Aegisub public beta

[ArchMageZeratuL]

Aegisub is the result of a few months of work in a project designed to completely replace Medusa for typesetting, as well as add a large number of useful and/or vital features that Medusa lacks.

After more time than I wanted in Alpha testing phase, it should now be stable enough for public distribution and testing - it is, after all, almost impossible to find all bugs without a large number of testers.

I hope you will find Aegisub to be useful - if you do, please contribute with bug reports, feature suggestions and join our little IRC channel! Thanks.

Please follow link to official website below for download.


Highlights

• Unicode in UTF-7, UTF-8, UTF-16LE and UTF-16BE encodings
• Importing non-Unicode subtitles from over 30 different character encodings, including Shift_JIS, allowing you to resume work on any subtitles regardless of your system locale settings
• Loading advanced Substation Alpha (.ass), Substation Alpha (.ssa), Subrip (.srt) and Plain-Text (.txt) (with actor data) subtitle formats
• Full tag conversion from SRT to ASS
• Multiple Undo levels
• Powerful Automation module using the Lua scripting language to create advanced karaoke effects (including per-syllable effects, similar to AssCalc) - or any other kind of subtitles manipulation
• Syntax highlighting
• Fonts collector to collect all fonts used in a script into a folder
• Opening of videos via Avisynth for true preview of subtitles using VSFilter, in any format supported by DirectShow
• Aspect ratio overriding for anamorphic videos
• Variable Frame Rate videos via timecodes file (v1 and v2), including exporting processed times for hardsubbing
• Display of video coordinates under mouse and simple auto positioning support
• Audio mode which allows you to open files in any format that your DirectShow can decode, including Wave, mp3, Ogg Vorbis, AAC, AC3 and others
• Powerful audio timing and playback mode, including full support for karaoke manipulation
• Functions and hotkeys easily synchronizing audio/video and subtitles
• Time shifting and manipulation by timestamps or frame numbers (once video or timecodes are loaded)
• Advanced find and replace with support for Regular Expressions
• Cut/Copy/Paste subtitles via plain-text format, so it can be pasted in any plain-text editor
• Translation assistant to easily translate subtitles to another language
• Styling assistant to quickly pick styles for each line in subtitles
• Styles manager that allows you to categorize subtitles styles, to make working on several projects painless
• Keyframe navigation for fast seeking to scene boundaries
• Several subtitles manipulation macros, such as making timing continuous or recombining double lines after OCR
• Highlighting of subtitles that are visible on current video frame, and a button to select those
• Preview subtitle changes imediately by pressing Ctrl+Enter
• Shortcut buttons to set text as bold, italic, underline, strikeout, change font or any of the 4 colors
• Extensive manual with detailed instructions and tutorials for typesetting and program usage

Screenshots




System requirements

• Windows NT 5.0 or later, that is, Windows 2000, XP, 2003 or Vista. (It has been widely tested on 2000, XP and 2003, and is expected to work properly on Vista as well). Windows 95, 98 and ME are NOT supported (due to unicows not feeling like working). Presently, there is no Linux port, either, due to dependency on Avisynth and VSFilter - please consider using Sabbu instead if you want to work on an Unix environment.
• 1024x768 display (more is recommeneded) on either 16 or 32 bits per pixel
• 256 MB RAM for good performance, 512+ is recommended especially when dealing with long audio in RAM cache.
• A fast processor might be required for decent video playback speeds.

Links
Official Aegisub website
Official Aegisub forums
Official Aegisub IRC channel

[ArchMageZeratuL]

Aegisub 1.07 is out. Grab it at the usual place, and changelog is available on http://aegisub.cellosoft.com/changelog.txt

Also, the forums are now available. Please post bug reports and feature requests there to keep things organized.

