Aegisub Help thread

[Kurre]

I prefer to give older aegisub at this time so I make sure that timing works like it should and not giving any head aches for them. All my tutorials made how to time uses older system and now if I give this new aegisub to somebody, my tutorial tells easy way and they have problems right away.

I think that the problem here is that you have been away from the scene for too long. In that time some things have changed and some improved. You just need to update your tutorials and add the step where you zero the wanted lines (that or use the older version, etc.). As a beginner timer I would rate Aegisub’s timing capabilities very high. Just tell your new timers to read the Aegisub’s timing manual and give them some preferable TPP values. You really can't go wrong - except the noobies fill always fail at the beginning but that's just life.

Working with audio
TPP

That's just my two cents. Not that I’m telling you how to run your group. Good luck and hopefully I'll be seeing you release stuff in the near future.

[asunder]

<jfs> that's intentional and not changing. if your subtitles are already timed, meaning not zero-timed, then you're modifying the existing timing and therefore that existing timing is grabbed when moving to a line. it's intentional.

If you already got the reason from jfs beforehand, then what was the point of complaining on this thread? Was it to show the injustices and "poor behaviour" of evil jfs and his diabolical crew of aegisub programmers??? (They eat their own children you know!!)

Anyways, you're right aegisub is complete shit. There's this new subtitle software on the fansub scene called substation alpha 4. You should try it out.

[PositronCannon]

Originally Posted by asunder

Anyways, you're right aegisub is complete shit. There's this new subtitle software on the fansub scene called substation alpha 4. You should try it out.

It's like I'm really back in 2000.

Like said plenty of times already, I fail to see the problem. Aegisub's timing features are the best I've seen and I can't really think of anything to improve them beyond the software just plain reading your mind. Not knowing how to use those features is a whole different issue.

And I gotta say it's kind of hard to decypher all that broken English. Not a native speaker, I know, but really... I imagine it's just as hard for the developers to understand what you're trying to get at.

[Zalis]

Wait, can't Jeroi's problem be solved with one little toggle button?



If the "Automatically scroll audio display to selected line" option is disabled, then committing a line will put the red and orange marks for the next line at some arbitrary point right after the line you just committed. In short, it'll make Aegisub behave exactly the way Jeroi wants it to behave. I haven't tried it, but would disabling the "Grab times from line upon selection" option under Options > Audio accomplish the same thing?

I remember having the same problem when I made the jump from 1.10 to 2.17. But that one toggle button made timing feel just like it did before. Just turn the option back on for tasks other than initial timing. I've never even felt it necessary to zero out all the times before timing a script. Granted, it still takes me ~45 minutes to time a 24 minute script, so what do I know...

[Krycek7o2]

I recently re-formatted my computer and re-installed all of my previous apps.
After installing Aegisub I can't load a D2V file, I am wondering anyone here could help with this problem.

Here's the error I receive when I attempt to load a D2V file. I have created a file with DGIndex and with MeGUI and nothing seems to work.


Any help will be appreciated.

[jfs]

You're probably best off creating an Avisynth script (.avs file) that loads the D2V file using the appropriate Mpeg2Source filter, and then open that AVS file in Aegisub.

[Krycek7o2]

Originally Posted by jfs

You're probably best off creating an Avisynth script (.avs file) that loads the D2V file using the appropriate Mpeg2Source filter, and then open that AVS file in Aegisub.

I've never created an avs script before, any tips that you could provide?

Also, any idea why/how this happened? On my last setup all of my D2V files were loading properly. I am using the current release of Aegisub.

[jfs]

Well you can try switching video provider to Avisynth. (You do that in Options.)

But actually, you're better off just encoding the video into H.264 dumped into MKV and load that in Aegisub. It's much more reliable.

[Krycek7o2]

Thanks for the help!

Upon switching to AviSynth and creating a script the movie file was loaded properly. The advice was much appreciated.

