Fix int overflow that can happen when computing PAR
If the source has large non-reduced PAR values, our computed value
was overflowing an int. Compute it in an int64_t then reduce it.
Also, keep num and den below 65535. Larger values just aren't really
significant and will cause more overflow issues.
MacGui: implement Frame Rate Mode as per the Lingui.
- Removes the current Peak Framerate checkbox and replaces it with appropriate Radio Buttons depending on if Same as Source or a specific framerate is chosen in the Frame Rate popup widget.
- Gets the macgui up to speed with https://trac.handbrake.fr/changeset/3770
WinGui:
- Added "Advanced" audio options floating popup window.
* This exposes a new Audio Gain Control (-20 to +20 dB)
- Added support for (--gain=x,y,z) to the preset Loader.
WinGui:
- Add "Send to" option. This will pass the encoded output file name of each completed job on the queue to a 3rd party application with optional arguments passed.
Change HB.version from "svnXXXX" to "XXXXsvn"
This fixes an issue with packaging for Ubuntu 11.04. The packaging tool
now requires that version numbers start with a digit. They used to just
issue a warning.
Each clip of a BD are allowed to have different audios if the clip
does not have a seamless connection to the previous clip. Most titles
are a series of seamless clips that all have the exact same audio. But
I found some that have a final non-seamless clip that has completely
different audios and broke the old algorithm.
New algorithm, look at each clip and count the number of other clips have
the same audio. Use the clip that has the most matches.
Fix a deadlock that can happen with p-to-p encoding
With p-to-p, the audio sync thread waits for the video sync thread to
reach the designated start point. There is a possibility that the video
decoder will drop so many frames that the audio sync fifo fills before
any frames reach the video sync thread. When this happens, drop some
audio to unplug the pipeline.
Also, to make this less likely to happen, start sending data to the video
decoder 2 seconds before the actual desired start point. This will allow
the decoder to find an initial i-frame before the audio stalls since the
audio sync thread drops any audio that is before the designated start point.
A side effect of this is our start time now more accurate since the decoder
is only dropping frames before the start point instead of after.
WinGui:
- Fix a culture issue with user settings (x264)
- Made the Preset loader more strict with what settings it can and can't load with regards to video settings.
These boundaries are always discontinuities. But sometimes we were
not detecting them as such and would drop frames. So set a flag
in the buffer when libbluray tells us a new clip is starting and
use that to trigger computation of a new scr offset.
WinGui: AudioPanel refactored.
- Switch the Audio panel to use DataBinding to a BindingList collection leading to much cleaner code.
- Added gain support (--gain=3,4) to the services library. (UI has not been done yet)
Use bibiliographic form of ISO-639-2 lang codes in muxmkv
The Mkvtoolnix developer claims that MKV only allows the bibliographic form
of ISO-639-2 lang codes: https://www.bunkus.org/bugzilla/show_bug.cgi?id=598
http://matroska.org/technical/specs/index.html#languages seems to confirm this.
Change internal audio representation range
...from float [-32768...32767] to float [-1.0...1.0]
Using the range [-1.0..1.0] requires fewer translations of the range for our
various encoders and decoders. This also gets rid of a hacky
translation from float to int to float in decavcodec audio decoding.
Fix dts downmix when the output encoder is lame
libdca downmix is broken if you ask for dolby and DCA_ADJUST_LEVEL.
Since we fixed the clipping problem that DCA_ADJUST_LEVEL is used for
with changeset 3294, we can just disable this.
LinGui: add audio gain adjustment and create advanced audio dialog.
Gain, DRC, and Sample Rate are now in an advanced audio dialog.
This makes the audio tab less crowded and makes room for future
improvements.
Add audio gain adjustment to libhb and CLI
New CLI option is --gain <float>. Value is measured in dB. Negative values are
quieter, positive values are louder.
Start mkv clusters with a video keyframe whenever possible
libmkv now starts a new cluster whever there is a video keyframe if
the new cluster would not be closer than 2 seconds to the previous cluster.
