PDA

View Full Version : Timing Guide


Sakaki-
2003-11-09, 04:48
Timing
"The Art of putting subtitles
On a video and audio stream"


Apprehending Tool's

First of all you will need a program called "Sub Station Alpha" (Referred to as SSA)
This program is used to apply times to lines in a text file, using a mono wave stream.
http://www.backupdvd.info/Editing/SSAinstall.exe

Second you will need VirtualDubMod (Referred to as VDub)
This is a program that you can encode video in, mux or demux audio.
This program also has a thing that is useful to timers; it can be used to apply filters to your video stream.
And preview it in real-time, this is where the third tool you need comes in.
http://virtualdubmod.sourceforge.net/

Third thing you will need is VobSub
This is a directx filter, used to be able to view soft subs with your video files.
But what’s more important for you as a timer, is the subtitle filter that comes with vobsub.
This subtitle filter can be used as an AVS (Avisynth) or VDub filter.
http://www.doom9.org/software2.htm#subs

So to get started.
Basically you now have all the tools on your computer needed to subtitle an anime.
I will assume the following.
1. You have a Video file.
2. You have a script with the subtitles that needs timing.
3. You have an obtained the above tool's.
The first step. "Getting material to work with."
Start VDub.
Then go to "File" ---> Open video file, and select your video file.
(Note: If you get a warning that the video stream uses vbr audio redo this step selecting "Nandub VBR MP3 compatibility mode”)
Now you opened your video file,
Go to "AVI" ---> Save WAV...
Save your wav files as something like this: (<series><episode>mp3.wav). Eg: PMK02mp3.wav
Now that you have saved the audio stream in a separate file, you can close down VDub.

The second step. "Conversion"
Go to Start ---> Run, and enter sndrec32.
Then go to "File" ---> open, and select the (<series><episode>mp3.wav).
Then go to "File" ---> Save As,
And click the change button next to Format: and for "Select Format PCM”, use 8,000 KHz Mono 8 bit.
(Note: you can select whatever here, as long as it is 8 bit Mono)
Then save the files as (<series><episode>mono.wav)

The Third Step. "First step of Actual Timing"
Now we are all done with preparations. ~^_^~ phew...
Open SSA.
Then go to "File" ---> Open, and in the format box select: (Plain Text), then select the script you're going to time.
If your translator already has an .ssa file ready for you, just open that instead.
Then click "Timing" ---> Time from WAV file 0
Click the new Open button, and Select the (<series><episode>mono.wav) file.
Use Play to play the audio, and Stop for when you get to the dialogue that matches the text that is selected.
Select the area where the dialogue is on the wave track and click Grab Times.
You can use the Play Selected and! Buttons to test out the selection; play around and find what works best for you.
(Note: You can use the zoom and y-scale bars to make it easier for you to see where the dialogue starts and ends)
You continue like this until all the lines are timed.
(Tip: When doing this, adding an amount of time after every line that you timed is good; that will make it easier to read later.)
(Tip: Save often and/or use the Auto-Save option under “File”)

The Fourth Step "Polishing your timing."
Now you have a timed script.
But there’s one step left if you want your subtitles to look nicely timed and tidy.
This step is referred to as "Scene Timing".
This step is basically just checking so that your timing doesn’t overstep scene changes if the talking doesn’t.
This is done by looking at your video in VDub using the TextSub filter that came with VobSub.
Start VDub
Go to "Video" ---> Filters.
Click the Add button.
Select TextSub
Click the Open button and select SSA file you saved.
Exit the filtering area and make sure the TextSub filter is loaded.
Now open the SSA file with notepad or any good text editors.
Now watch through your video and look for places where the text goes over a scene change.
When you find a spot like this, stop the video, and copy the time of the frame directly after the scene change.
Dialogue: Marked=0,0:03:14.97,0:03:17.94,*Default,,0000,0000 ,0000,,Next door's welcoming is amazing, right Aione-kun?
Let’s say this is the line that oversteps a scene change, and the time directly after that scene change is 0:03:17.23
Dialogue: Marked=0,0:03:14.97,0:03:17.23,*Default,,0000,0000 ,0000,,Next door's welcoming is amazing, right Aione-kun?
This is how you change that line then; this will make the text end exactly before that scene change.
(Note: If you notice a line ending around 2 or 3 frames before a scene change, try to drag that one forward changing just like this one but making the line show longer instead)
Continue like this until you made sure no scene change has lines overstepping them unless the corresponding dialogue is doing it to.
(Tip: Often the frame directly after a scene change is a key frame; you can use the vdub button with a key and arrows, or the shift-<arrow key> hotkeys to jump to them.)