[simonrule]

when i want to recent audio or open audio from video
aegisub crashed
====================
---2010-07-18 15:32:30------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x0040298A:
001 - 0x0040284F:
002 - 0x0040286F:
003 - 0x004028DF:
004 - 0x00402ED3:
005 - 0x00648A2C:
006 - 0x00648FA8:
007 - 0x0049BE3D:
End of stack dump.
----------------------------------------

---2010-08-22 00:31:22------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x33008600:
001 - 0x0046DEC4:
002 - 0x0046E462:
003 - 0x00470FCC:
004 - 0x00471A41:
005 - 0x00472A2A:
006 - 0x0046E474:
007 - 0x00462F40:
008 - 0x0046D807:
009 - 0x0046E72C:
010 - 0x00462EDE:
011 - 0x006E6DC4:
012 - 0x0049D425:
013 - 0x0049D4DB:
014 - 0x78543433: endthreadex
015 - 0x785434C7: endthreadex
016 - 0x7C80B6D9: GetModuleFileNameA
End of stack dump.
----------------------------------------

---2010-08-26 12:49:55------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x0046F843:
001 - 0x0046F2ED:
002 - 0x0046F012:
003 - 0x004701EE:
004 - 0x00463159:
005 - 0x006E6ECF:
006 - 0x0049D425:
007 - 0x0049D4DB:
008 - 0x78543433: endthreadex
009 - 0x785434C7: endthreadex
010 - 0x7C80B6D9: GetModuleFileNameA
End of stack dump.
----------------------------------------

---2010-08-26 23:48:36------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0428:
002 - 0x10001CB0:
003 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-26 23:48:52------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04: FFMS_AudioSource::~FFMS_AudioSource on c:\c-projects\ffms2\src\core\audiosource.cpp:118
001 - 0x78559BD3: localtime64_s
002 - 0x003B0428:
003 - 0x10001CAF: FFMS_AudioSource::GetAudioCheck
004 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-26 23:49:53------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04: FFMS_AudioSource::~FFMS_AudioSource on c:\c-projects\ffms2\src\core\audiosource.cpp:118
001 - 0x78559BD3: localtime64_s
002 - 0x003B0538:
003 - 0x10001CAF: FFMS_AudioSource::GetAudioCheck
004 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-26 23:53:12------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0348:
002 - 0x10001CB0:
003 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-27 00:05:41------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0438:
002 - 0x10001CB0:
003 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-27 00:28:41------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0448:
002 - 0x10001CB0:
003 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-27 22:11:15------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0448:
002 - 0x10001CB0:
003 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-27 22:19:14------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04: FFMS_AudioSource::~FFMS_AudioSource on c:\c-projects\ffms2\src\core\audiosource.cpp:118
001 - 0x78559BD3: localtime64_s
002 - 0x003B0538:
003 - 0x10001CAF: FFMS_AudioSource::GetAudioCheck
004 - 0xF4E81051:
End of stack dump.
----------------------------------------

---2010-08-28 16:55:31------------------
VER - 2.1.8
FTL - Begining stack dump for "Fatal exception":
000 - 0x10001C04:
001 - 0x003B0428:
002 - 0x10001CB0:
003 - 0xF4E81051:

[guest0815]

Is softsub typesetting, animated with a new line for every frame and position, like in [Commie] The World God Only Knows - 01 [476E9732].mkv (http://pastebin.com/RSDTnaJw), done in an out-of-the-box aegisub (there would be an effect field, no?) or with private tools?

[Plorkyeran]

That was probably motion tracked in mocha or after effects then exported to ass using a script.

[guest0815]

Are moving tags and/or rotations with distant origins in transforms tags with accel parameter save for softsubbing? The few things I tried on my computer dont seem to lag :3

edit: looks like smplayer portable says no to \t

[jfs]

The \org tag has been in ASS since forever, so have \move, \pos, \t and the rotation tags. Rotation and movement shouldn't be any slower (or faster) when origin is specified.
One thing to be careful about is using decimal coordinates for \move and \pos, because was added only later, so there might be a few users that have old VSFilter versions, that don't handle non-whole coordinates for \move and \pos. (Those that don't handle non-whole coordinates will completely ignore the tag, making the text use the default position instead of the specified.)