[Eggh]

New build of aegisub is up!
1.08 beta: http://files.cellosoft.com/zeratul/Aegisub-1.08.exe

Changelog:
= 1.08 beta - 2006.01.03 ===========================

- Fixed crash when trying to shift times of selected rows with none selected (AMZ)
- Added option to make audio autocommit only work when SSA mode is disabled (AMZ)
- Implemented exporting to SSA (Substation Alpha) subtitles format (AMZ)
- Added *.mka to list of recognized audio formats (AMZ)
- Fixed (harmless) unhandled exception when cancelling audio loading (AMZ)
- Added recent lists to video and audio (AMZ)
- Added option to allow user to disable jumping to next line when commiting in SSA mode (AMZ)
- Fixed line jump to in SSA mode (AMZ)
- Aegisub no longer tries to open .vob files with mpeg2source (Myrsloik)
- Implemented saving of PNG snapshots from video (AMZ)
- Implemented customizeable hotkeys for most shortcuts (AMZ)
- Added hotkey customizing dialog (AMZ)
- Bugfix: simple-k-replacer.lua never worked (jfs)
- Added "apply only to this style" options to simple-k-replacer.lua and line-per-syllable.lua (jfs)
- Fixed selection being reset after lines were joined (AMZ)
- Duplicated lines are now created after selection, and first one is selected (AMZ)
- Added duplicate and shift by one frame option for easier frame-by-frame typesetting (AMZ)
- Fixed crash related to setting style overrides via buttons (AMZ)
- Main program window can now be freely resized (AMZ)
- Zoom slider under video was replaced with a drop-down box on toolbar (AMZ)
- Drawing command blocks are now parsed properly (not that it matters much) (AMZ)
- Video slider is now a custom control (AMZ)
- Automation: The frame_from_ms and ms_from_frame functions weren't actually available. They are now. Oops... (jfs)
- Automation: Fixed error reporting during script execution so you have a chance to see the error messages again (jfs)
- Mouse wheel now scrolls the audio display, or zooms if shift is held down (you can invert the two in config.dat) (AMZ)
- New lines now keep the style of 'parent' line (AMZ)
- First line of script will now be selected when it's loaded (AMZ)
- Renamed "Audio Draw Boundary Lines" into "Audio Draw Secondary Lines" (AMZ)
- Added Undo/Cut/Copy/Paste/Select All to subtitles edit box context menu (AMZ)
- Added a "OK" button to style editor dialog, and now "Apply" applies changes without closing it (AMZ)
- Paths to video/audio/timecodes files are now stored relative to subtitles (instead of as absolute paths) in [Script Info] (AMZ)
- Audio display now shows previous line in grey (this feature can be disabled (0) or extended to every line (2) with "Audio Inactive Lines Display Mode" on config.dat) (AMZ)
- Audio display height can now be adjusted by drawing the sash under it (AMZ)
- Holding shift while clicking on video seek bar now snaps to keyframes (AMZ)
- Fixed positioning of subs on video while shift is held down (AMZ)

[MeteorRain]

Very cool tool. Thankyou, Aegisub Developers!

[ArchMageZeratuL]

By the way... Version 1.08 has audio timing mode very similar to SSA (after SSA mode is enabled, and possibly with tweaking on config.dat). I would like to hear from timers used to Substation Alpha: what do you think of 1.08? Would you change anything? What needs to be implemented?

Thank you for your cooperation.

[devilcoelhodog]

Version 1.08 has audio timing mode very similar to SSA

That's great !!! I 'll check as soon I arrive in my home, testing Aegisub in some timing projects.

What needs to be implemented?

For animes, I think the people are enjoying the Karaoke & similar features.

I will try Aegisub in making some scripts using longer videos ( + 80 minutes) and see what happens. If something weird happens, I will put the results in these thread, as some people suggested before.

I know that Aegisub was intended for animers, karaokeining and fansubbers.

But, who knows, it can help people who wanna try the program for subtitling many releases from movies around internet.

As SSA could time almost all kind of things before in past ( not only animes I could do it using SSA, but longer sources, like 2 hours materials in lenght ), maybe Aegisub could be able to do the same.

Thanks for your effort in making Aegisub better and better.

devil (johner)

[RaistlinMajere]

General annoyance:
-I NEED the ability to select times without having a line selected. That restriction makes no sense.

1.08 bugs:
-I can't change the style. It just doesn't take effect. I click it, it changes, I can press enter, I can click commit, but no matter what, it does not stay as I changed it when I click another line, nor does it take effect in the preview. It just changes the combo box (until I click another line and click back)
-Auto commit for times doesn't seem to work. I've unclicked and reclicked it, but I still have to use the commit button for times.

