C program for converting arib subtitles from .ts files into .ass and .srt formats using libavcodec and libaribcaption.
You will need to have these dependencies before building:
- libavcodec
- libavformat
- libavutil
- freetype2
- cmake
- make
- pkg-config
- C and C++ compiler
then
git clone --recurse-submodules https://codeberg.org/moex3/arib2ass-cstyle
cd arib2ass-cstyle
make -j4
./a2ac -hWindows builds are available under the releases tab.
By default, it will try to match the original arib subtitle format as close as possible.
ass -o out.ass
Positioning within in the middle of the bounding box can be turned off with -no-center-spacing.
Look at く.
The difference could be more easily seen with the --debug-boxes option.
ass --no-center-spacing -o out.ass
The width of each character is assumed to be constant. If it is not, it will be padded to fit the width expected by the ARIB format.
This can be turned off with the --constant-spacing option, with a spacing value in pixels.
ass --constant-spacing 4 -o out.ass
As the ruby position is specified in absolute pixels in the ARIB format, the ruby text will not be at its expected place once the character positions change.
Using the --shift-ruby option, it will be moved to its correct place.
ass --constant-spacing 4 --shift-ruby -o out.ass
Some fonts have a smaller real character size than other fonts at the same font size. One of them is the IPAexGothic font. The font size is 36 here, with a real character width of 33.
ass -o out.ass --font fonts/ipaexg.ttf
Using the --fs-adjust option, the font size will be changed to 39, with a real character width of 36, which is the expected size.
ass --fs-adjust -o out.ass --font fonts/ipaexg.ttf
Selecting a specific font from a .ttc font collection file can be done with the --font-face option, which expects either a numerical index,
or the text name of the font.
The list of fonts contained in a .ttc file can be listed with the fc-query command.
To use the MS PGothic font for example, use ass -o out.ass -f fonts/MSGOTHIC.TTC --font-face 'MS PGothic'.
Some universal font should be used that most users will have installed, or bundle the font file with the mkv, otherwise the formatting might be weird (mainly the unaligned ruby text).
The default is just an srt file without any tags.
srt -o out.srt
Given the --tags option, additional tags will also be written, like color, bold, italic etc.
Here, the --furi option is also enabled, which will try to find the characters meant for the given furigana text, and write it after that.
It should not be used, as it will not be able to find the correct characters most of the time.
srt -o out.srt --tags --furi
It's possible to save all options into a config file.
The option --dump-config will save the currently set options into a file that can be later loaded with --config-file.
If you see something like Found no drcs replacement char for 06cb56043b9c4006bcfbe07cc831feaf. Writing image to file. when running the program,
that means an unhandled DRCS character has been encountered. The png image for the character is written into the folder named drcs
under the current working directory named with their md5 hash sum.
Unhandled characters are default replaced with a space. To use another replacement, use the --drcs-conv option with a file (can be specified multiple times),
or use the [drcs_conv] table in the config file. The format for the file, or the config table is as follows:
Each line is of the format key = replacement_value. key can be the hash sum of the character, or the word all or "*", in
which case all unhandled characters will be replaced by that replacement_value.
key can also be the word files in which case the value is a TOML array of files to load replacement characters from (this is only valid from the config file).
replacement_value can be either a unicode codepoint number, like 0x1F98B or the character itself in quotes like "🦋"
Currently only a single character can be used as replacement.
For example:
# Only in the config file
[drcs_conv]
files = [ "./file1.toml", "./file2.toml" ]
# Valid in config file or as a file to --drcs-conv
06cb56043b9c4006bcfbe07cc831feaf = 0x1F50A
33d4c5243a45503d43fbb858a728664d = "📱"
all = "🦋"This software is still in early developement, so it might crash or produce incorrect output. If that happens, feel free to create an issue :)