[Aimbotter33]

Anybody Know how to Encode The Karaoke Effects to .avs for the encoder to do the job..

[Quarkboy]

Well, I finally got around to it, and so I figured I might post it here:

Quarkboy's Smart Line Splitter v2.0

This lua automation script is now fully configurable with a configuration dialog.

Features include:

1. Adjustable maximum character length
2. Apply to selection only, and/or any subset of styles
3. Adjustable "overhang bias" from 1 = always top longer, to 0 = always bottom longer
4. Selectable behavior for existing \N... Either overwrite or ignore
5. Optional commenting out of any detected 3-liners for easy spotting and fixing.

Please pardon my crappy coding style, since I'm not exactly an LUA expert.

script_name = "Smart Linesplitter"
script_description = "Insert \N Linebreaks"
script_author = "Sam Pinansky"
script_version = "2.0"

function isininttab(tab,match)
	for i = 1, table.maxn(tab) do
		if tab[i]==match then return true end
	end
	return false
end

function smart_linebreaks(subtitles, selected_lines, active_line)
	local conf = {}
	conf[1]={class="label",label="Smart Linesplitter by Sam Pinansky v2.0",x=1,y=5,width=2}
	conf[2]={class="intedit",name="maxlinewidthconfig",hint="Lines will be split into two halves each less than this many characters.\nIf this is not possible, then a line break will be inserted such that the first half\nis less but the second half is too long, and a warning will be written.", value=45,min=1,max=3939,x=1,y=0}
	conf[3]={class="label",label="Max char per line",x=2,y=0}
	conf[4]={class="floatedit",name="lowerbias",hint="Determines whether to have the top line or bottom line longer.\nBased on character numbers and not actual width, so results may vary.",value=0.66,min=0,max=1,step=0.01,x=1,y=1}
	conf[5]={class="label",label="Overhang bias",x=2,y=1}
	conf[6]={class="label",label="Apply to styles:",x=0,y=0}
	conf[7]={class="checkbox",label="Apply to selection only",hint="Only apply to the currently selected lines.",name="selectedonly", value=false,x=1,y=2,width=2}
	conf[8]={class="checkbox",label="Overwrite existing linebreaks",hint="If checked, all \\N's in lines are replaced with new breaks at the desired max limit.\nUnchecked, any line with a \\N will be left as is.",name="overwrite",value=false,x=1,y=3,width=2}
	conf[9]={class="checkbox",label="Comment out 3-liners",hint="If checked, lines that are too long will be commented out.",name="markthree",value=false,x=1,y=4,width=2}
	local styleindex = 10
	for i = 1, subtitles.n do
		if subtitles[i].class == "style" then 
			conf[styleindex] = {class="checkbox",label=subtitles[i].name,name=subtitles[i].name, value=true,x=0,y=styleindex-9}
			styleindex = styleindex + 1
		end
	end
	local ok,configresult = aegisub.dialog.display(conf)
	local maxlinewidth = configresult.maxlinewidthconfig
	linecount = 0
	if not ok then return end
	for i = 1, subtitles.n do
		local l = subtitles[i]
		if l.class == "dialogue" then linecount = linecount + 1 end
		local process = true
		if configresult.selectedonly then 
			if isininttab(selected_lines,i) then 
				process = true 
			else process = false
			end
		else process = true 
		end
		if l.class == "dialogue" and configresult.overwrite then l.text = l.text:gsub("\\N","") end
		if l.class == "dialogue" and not l.comment and configresult[l.style] and process and not l.text:find("\\N") then
			local tstrip = l.text:gsub("%{.-%}","")
			local linelength = tstrip:len()
			local commentarray = {}
			local commentlocarray = {}
			for comit in l.text:gmatch("%{.-%}") do table.insert(commentarray,comit) end
			local loc1 , loc2 = l.text:find("%{.-%}",1)
			while loc1 do table.insert(commentlocarray,{loc1,loc2}) loc1 , loc2 = l.text:find("%{.-%}",loc2)  end
			local rebuilt = ""
			local pos = 0
			local split = false
			local insertionpoint = 0
			if linelength > maxlinewidth then
				for w in tstrip:gmatch("[^%s]+%s*") do
					pos = pos + w:len()
					if pos > linelength/2 and not split then 
						if pos > maxlinewidth or (pos-linelength/2) > (w:len()*configresult.lowerbias) then
							rebuilt = rebuilt .. "\\N" .. w 
							split = true
							insertionpoint = pos - w:len()
							pos = 0
							else
							rebuilt = rebuilt .. w .. "\\N" 
							split = true
							insertionpoint = pos
							pos = 0
							