Brought to you by
Sakaki- & Kirika
#Timers@irc.mircx.com

uLTraCarL-
2003-11-09, 05:43
brand spanking new timing guide

hurrah hurrah


#timers @ mirc-x
its so lonely in there, liven it up guys

GipFace
2003-11-09, 13:10
I'll add my two cents into the fire ...

A bad alternative (http://nnl1.com/t00t-04.html)

runpsicat
2003-11-11, 10:17
I'll add my two cents into the fire ...

A bad alternative (http://nnl1.com/t00t-04.html)

I have a question to ask timers regarding the following:

"Quicksubs are bad. While you're timing someone's speech, the subs change every 1-2 seconds to reflect a new line. This pitfall happens to a lot of newbie timers - don't let it happen to you. Create a soothing melody of sub changes that change less often during a long speech, ranging from 3.0-6.5 seconds."

I've heard arguments both for and against this viewpoint. Some viewers complain that the more "inclusive" subs often spoil upcoming dialogue and prevent them from relishing "moments" in the episode. Shorter subs are indeed annoying if they are accompanied by sub-blinking, but if the transition is smooth, where is the harm in having shorter subs to enable a closer approximation of the "listening" experience (i.e., information only enters the brain as the dialogue is said)? Any thoughts?

Also, do normal viewers even notice the difference between blipping/non-blipping subs and subs that "bleed" over scene changes versus those that do not? Most people tell me they don't start noticing until these factors are specifically brought to their attention, and that their viewing experience is then compromised by the fact that they start noticing all these "glitches" they were not aware of before (the fate of all QCers, apparently). Most people seem to be concerned only with whether the subs are accurate and stay on long enough for them to read. If so, are these fine-tuning processes truly worth tearing one's hair out over? I suppose the guide is but a guide and exceptions are to be expected, but I just felt like asking these questions anyway. ^^


Edit: The VBulletin tags are not working for some reason (they are enabled, but not working), so I used the time-honored quotation marks instead ^^;

GipFace
2003-11-11, 11:04
I've heard arguments both for and against this viewpoint. Some viewers complain that the more "inclusive" subs often spoil upcoming dialogue and prevent them from relishing "moments" in the episode. Shorter subs are indeed annoying if they are accompanied by sub-blinking, but if the transition is smooth, where is the harm in having shorter subs to enable a closer approximation of the "listening" experience (i.e., information only enters the brain as the dialogue is said)? Any thoughts?

The editor should be able to figure out when to quicksub a line to prevent upcoming spoilage. However, what I have mostly seen is both ends of the extreme ... too many quicksubs or too many longsubs. This is really the editor's job, though; he or she should tell the timer what to adjust.

However, the worst example is a sub change when the character doesn't even pause for a breath. You can use "..." to extend it to the next line but some people don't even do that.

Also, do normal viewers even notice the difference between blipping/non-blipping subs and subs that "bleed" over scene changes versus those that do not? Most people tell me they don't start noticing until these factors are specifically brought to their attention, and that their viewing experience is then compromised by the fact that they start noticing all these "glitches" they were not aware of before (the fate of all QCers, apparently). Most people seem to be concerned only with whether the subs are accurate and stay on long enough for them to read. If so, are these fine-tuning processes truly worth tearing one's hair out over? I suppose the guide is but a guide and exceptions are to be expected, but I just felt like asking these questions anyway. ^^

I'm anal, so I would. You don't NEED to do it, but if you're not giving 100% effort, why do it at all? ^_^

SirCanealot
2003-11-11, 13:34
I myself like to break up lines by only including usely at most 2 sentances on screen at once, or breaking the lines up with "...", commas and of course following the subs on with no "blink". While I have nothing against shoving more on screen at once, I think it's best to try and give the viewer what is said as it is said. Allthough my timing style is still very much in development right now, from only having put out 4 episides thus far with my timing.

I'm anal, so I would. You don't NEED to do it, but if you're not giving 100% effort, why do it at all? ^_^

Yup, I always do my best when timing, I don't see any point in giving it anything other than my full. Except for previews, I'm usely suffering from timing insanity (something all but the most toughest and experienced timers suffer from :P) about then, so I sub the previews in about 4 seconds....
I guess some people are more concered with getting the product out than QCing it.
I'm anal about subs bleeding into the next scene too, not really because it looks gay (allthough it can, when the next scene has NOTHING to do with the last), but because cutting subs at a scene chage looks cool. Lately I've been extending subs to scene changes if they're close to them, so the sub disapears with the scene change. I think it looks quite good...

