Then in sync, syncAudioWork stalled until the correct start of video was
found, causing the audio fifo to fill and stall the whole pipeline.
dynaflash [Tue, 25 May 2010 17:09:10 +0000 (17:09 +0000)]
MacGui: Enable soft vobsubs in mp4
- Initial patch by Rodeo. Thanks Rodeo!
- Enables soft vobsubsin macgui as per http://trac.handbrake.fr/changeset/3325
- Due to the fact that currently no quicktime based playback devices (yes, including computers and iphone/ipod/ipad) supports soft vobsubs, "Burned In" is on by default.
- Currently soft vobsubs are shown as supported by VLC (Lunettes) and Mplayer on the mac platform.
jstebbins [Mon, 24 May 2010 21:53:48 +0000 (21:53 +0000)]
add support for Nero vobsubs in mp4
Note that these do not work with any apple products that I know of. Perian
might be able to do something with them. MPlayer and VLC both grok nero
vobsubs.
jstebbins [Mon, 24 May 2010 18:15:40 +0000 (18:15 +0000)]
LinGui: add option to completely disable update checking
configure --disable-gtk-update-checks removes the code that does update checks
including the webkit library used for displaying the appcast. this is
useful when packaging for a repository that has it's own update mechanism.
sr55 [Sun, 23 May 2010 17:32:45 +0000 (17:32 +0000)]
WinGui:
- Improvements to the Installer. (Both Nightly and Primary nsi scripts)
* Check for .NET 3.5 SP1 and if it doesn't exist, download and run the web Installer
* Uninstall the current version before trying to install the new version.
jstebbins [Sun, 23 May 2010 15:38:21 +0000 (15:38 +0000)]
add ability to probe for existance of headers and libraries to configure.py
use this to check for system versions of bz2 z pthreadGC2 iberty and dl when
building with mingw. solves a problem with conflicts when building on fedora with
a full standard install of their mingw packages.
dynaflash [Thu, 20 May 2010 15:28:27 +0000 (15:28 +0000)]
Support for reading VOB subtitle tracks from file inputs initital implementation.
- Patch by davidfstr, Nice Work!
- Adds support for reading VOB subtitle tracks from file inputs.
Tested with:
- MKV VOB -> MKV VOB passthru.
- MKV VOB -> MKV VOB burned in.
VOB subtitle palette moved from per-title to per-track.
Discussion leading up to commit can be referenced here: http://forum.handbrake.fr/viewtopic.php?f=4&t=16267
sr55 [Wed, 19 May 2010 20:13:53 +0000 (20:13 +0000)]
WinGui:
- In-Gui encode status is now own by default.
- Added a progress bar and tidied up the appearance of the status text.
- Tweaked the preview window design.
dynaflash [Tue, 18 May 2010 19:36:55 +0000 (19:36 +0000)]
x264 bump from r1583 to r1592
- Faster cabac
- Various bugfixes and enhancements
- http://git.videolan.org/gitweb.cgi?p=x264.git;a=commit;h=95aacc4fb0ed4477a8c8b7cbf3f8c89229b03dc7
sr55 [Tue, 18 May 2010 18:34:59 +0000 (18:34 +0000)]
WinGui:
- Changed the ActivityWindow to be event driven which should make it work a bit better. Please report any bugs / issues you see with this window.
sr55 [Tue, 18 May 2010 16:23:25 +0000 (16:23 +0000)]
WinGui:
- Added an exception window which allows the exceptions to be copied by the user rather than using a message box. (Long overdue). Moved some of the most common exceptions over to using it.
dynaflash [Tue, 18 May 2010 16:22:00 +0000 (16:22 +0000)]
MacGui: Live Preview HUD Overlay playback controller initial implementation
- Fixes current issue where a live preview with the display res larger than storage res would not show the built in QTKit control bar.
- Replaces the stock QTKit movie controller.
- Adds playback time code information to HUD display.
- Allows frame by frame scrubbing via keyboard arrow keys.
-- Methods exist for frame by frame step through but buttons are not on the HUD overlay yet.
Known Bugs:
- If movie is playing, first click on the scrubber bar can cause the movie to jump to beginning or end unless scrubber is held and dragged.
- Probably some others I haven't seen yet.