[ArchMageZeratuL]

Originally Posted by RaistlinMajere

General annoyance:
-I NEED the ability to select times without having a line selected. That restriction makes no sense.

I don't quite understand... what time, if there is no line selected?

Originally Posted by RaistlinMajere

-I can't change the style. It just doesn't take effect. I click it, it changes, I can press enter, I can click commit, but no matter what, it does not stay as I changed it when I click another line, nor does it take effect in the preview. It just changes the combo box (until I click another line and click back)

Uh, it works fine for me... the style dropdown imediately applies effect to every line selected in grid below, not necessarily what's visible in the editbox.

Originally Posted by RaistlinMajere

-Auto commit for times doesn't seem to work. I've unclicked and reclicked it, but I still have to use the commit button for times.

There's an option in config.dat that blocks autocommit on SSA mode if you have "go to next line on commit" checked. You can invert them if you want... I'm not sure if I want to make autocommit just not go to next line or what...

[devilcoelhodog]

select times without having a line selected.

It's possible to create an option to activate or disable the "color selection" that select the lines (begginnig and the end of time line) in Aegisub?

eg: In SSA i can select the begining and the end WITHOUT any color or "shadow" showing my selection on the timing line. In the old Medusa, for example, a green color appears to show my selected times. I do not like this "feature".

I really think ( for me, the other people may say different opinions) its better to timing just having to mark for the beginning and other, the end, without nothing on time line.

That's more easy to see the proper positions. So, if that feature could be activated or not, according the needs and tastes of each timer, it will be nice.

Thanks.

[TheFluff]

@devilcoelhodog: I'm not sure that I understand what you're talking about, but maybe one of these options are what you're looking for...:

Originally Posted by Aegisub manual, Options section

audio draw selection background=BOOL
Controls whether the background for the current audio selection is shown in different colors or not. Defaults to true (1).

audio inactive lines display mode=NUM
Controls how Aegisub displays the lines surrounding the currently selected one in audio mode. Valid modes are:
0 - Disabled (only show the current line)
1 - Show the previous and next lines in grey (this is the default behaviour)
2 - Show all unselected lines in grey

audio line boundary start=HEX
audio line boundary end=HEX
The color of respectively the start and end line boundaries in the audio view, as hexadecimal RRGGBB values. Both default to #FF0000.

audio line boundary inactive line=HEX
The color of the lines surrounding the currently selected one in the audio waveform (see the audio inactive lines display mode setting). Defaults to #808080

audio line boundaries thickness=NUM
Thickness in pixels of the boundary lines for the current audio selection. Defaults to 2.

audio selection background modified=HEX
The color of the background of the selected area, when modified but not yet committed, in the audio view, as a hexadecimal RRGGBB value. Defaults to #5C0000.

audio waveform selected=HEX
The color of the selected area of audio waveform, as a hexadecimal RRGGBB value. Defaults to #FFFFFF.

grid inframe background=HEX
The color of a line in the grid that is currently visible in the video display. Defaults to #FFFDEA.

grid selection background=HEX
The background color of a selected line in the grid. Defaults to #CEFFE7.

grid selection foreground=HEX
The foreground (text) color of a selected line in the grid. Defaults to #000000.

Refer to the aegisub manual, options section, for more details on how to set these options.

[devilcoelhodog]

@TheFluff:

Yes, I guess I've found what I'm looking for in your suggestions.

Basically, I want to time without any color selection in time line. To me, just to set the beginning and the end of the speeches is enough.

No "shadows" or "color selection" to mark them. In Aegisub's case, that color selection is red.

Then, Aegisub can unable or disable that option I'm looking for?

I'll check the manual, because I didn't see it yet, I must confess.

Thanks for help.

[devilcoelhodog]

Oh, I forget to mention: timing in Aegisub improve nice this time.

I'll test more to say what I think about timing using this Aegisub's version.

Thanks.

[TheFluff]

Originally Posted by devilcoelhodog

Basically, I want to time without any color selection in time line. To me, just to set the beginning and the end of the speeches is enough.