Fix the volume output level of our dts decoder
dcadec returns samples that have values in the range -1.0 to 1.0.
We need these to be converted to the range -32768 to 32767. For some
reason, decdca was scaling by 16768 instead of 32767. This has been
like this since dts support was initially added by maurj.
Make logging of audio and subtitle id's more consistant
1) whenever we log audio->id or subtitle->id using hex formatting, precede
the hex with 0x (which was already done in some places but not others)
2) format audio->id as hex instead of decimal in sync.c (makes it much easier
to see which track "went backwards" or had silence added to it, checking
the job configuration logged by work.c)
Shorten instead of dropping overlapping subtitles in mp4
If the non-overlaping remainder is greater than 0.5 seconds, shorten
the subtitle instead of completely dropping it.
Flush ffmpeg codec buffers after every seek.
According to several e-mails I've read on ffmpeg-devel, avcodec_flush_buffers
should be called after any seek. It appears this is even more critical to do
when using frame based multi-threading. I don't see any immediate difference
in functionality by adding this, but it may prevent surprises in the future.
Forome reason, frames that are tagged as recovery points in many BD h.264
streams do not result in complete frames when decoded. Pushing 2 extra
frames through the decoder seems to always fix this. This patch extends
something I was already doing when generating previews from a BD structure.
This just applies the same logic to ffmpeg streams that have h.264 video.
jstebbins [Thu, 31 Mar 2011 00:05:23 +0000 (00:05 +0000)]
True SSA passthru for MKV.
We were converting SSA to UTF8 subs which looses a lot of formatting.
Now we pass through the ssa unmodified and add all fonts as attachments
to the mkv.
davidfstr [Wed, 30 Mar 2011 02:50:42 +0000 (02:50 +0000)]
MP4 muxer ignores all but the first of overlapping UTF-8 subtitles.
This is necessary because MP4 does not support overlapping subtitles. Attempting to use overlapping subtitles causes the display of subsequent subtitles to be delayed incorrectly.
Subsequent patches may merge UTF-8 subtitles (upstream) so that this case does not occur.
jstebbins [Tue, 29 Mar 2011 22:43:07 +0000 (22:43 +0000)]
Fix chapter passthru problems with some files.
If the video stream is not the first track in the file, chapters were lost.
During scan, we identify which track is video and stash this in title.
While reading, when a chapter is found we want to tag the next video buffer.
But the video track id stored in the title was not being applied when
opening the file for reading, so the chapter mark always went on track id 0.
dynaflash [Tue, 29 Mar 2011 17:55:25 +0000 (17:55 +0000)]
MacGui: Never disable the chapter markers checkbox.
Patch by Rodeo ... Thanks!
- Discussed here https://reviews.handbrake.fr/r/58/
This allows users to enable/disable chapter markers regardless of whether the source has chapters, which is very useful when saving a preset.
This will not cause chapter markers to be inserted if we only have one chapter, as this is sanitized twice:
- once in macosx/Controller.m, in createQueueFileItem
- once in libhb/work.c, in do_job()
In autoSetM4vExtension, update the condition for chapter markers being on or off, since we can no longer rely on the checkbox being enabled/disabled.
jstebbins [Mon, 28 Mar 2011 17:47:21 +0000 (17:47 +0000)]
Fix problem with qt7 playback of mpeg4 video
The reference frames were not being tagged correctly during muxing
which really screwes up qt7 but appears to have little effect on qtx
or other players.
sr55 [Sun, 27 Mar 2011 13:52:36 +0000 (13:52 +0000)]
WinGui:
- Remove built-in quicktime based video preview support.