Thing I should emphasis as more of a critique than I'll ever be a timer is, yes, blinking subs. What the hell is wrong with people who have a 0.2 second sub for "hai" or someones name? Surely you must relise I'm going to have to move my body to my keyboard, which will be next to the sofa somewhere, press the left arrow key to go bacwards 5 seconds then pause to read the sub. Just don't do it. Give us some lead out time, especially on shorter sentances. I try to atleast make sure my subs don't disapear instantly when the dialog ends. Also, with trans notes, don't have a trans note on the screen explaining something about a sub that stays on screen for 1 second. I don't want to have to again, move my body to read it. Keep the translation note on screen as long as possible!

Also, do normal viewers even notice the difference between blipping/non-blipping subs and subs that "bleed" over scene changes versus those that do not? Most people tell me they don't start noticing until these factors are specifically brought to their attention, and that their viewing experience is then compromised by the fact that they start noticing all these "glitches" they were not aware of before (the fate of all QCers, apparently).

I've been thinking about that before and I'd say yes and no. Learning about poetry, all the subtle things the poet has (allegedly) put into the poem to convey it's meaning. Unless you learn about them you wont really notice them, but you'll still get the effect of them. I guess the same could be said about subs. The subs look a lot more tidy when you cut them at the scene, I think it's slightly more enjoyable to watch a episode that has been timed well than one that hassen't. (edit: actually, see my comment about eyes at the bottom of the screen bellow, I think this plays a factor)
And I would say my viewing experience has been compimised slightly by knowing all this, allthough saying that I feel all the more enjoyment now watching a well timed episode, knowing a fellow timer has put hours into the script and knowing I wont have to keep an eye out for funny subtitle timings (keep my eyes glued to the bottom of the screen) and can enjoy the artwork a lot more.

What really pisses me off these days is R1 DVDs... I don't watch too many of them, due to lack of money, but I haven't seen a well timed DVD yet since becoming a timer/anal about timing :P
The person that timed Card Captor Sakura seems to take peverse pleasure in making sure even the smallest of subs stay on the screen for 2 seconds and almost every subtitle bleeds over into the next scene. Really annoying...

Edit: Sorry about all the typos there probebly is in this post... just got back from college = TIRED

Sakaki-
2003-11-11, 13:54
I've heard arguments both for and against this viewpoint. Some viewers complain that the more "inclusive" subs often spoil upcoming dialogue and prevent them from relishing "moments" in the episode. Shorter subs are indeed annoying if they are accompanied by sub-blinking, but if the transition is smooth, where is the harm in having shorter subs to enable a closer approximation of the "listening" experience (i.e., information only enters the brain as the dialogue is said)? Any thoughts?

Well i would say, There are times when the gap betwene the lines are to little so i don't leave a space betwene the subs, but when theres fast dialouge it gets very annoying.

"where is the harm in having shorter subs to enable a closer approximation of the "listening" experience"

Well i want kind of short subs, But there are times when you need to have loger ones too.

Also, do normal viewers even notice the difference between blipping/non-blipping subs and subs that "bleed" over scene changes versus those that do not? Most people tell me they don't start noticing until these factors are specifically brought to their attention, and that their viewing experience is then compromised by the fact that they start noticing all these "glitches" they were not aware of before (the fate of all QCers, apparently). Most people seem to be concerned only with whether the subs are accurate and stay on long enough for them to read. If so, are these fine-tuning processes truly worth tearing one's hair out over? I suppose the guide is but a guide and exceptions are to be expected, but I just felt like asking these questions anyway. ^^
I've been thinking about that before and I'd say yes and no. Learning about poetry, all the subtle things the poet has (allegedly) put into the poem to convey it's meaning. Unless you learn about them you wont really notice them, but you'll still get the effect of them. I guess the same could be said about subs. The subs look a lot more tidy when you cut them at the scene, I think it's slightly more enjoyable to watch a episode that has been timed well than one that hassen't. (edit: actually, see my comment about eyes at the bottom of the screen bellow, I think this plays a factor)
And I would say my viewing experience has been compimised slightly by knowing all this, allthough saying that I feel all the more enjoyment now watching a well timed episode, knowing a fellow timer has put hours into the script and knowing I wont have to keep an eye out for funny subtitle timings (keep my eyes glued to the bottom of the screen) and can enjoy the artwork a lot more.


I get slightly annoyed after watching a non scene timed sub.
And i got annoyed even before becoming a timer of that, It feels like getting the sub out was rushed and gives a bad feeling.
As you say i feel more relaxed when it scene timed, I don't notice the subtitles as much then.

Shii
2003-11-11, 15:35
SSA is ancient; its features reflect that it was made for the era of VHS fansubs. I like XombieSub, which has the nice feature of being able to see the video you're editing as well as hear the audio.