To Do (not that it ever will get done):
- Replace the current crappy stock buttons with some decent ones.
- Display accurate frame number hopefully though QTKit offers little documentation to get that exact info.
- Add a volume slider and remember the volume setting in the prefs.
- Add a subtitle control widget hopefully so that users can see what the subtitles would look like.
Other small changes include adding some tooltips to the hud overlay controller widgets to hopefully make it a bit more intuitive.
sr55 [Tue, 18 May 2010 15:19:42 +0000 (15:19 +0000)]
WinGui:
- Switch the Audio Panel over to a DataGridView for more flexibility in possible upcoming changes.
- Few UI tweaks to other panels to make look / behave similarly.
jstebbins [Sun, 16 May 2010 23:12:12 +0000 (23:12 +0000)]
patch ffmpeg avi seek problem
when seeking to a timestamp, ffmpeg seeks to the position in the primary
stream where that timestamp starts. Then seeks to the next interleaved packet
for all other streams. When seeking back to the beginning of the file, this
means the first packet of some streams are missed if those streams start
before the video stream in the file.
jstebbins [Thu, 13 May 2010 14:48:18 +0000 (14:48 +0000)]
allow mono mixdown with lame encoder
enable mono mode in lame encoder when mixdown is mono.
use lame_encode_buffer_float instead of lame_encode_buffer_interleave. this
eliminates the clipping issue in lame without reducing the level of the input.
sr55 [Sun, 9 May 2010 16:56:56 +0000 (16:56 +0000)]
WinGui:
- Fixed Exception on restoring window from system tray
- Fixed Exception when trying to edit an item on the queue without first selecting an icon
- Replaced old folder icon on the source menu with the newer windows equiv
dynaflash [Thu, 6 May 2010 18:30:32 +0000 (18:30 +0000)]
MacGui: Chapter Title import/export functionality
- Saves the chapter list to a .csv file which is compatible with the cli and wingui.
- Note: comma's in the chapter name are escaped with a "\" to maintain cli compatibility then if necessary stripped upon re import into macgui.
- Original patch by borgclone and added to by realityking. Thanks!
- Feature development referenced here http://forum.handbrake.fr/viewtopic.php?f=4&t=4146&start=0
WinGui:
- Strip any illegal characters from the autoname path
- Display build number in program window title. Also added a hack to force this to display the correct number.
- Imported some settings for Stylecop for resharper 5
LinGui: drop dependency on hal, use udev instead
this means support for older linux versions is being dropped.
ubuntu 9.10 and above and fedora 12 and above are known to work.
MacGui: reset hb_filter_decomb after use, so switching between default decomb and a custom decomb string does cause a garbage string. Thanks j45 for the patch!
bump ffmpeg to rev 22950
offset ffmpeg seeks by value of initial timestamp since it can be non-zero
catch pix fmt that is unsupported by swscale, log it, skip the track
use new ffmpeg avg_frame_rate for more accurate framerate estimate
fix qdm2 audio decoding
it seems ffmpeg wants to be passed the same buffer repeatedly
while decoding this audio type. we were exiting if ffmpeg said it
consumed 0 bytes. Now we continue to feed the same buffer when
this happens. I added a loop limit to protect against an hypothetical
ffmpeg bug that would never consume anything. I wonder if any other
codecs behave this way *scratches head in bewilderment*
fix decoding of qtrle codec and possibly others
we rely on ffmpeg calling context->get_buffer, which we override to
point to our get_frame_buf(), to stuff our pts values into frames it
is decoding. but some decoders call context->reget_buffer
instead. So I added an override for this callback as well.
make sure HB_STATE_WORKING always gets set for a job
very short video, or video that has no valid frames, can cause UpdateState
in sync.c to never be called. This is the only way the sequence_id would
ever be set, so when the ui sees HB_STATE_WORKDONE, it could not look up
the job in it's queue based on the sequence_id causing it to never be
marked as complete.
add test to ensure that ffmpeg's reported audio layout agrees with channel count
a crash report from a user leads me to believe ffmpeg is giving us
a channel layout that has a different number of channels than the
channel count it tells us.
fix LATM aac processing
somewhere in the series of patch changes, setting the sample_fmt got
lost. we were being told SAMPLE_FMT_NONE by ffmpeg for latm