- Resigned the Preview window as a result. Currently only supports VLC launching, but will later add an option to use the systems default player for the format produced.
jstebbins [Tue, 22 Mar 2011 02:25:05 +0000 (02:25 +0000)]
LinGui: use g_object_set to set widget properties
instead of directly setting the member variables. The member variables
dissepear into priv data in gtk+-3.0.
jstebbins [Tue, 22 Mar 2011 01:08:37 +0000 (01:08 +0000)]
LinGui: No need to call gtk_set_locale
gtk_init does it for us. And since gtk_set_locale is being removed
from gtk+-3.0, might as well remove the call.
jstebbins [Mon, 21 Mar 2011 18:14:04 +0000 (18:14 +0000)]
Fix linking on mingw when using libtool
Linking the gtk gui uses libtool. The *.la files libtool uses specify
the dependency libraries that libraries are built against. Our
libbluray contrib was being built against the system libxml2 due to
dependencies that were not set in our build system and PKG_CONFIG_PATH
not being set when configuring.
Note that on fedora (and possibly other systems) the ${cross}-pkg-config
script provided is broken and explicitely sets PKG_CONFIG_PATH="". You
need to replace this script with something like this:
PKG_CONFIG_LIBDIR=$(${cross}-gcc --print-sysroot)/mingw/lib/pkgconfig pkg-config $*
jstebbins [Mon, 21 Mar 2011 02:20:48 +0000 (02:20 +0000)]
Remove legacy % to RF mapping.
Although the % option has been gone for a while in the cli and gui's,
there were some mappings happening in libhb and for preset imports.
This removes the last vestages of % quality mapping.
Thanks to Rodeo for the patch.
jstebbins [Sun, 13 Mar 2011 21:51:14 +0000 (21:51 +0000)]
fix 2 pass cfr x264 crash
An error in interjob->vrate calculation lead to specifying a different
timebase for the 1st and 2nd pass which x264 does not allow. This
improves the interjob->vrate calculation accuracy and also guarantees
the timebase is the same on both passes regardless of the calculations
accuracy.
jstebbins [Sun, 13 Mar 2011 17:14:36 +0000 (17:14 +0000)]
Fix source dimensions problem with some mpeg2 streams
Was using the incorrect field for actual "display" dimensions of video.
The field I was using is most often used for pan and scan which
crops a 16:9 image to fit a 4:3 display.
jstebbins [Fri, 11 Mar 2011 22:40:30 +0000 (22:40 +0000)]
Add parameter parsing and b-frame support to ffmpeg mpeg-4 encoder
The cli will now accept ':' separated parameters using the '-x' option
for ffmpeg mpeg-4. The linux gui has an entry box on the advanced tab
to add options. The option keys and values are the same as what the
ffmpeg command line allows.
Calculation of DTS timestamps was added to encavcodec.c in order to allow
out of order b-frames. The algorithm is similar to what x264 uses.
jstebbins [Fri, 11 Mar 2011 22:09:18 +0000 (22:09 +0000)]
Fix PAR and height of mpeg-2 1080 content.
mpeg-2 dimensions must be multiples of 16, so the actual coded height
is 1088. But there are fields in the sequence header that define
the "display" height and width. We were ignoring these which lead
to slightly incorrect height and PAR.
jstebbins [Fri, 11 Mar 2011 15:26:31 +0000 (15:26 +0000)]
Fix mkv timestamps that derive from SimpleBlock and default track duration.
Since I've gotten no response to this patch yet on ffmpeg-devel, I'm
committing it here till someone has the time to look at it.
When a SimpleBlock has lacing (multiple frames in a block) the duration
of each frame should come from the default track duration. An initialization
error in ffmpeg causes them to set the duration to 0 which triggers
code higher in the stack to estimate the duration based on bitrate
and frame size. This is wildly inaccurate and leads to very bad timestamps.
jstebbins [Thu, 10 Mar 2011 23:34:08 +0000 (23:34 +0000)]
Use ffmpeg's lock manager for locking needed by ffmpeg
This lets ffmpeg tell us when it needs a lock instead of
us trying to guess which functions we need to wrap in a mutex.
jstebbins [Tue, 8 Mar 2011 17:05:34 +0000 (17:05 +0000)]
Remove a workaround for an ffmpeg bug that has been fixed.
Removal of the workaround also removes the need for a patch
that fails to apply cleanly to latest ffmpeg git. So remove
the patch as well.