Enragin_Angel
2003-11-11, 16:43
ashibaka: what are you talking about? xombie never even finished developing it.

Blipping subs? I thought blipping subs were subs that ended like .05 seconds before another line started. .5 seconds is more than enough time to read a "Hai"

To illustrate what I mean...

...Hello world.It's a new day...<--non-blip

...Hello world...It's a new day...<--blip

Its when the first line "Hello world" disappears for a brief moment and the new line "It's a new day" appears. Whereas the in the non-blip version, the new line appears right on the next frame so you can continue to read without losing a thought.

Personally, I think that doing it in that method can certainly help readability for sentences that are too long to fit on two lines and the thought continues on the following subtitle. However, there are drawbacks to this method...you have to check for scene changes...which is even more annoying when you switch raws and have to check them all again. Also, chaining together different character's dialogue can be thought of as one character's thought...It also annoys me sometimes when there are dialogue subtitles on the screen and no one is talking. But of course sometimes it is necessary for the shorter lines.

I'm ranting too long...maybe I'll finish this reply later...or someone will finish it for me.

Shii
2003-11-11, 23:04
ashibaka: what are you talking about? xombie never even finished developing it.

So? [shrug] It still beats SSA for positioning subtitles, etc.

Sakaki-
2003-11-12, 00:33
ashibaka

VdubMod + SSA + Notepad beats that in efficency easy ;)

Kasshin
2003-11-12, 00:43
I personally like notepad or any text editor for positioning / typesetting stuff. Xombie also lags too much on my kind-of-slow computer for timing.

As for bleeds and blips, as a timer I would feel that I didn't give it my all when timing the episode if it had bleeds and blips in it. The main reason I check for them is so that I will be happy with my own work; so I can watch something I worked on and be satisfied. It really doesn't matter whether the viewers notice it or not, I'm doing it so I won't get annoyed at it when watching.

Shii
2003-11-12, 09:49
I personally like notepad or any text editor for positioning / typesetting stuff.

That's probably more accurate :)

runpsicat
2003-11-12, 10:26
I guess I was not specific enough in my post about what I meant. I fully agree that one should not do a sloppy job (QCing your own timings is a must to pick up stutters and bleeds), but how much time one should sink into one's part in fansubbing always seems to be a matter of debate. The ideal of making something "as perfect as you could possibly can" is always tempered by practical concerns, and people tend to label such pursuits "obsessive" once it crosses a certain line. The issue is that individuals have differences in opinion regarding what they consider to be the threshold, and that was essentially what I was trying to point out (aside from trying to ask for possible exceptions to the stated guidelines). Now that I think about it, bringing up an issue without a real solution was not a good idea :heh:. Still, staff frequently criticize whether fellow staff (and staff in other groups) spend too much/little time on his/her step, and subbing angst can consequently ensue... In that sense, I always find it interesting what other people find "acceptable" and such. Eh, my apologies for asking ambiguously worded questions in my previous post and for making this post rather general.

Sakaki-
2003-11-12, 15:45
I guess I was not specific enough in my post about what I meant. I fully agree that one should not do a sloppy job (QCing your own timings is a must to pick up stutters and bleeds), but how much time one should sink into one's part in fansubbing always seems to be a matter of debate. The ideal of making something "as perfect as you could possibly can" is always tempered by practical concerns, and people tend to label such pursuits "obsessive" once it crosses a certain line. The issue is that individuals have differences in opinion regarding what they consider to be the threshold, and that was essentially what I was trying to point out (aside from trying to ask for possible exceptions to the stated guidelines). Now that I think about it, bringing up an issue without a real solution was not a good idea :heh:. Still, staff frequently criticize whether fellow staff (and staff in other groups) spend too much/little time on his/her step, and subbing angst can consequently ensue... In that sense, I always find it interesting what other people find "acceptable" and such. Eh, my apologies for asking ambiguously worded questions in my previous post and for making this post rather general.

Well theres a limit to what time i would put down on what i do, And i must say i always find something i could have touched up a bit.
Theres always a expectation to the rule or the guidlines.
Fast Blips sux, easy to avoid
Blips with a medium space i think is good depending on how long the line before is.

I give it my best whatever i do i cant live with doing a half ...ed job.
Even if it's kind of a hobby.

" Still, staff frequently criticize whether fellow staff (and staff in other groups) spend too much/little time on his/her step, and subbing angst can consequently ensue... "

Well i admit i have got alot more sensitive to how stuff looks now then before i started timing, That includes every step.