						end
						else					
						rebuilt = rebuilt .. w
					end
				end
				if pos > maxlinewidth then aegisub.debug.out(2,"WARNING, 3+ line: %i\n",linecount) end
				if pos > maxlinewidth and configresult.markthree then l.comment = true end
				for j,com in ipairs(commentarray) do 
					if commentlocarray[j][1] < insertionpoint then 
						rebuilt = rebuilt:sub(1,commentlocarray[j][1]-1) .. com .. rebuilt:sub(commentlocarray[j][1])
						insertionpoint = insertionpoint + com:len()
						else
						rebuilt = rebuilt:sub(1,commentlocarray[j][1]+1) .. com .. rebuilt:sub(commentlocarray[j][1]+2)
					end
				end
				l.text = rebuilt
			end
			subtitles[i] = l
		end
	end
	aegisub.set_undo_point("Insert Linebreaks")
end

aegisub.register_macro("Smart Line Splitter", "Insert Manual \N Linebreaks", smart_linebreaks)

[idiffer]

does aegisub play .wmv? if not, then is it planned in the future?
i haven't used it in a long time. hoping for my dream to come true.

[jfs]

I can't remember if FFmpegSource handles Windows Media stuff. You might have some luck if you can remux the file to MKV, rather than the closed ASF container. (The WMV9 video coding is published as a standard, the container format is not. Correct me if I'm wrong.)

[idiffer]

Originally Posted by jfs

I can't remember if FFmpegSource handles Windows Media stuff. You might have some luck if you can remux the file to MKV, rather than the closed ASF container. (The WMV9 video coding is published as a standard, the container format is not. Correct me if I'm wrong.)

and...nope, doesn't work sadly. oh well...
btw, why does aegisub state that it is not a player, thus everyone can shut up about it?
the problem i'm having with .wmv is related to subbing, not just convenience. i can't typeset moving signs when my video is .wmv. if i reencode it to other formats, typeset that, then mux the subs with the original .wmv the result is a mess with the subs being off. this is probably related to .wmv being VFR and the reencoded shit is always constant frame rate. so you can't just shift the subs, not moving signs.
soooo, why the stubornness?
just in case, cause i'm a noob, i'll ask this - is there a way to reencode .wmv retaining the variable frame rate? although this probably belongs in another thread...
PS. i know i sound rude, but really i'm just curious. i swear.

[Quarkboy]

Originally Posted by idiffer

just in case, cause i'm a noob, i'll ask this - is there a way to reencode .wmv retaining the variable frame rate? although this probably belongs in another thread...
PS. i know i sound rude, but really i'm just curious. i swear.

There isn't any _good_ way, but one not so good way is to open it up in avisynth with a directshowsource and an fps = 120 (or, 119.XXX)...

Then encode that as Constant frame rate intermediary, then use a filter like dedup to get a vfr timecodes file and mux into a reasonable container like mkv.

It depends exactly how screwed up the vfr is in the wmv. If it's something reasonable that should work fine