Then go to the aegisub folder, open config.dat in a text editor (i.e. notepad) and change the following lines:

Originally Posted by config.dat

audio selection background=#404040
audio selection background modified=#5C0000

to

Originally Posted by config.dat

audio selection background=#000000
audio selection background modified=#000000

And then save the file. Changes will take effect when you restart Aegisub.

If you don't want the waveform to turn white when you have it selected, or greyish when it's modified, change

Originally Posted by config.dat

audio waveform modified=#FFC8C8
audio waveform selected=#FFFFFF

to

Originally Posted by config.dat

audio waveform modified=#00C800
audio waveform selected=#00C800

Do note that if you make all the above changes, there's no way of telling the start and end markers of a line apart.

[RaistlinMajere]

Originally Posted by ArchMageZeratuL

I don't quite understand... what time, if there is no line selected?

If I just want to play a section of the audio, it seems that I have to either select a line first and do it (which isn't a huge deal w/o autocommit) or insert a new line, both more trouble than if I could select times w/o having a line selected.

[ArchMageZeratuL]

I see what you mean. I've just fixed it so there will always be a selected line... it's not -exactly- what you want, but it was a bug, anyway.

Maybe I could make middle click play from current cursor position, and stop when you click it again?

[RaistlinMajere]

I suppose that would be better than nothing. But if you do that, make sure to put some sort of marker where you started. It can go away when you click a line again. Or hell, you could click a line, then middle click, and somehow have the option "shift line to this time" where it'd shift the entire file such that the selected line starts at the place you selected.

Also, a request: Precision to 1/1000s. SSA/ASS spec precision down to 1/100s. And indeed VSFilter reads it as an integer, so going to 1/1000s with a normal version of VSFilter will not work. However, if one changes a line that allows VSFilter to read times precise to 1/100s or 1/1000s, then it becomes a good option, because it is not possible to fine-time 120FPS files with a precision of 1/100s. I request an option (even if it's hidden) to allow for this. I've added the option to SSATool, though that version hasn't been released yet. Catch me on Chatspike if you want the modifed VSFilter or change lines 1488/1489 of STS.cpp inside of Guliverkli's src\subtitles directory to:
(((hh1*60 + mm1)*60) + ss1)*1000 + (ms1_div10<100?(ms1_div10*10):ms1_div10),
(((hh2*60 + mm2)*60) + ss2)*1000 + (ms2_div10<100?(ms2_div10*10):ms2_div10),
and compile VSFilter.

[jfs]

For not having the selecting background change, I think Fluffy missed one of the options he pasted himself:

Originally Posted by Aegisub manual

audio draw selection background=BOOL
Controls whether the background for the current audio selection is shown in different colors or not. Defaults to true (1).

Set that one to false (0) to avoid the current selection changing color. IIRC it also works for spectrum analyser mode.
After you do that, you might also want to change the color of the selection start/end markers. (They can be set to different colors.)

Originally Posted by RaistlinMajere

because it is not possible to fine-time 120FPS files with a precision of 1/100s.

There is no such thing as 120fps video. Those are AVI files with "fake variable framerate", achieved by inserting 3 or 4 drop-frames between every coded frame. Thus, they are in fact 29.97 fps at best, same as everything else. In rare cases they might have 59.94 fps sections, in case the encoder found places where the producers did weird stuff. (I think I saw two 59.94 fps sections in Magical Canan OP, in the 1024x576 HDTV raws.)

[RaistlinMajere]

I know what a 120FPS file really is, but they do exist. Stuff is not broadcast in 120FPS, but 120FPS files exist. Also, 60FPS is possible in a broadcast. If you want a common example in anime, for shows that have a "don't sit close to the TV" thing that scrolls across, that's usually 60FPS (if it's an HD channel).

I suppose on the other hand, it would be a pain to have to zoom in enough to deal with 1/1000s precision... But VFR timing can be a pain as well...

[ArchMageZeratuL]

VFR timing is (should be) painless with Aegisub... is there any reason not to just make it VFR and work with demuxed video+timecodes file?

Because that's what I recommend.

[RaistlinMajere]

My timers don't like it :/

Hell, they don't even want me doing VFR (on a hybrid show) - to which I say too bad But 1/1000s precision would allow them to time from the original raw...