Take Care
Sakaki-

Sakuya
2003-11-13, 04:19
I have a problem. In Vdub, after I select the SSA subs and click Ok, it won't play my subtitles on the video! When I press play, only the raw plays in the input video display and the output display doesn't move. When I click stop, the output display image suddenly changes to the frame where I clicked stop. And the subtitles are there in the output display. How do I get it to play along with the subs?

Enragin_Angel
2003-11-13, 04:31
Encode it. But when you preview in vdub you're only supposed to press right or left (or shift+right or shift+left to snap to keyframes). that is the only way to get the output without having to waste time encoding it.

TaMz
2003-11-13, 04:38
I have a problem. In Vdub, after I select the SSA subs and click Ok, it won't play my subtitles on the video! When I press play, only the raw plays in the input video display and the output display doesn't move. When I click stop, the output display image suddenly changes to the frame where I clicked stop. And the subtitles are there in the output display. How do I get it to play along with the subs?
right click on the 1st screen (RAW screen) and choose 1/4 size...
then (or before that) put VDub Maximized...
then push the second play button at the bottom (the second play button is "play output video" or something like that)...

Videric
2003-11-13, 04:41
Sakaki min soetnos :love:

TaMz
2003-11-13, 04:48
Oh ya... and a better way for previewing subs than VDub is using MPC (Media Player Classic)...
I use it...
Works like this:
1 get MPC and VobSub
2 put the video and subs in the same folder
3 give them the same name (example: video.avi & video.ssa)
4 open MPC
5 load the video file
6 watch the video
;)

Sakaki-
2003-11-13, 08:41
I have a problem. In Vdub, after I select the SSA subs and click Ok, it won't play my subtitles on the video! When I press play, only the raw plays in the input video display and the output display doesn't move. When I click stop, the output display image suddenly changes to the frame where I clicked stop. And the subtitles are there in the output display. How do I get it to play along with the subs?

Enable DirectDraw Acceleration.
You can also change the size of the original video to the smallest ;)
And it schould be fixed

Use the |>. button to play.

Take Care
Sakaki-

zalas
2003-11-13, 17:26
Oh ya... and a better way for previewing subs than VDub is using MPC (Media Player Classic)...
I use it...
Works like this:
1 get MPC and VobSub
2 put the video and subs in the same folder
3 give them the same name (example: video.avi & video.ssa)
4 open MPC
5 load the video file
6 watch the video
;)

Um... Directshow isn't proven to be frame accurate. :/
However, DVobSub will let you shift subtitles, which may be good for some instances.
Besides, seeking around gets annoying.
For virtualdub, you want to do:

1) load up your filters
2) disable "Display input video" in options
3) enable "Swap input/output panes" in options
4) press |>O (that is the O play button, not the I play button)

Animaniac
2003-11-13, 18:35
SSA is ancient; its features reflect that it was made for the era of VHS fansubs. I like XombieSub, which has the nice feature of being able to see the video you're editing as well as hear the audio.You'd probably like USF (http://usf.corecodec.org/). There are two (http://corecodec.org/projects/u96/) tools (http://corecodec.org/projects/use/) out there so far. They aim to render SSA/ASS ancient. Gabest has had a lot of influence on the specs too, and he has had USF support in MPC and VSFilter for a long while now.

TaMz
2003-11-13, 19:12
Um... Directshow isn't proven to be frame accurate. :/
However, DVobSub will let you shift subtitles, which may be good for some instances.
Besides, seeking around gets annoying.
Well, my method is SSA -> watch with MPC -> use Notepad to change timings
dunno about that frame accuracy thing, but at least this works really fine for me...

dmhypopgo
2003-11-14, 09:47
All u need is SUBCREATOR, yeah, thats all u need. open a avi file and add timing.

Sakaki-
2003-11-14, 11:34
All u need is SUBCREATOR, yeah, thats all u need. open a avi file and add timing.


Subcreator is for *.srt :O

zalas
2003-11-14, 13:29
You'd probably like USF (http://usf.corecodec.org/). There are two (http://corecodec.org/projects/u96/) tools (http://corecodec.org/projects/use/) out there so far. They aim to render SSA/ASS ancient. Gabest has had a lot of influence on the specs too, and he has had USF support in MPC and VSFilter for a long while now.

When do you suppose they'd be able to add in the 'effects' implementation? That's the only thing I'm using in ASS (which is kinda broken for some effects, the interpolation is horrible for \t()) which cannot be done in USF at the moment.

SirCanealot
2003-11-14, 19:39
Heh. I actually just watched a episode I timed properly for the first time (eg, in my sofa on my TV-Out). I think I'm going to try and standardise my lead-out times. My subs were random in where they would go out. It looked a bit crapy when you had your eyes glued to the bottom of the screen. I think I'll try 0.20 for normal subs and 0.40 for those crapy short subs and only go off this if a setance is weird...

Anyone reading this about to time for the first time could try and follow this... my subs definately did look too random... and my dialog needs work too... gah...

Ramza
2004-05-15, 07:46
Hi, i've got a question. Can anybody tell me of a site which has tutorials for some cool sub effects, such as those in openings or endings? Dissapearing, 'moving' text etc. I'd appreciate any help :)

Enragin_Angel
2004-05-15, 14:18
Sure, download vobsub 2.23 and look in the docs folder. You'll find ass-specs.doc and assquickref.txt or something along the lines of those two filenames. Read those two docs and they will tell you how to do all the commands you've listed. Although, the creativeness is up to you.

Ramza
2004-05-15, 15:11
Thx man, like u said - there's all i want ^^ Big thank you once again.

clem-kun
2004-05-30, 22:26
"No audio decompressor could be found to decompress the source audio format"

=> That's what VirtualDub says when I try to save the .wav file of an .avi raw... what do I do?

getfresh
2004-05-30, 23:38
"No audio decompressor could be found to decompress the source audio format"

=> That's what VirtualDub says when I try to save the .wav file of an .avi raw... what do I do?


Do you have the ac3dec decoder?

curlyconnor
2004-05-30, 23:48
What I do is I take SSA and VDub (equipped with Avery Lee's Subtitle Filter), and then play the movie through on VDub, stopping where the beginning and end of dialogue is (being extremely precise of course) and copying down the time displayed on the counter into SSA (rounding to the nearest 1/100th of a second, which is as precise as SSA gets). Then I watch through it, and if something's bad, I redo it. This works perfectly. I don't save the audio as a WAV and just listen to that, I watch it just like I would regularly, just on a program that has a very, VERY precise counter :heh: . I don't have to worry about the scene changes this way.

getfresh
2004-05-31, 03:42
The fastest and most precise way of going about it is still by timing using waveforms as ref. Timing using a linear method is obsolete and has been for some time. Also having to stop to type in values is a waste of time when you can just use the capture button in the time from wave mode within ssa. Not tring to flame you or anything, but why in hell are you using that method? If you can explain to me in a convincing manor why that method is useful then more power to you. I say this because I've been timing for quite awhile and never seen a benefit to linear sub time capture...

SirCanealot
2004-05-31, 06:57
Yup, I don't mean offense, but I'd have to say someone that doesn't use some form of audio (bet it .wav for SSA, for the original mp3 for other sub programs) to be completely insane - far that which higher than timers usually are :P
SSA's set out pretty simply, so you can get good speed on rough timing using a wave file, so I don't see any reason why anyone'd want to do something different...?

clem-kun
2004-05-31, 09:54
Do you have the ac3dec decoder?

Where do I get that?

getfresh
2004-05-31, 09:57
just google for it, it's not that hard to find.

clem-kun
2004-05-31, 09:58
just google for it, it's not that hard to find.

Yeah... I got it already. Thanks for your help! :)

getfresh
2004-05-31, 10:11
np, if that wasn't it tell me and I'll see if what else it may be ^^

clem-kun
2004-05-31, 10:34
Okay, it wasn't it. I think Ac3Dec only works with .vob files, which I do not have. I have an .avi file...

Is there some filter for VirtualDub that can decompress audio in .avi files?

curlyconnor
2004-05-31, 11:52
The reason why is because I didn't know you could do that on SSA (I'm still fairly new to this). Now that I do, I feel like banging my head into my keyboard, typing in every one of those values. I've gotcha now, but what's that thing about scenechanges?

getfresh
2004-05-31, 13:14
http://www.evabeast.com/flf/wbboard/thread.php?threadid=10&boardid=10&styleid=1

read section 3 of this guide for an overview on scene change overlaps ^^

StormD
2004-05-31, 22:57
The fastest and most precise way of going about it is still by timing using waveforms as ref. Timing using a linear method is obsolete and has been for some time. Also having to stop to type in values is a waste of time when you can just use the capture button in the time from wave mode within ssa. Not tring to flame you or anything, but why in hell are you using that method? If you can explain to me in a convincing manor why that method is useful then more power to you. I say this because I've been timing for quite awhile and never seen a benefit to linear sub time capture...
I like the results I get from using Subtitle Workshop to capture timings. I can see the benefit of having a visual waveform queue, but honestly, you can get just as good by using your ears and the capture button while watching the video (especially since SW lets you play at reduced speed), and tweaking times up and down slightly to correct irregularities, and get sub in/outs exactly on key-frame changes when necessary. You get the best of both worlds, in that you can click to capture/adjust timings, while still getting to watch the video and use on-screen queues effectively. Once I'm done with SW, All I have to do manually is define my styles (SW doesn't support multiple-styles in output).

Of course, this method is no good for karaoke if you're into that (which I'm not), but for basic dialogue timing, and especially for things like on-screen text that is totally reliant on what's in the video display, I think it's currently the best solution out there.

ichido reichan
2004-05-31, 23:32
actually timing for me become even more fun, I hook up my genlock and I do the timing "on the fly" that is watching the movie in real time and turn on-off the subtitles depending on what character is talking.

after I got a "beta timing" I go back and check those missing lines and adjusting the actual subtitle until is satusfactory.

when I finish the script, the last thing I do is put the credits on it.

I can understand those fancy karaokes that everybody likes nowadays, I think those are far distracting with all the funky colors and effects, it gets mixed up with the opening and finally you never cant catch the words to sing at the same time.

SirCanealot
2004-06-01, 07:00
Capturing the timings "on the fly" like that, is in no way going to be as fast or as acurate as using the wave file. Human reaction time hangs around 25-30ms (microseconds) on average. So, you play the file back at half speed, you half the problem with human reaction time down to around 10-15ms, and you double the ammount of time it takes to time it - therefore completely getting rid of any advantage in doing it this way. Add to that - I know I'd notice if there was a 10ms delay before a subtitle came up whenever someone spoke. Anything other than grabing your start times with a wave is going to be worse imo. The wave allows you to(usually) grab the start time within a few ms, not 10-30.
And going back and having to tweak every other subtitle's start time is either going to be even more work than fine timing usually is, or look horrible (as the timing all of ichido reichan's fansubs does imo).

exedore
2004-06-01, 07:09
Capturing the timings "on the fly" like that, is in no way going to be as fast or as acurate as using the wave file.

Quoth someone who's never seen a tape from Central Anime. Every one of their subs was done on an Amiga (no wave capabilities, 20MB hard drives) and timed by Todd Perkins sitting in front of the LD player with a joystick. Sure it took him 2-3 passes to get it done, but Central never was (is?) about speed subbing, just quality scripts. And damned if their scripts aren't quality.

StormD
2004-06-01, 09:07
Well, my standard procedure allows for the delay in my reaction time. As soon as someone starts talking, I click on the video screen to pause it, click 'set start time' and then click the down arrow on the actual time drop once or twice depending on how long it seemed I missed the timing by (each click adjusts .1 seconds), then click on the video to re-start it. Rinse repeat for closing the subs. I'd say it probably takes the same amount of time as clicking through a .wav file and setting the times that way. I'm not claiming to save time in the actual timing, per se, but I do think I'm cutting some corners in the cleanup process (because I can check as I go that I'm not overlapping key-frame transitions, as well as being able to time on-screen text accurately during the same pass. Plus, I don't have to futz with extracting the .wav file in the first place. On average it takes me about 2h to completely time (and by completely, I mean good enough to pass QC) a 30m episode.

SirCanealot
2004-06-01, 11:05
And how long did it take Mr Perkins? I'm not saying it can't be done, it's just that a wave file is the faster of the methods of rough timing anyway.

It takes me about 2-4 seconds to time a short line in SSA - identify and select start point, end point and grab times. The same on the longer lines, except I actually have to listen to the line. And that's a very acurate grabing of the start and whatever end times you want to give a sub. There's no way you're matching my speed doing it any other way. A rough time should be perfectly rough - it's not about having to adjust anything up or down when you're just rough timing it.

Of course, in a race you'd still beat me, since I take AGES in fine timing, and like to get things timed PERFECTLY (which is something I give great respect for when I see other groups do it) - which accounts for a lot of crap I don't have names for, not just checking for scene bleeds and subtitle blinks. But that's just me. I'm stupidly fast when I want to be when rough timing.

Edit: Please don't take this as me saying I'm better than you, I'm not.
I have to say timing is about finding a method that most suits you. I have done, and being the type of PC user I am, I couldn't see any other way being better, apart from taking short cuts. If you time sitting upside down, attached to the ceiling, then bloody well so be it! :)

zalas
2004-06-01, 11:29
And how long did it take Mr Perkins? I'm not saying it can't be done, it's just that a wave file is the faster of the methods of rough timing anyway.

It takes me about 2-4 seconds to time a short line in SSA - identify and select start point, end point and grab times. The same on the longer lines, except I actually have to listen to the line. And that's a very acurate grabing of the start and whatever end times you want to give a sub. There's no way you're matching my speed doing it any other way. A rough time should be perfectly rough - it's not about having to adjust anything up or down when you're just rough timing it.

If someone is timing using a genlock in real time, he'll definitely beat you in rough timing speed for shows with a moderate to heavy amount of dialogue. This is because the shorter lines he'll get through real time, and for the longer lines you'd spend at least the amount of time he's spending, since you have to listen to everything anyways.

SirCanealot
2004-06-01, 11:36
He can't time it in real time though - he has to hit the button when they start talking, hit it when they stop, then at somepoint they need to ajust the timing backwards to account for human reaction time. Even seasoned gamers I know don't go under 15ms when tested :P

Mr_Paper
2004-06-01, 11:47
Just to play the devil's advocate for a moment...

You're assuming he's timing it in real time using only the audio for the timing ques. It would be safe to assume that he might also be using video ques to aid in the timing. Particularly with older series, as the mouth tends to open about 13ms before the coresponding audio is played. If he's fast enough to pick up on that, the rough timing can still be fairly accurate. Seasoned gamers might only have reaction times of 15ms, but I'm sure their ability to predict timings is far more accurate. =P

ichido reichan
2004-06-01, 13:21
Just to play the devil's advocate for a moment...

You're assuming he's timing it in real time using only the audio for the timing ques. It would be safe to assume that he might also be using video ques to aid in the timing. Particularly with older series, as the mouth tends to open about 13ms before the coresponding audio is played. If he's fast enough to pick up on that, the rough timing can still be fairly accurate. Seasoned gamers might only have reaction times of 15ms, but I'm sure their ability to predict timings is far more accurate. =P

actually timing in real time is preety cool, not because you are following the story, but because you have to listen to the breath and the animation to know when they will stop and when the next person start talking.

actually on SSA I use 2 buttons instead of one, the start button (space) and the tab button (end) but after finishing timing, when you do a mistake you, you have to go back and time it again with the wav file, I usually throw the script backwards
50 ms (the natural delay that has been talked about) and after that I adjust the lines that were out of the context.

One of my rules is:

-you must leave some time to people to read the subs

-the subtitle cant be as short as 90 ms or as long as 5-6 seconds

-lines must be mandatory 2 per page, 3 lines on a screen is distracting.

-subtitle must have at least 10-10-30 (wide and height) or 10-10-35 on digisub

-Script must be corrected at least 3 times.

ff7799
2004-06-01, 15:50
Mods This topic needs to be made into a sticky.

LytHka
2004-06-01, 23:02
Sakaki-'s way is still the best way of timing anime, no matter how you look at it. o.o

zalas
2004-06-02, 17:48
He can't time it in real time though - he has to hit the button when they start talking, hit it when they stop, then at somepoint they need to ajust the timing backwards to account for human reaction time. Even seasoned gamers I know don't go under 15ms when tested :P
Reaction lag is usually around the same amount, especially if it's a repetitive task, so he/she can just brush through once timing real time and then tell SSA to shift everything earlier by say 200ms, depending on his/her reflexes.

ZeppelinJ0
2004-06-18, 00:09
"(Note: If you get a warning that the video stream uses vbr audio redo this step selecting "Nandub VBR MP3 compatibility mode”)"

Where do i find this???

Secret Squirrel
2004-06-20, 05:50
....their viewing experience is then compromised by the fact that they start noticing all these "glitches" they were not aware of before (the fate of all QCers, apparently).

..tell me about it! :twitch: I used to enjoy most groups fansubs without complaint. Now I can't help but automatically QC every damn ep I watch

..and as a "trainee timer" teaching myself, this thread is one of the best reference guides out there! especially with people posting their "pet hates" to watch for things to avoid.

SirCanealot
2004-06-20, 08:59
..tell me about it! :twitch: I used to enjoy most groups fansubs without complaint. Now I can't help but automatically QC every damn ep I watch



Heh, that allows me to really appreciate a quality product though - whenever they do come along.

Imaginer
2004-07-31, 06:39
I like the results I get from using Subtitle Workshop to capture timings. I can see the benefit of having a visual waveform queue, but honestly, you can get just as good by using your ears and the capture button while watching the video (especially since SW lets you play at reduced speed), and tweaking times up and down slightly to correct irregularities, and get sub in/outs exactly on key-frame changes when necessary. You get the best of both worlds, in that you can click to capture/adjust timings, while still getting to watch the video and use on-screen queues effectively. Once I'm done with SW, All I have to do manually is define my styles (SW doesn't support multiple-styles in output).

Of course, this method is no good for karaoke if you're into that (which I'm not), but for basic dialogue timing, and especially for things like on-screen text that is totally reliant on what's in the video display, I think it's currently the best solution out there.

I have just started to play around in the subbing and timing field and yes I do agree. SW is better in getting down the times of the subs and alot better interface for that as well as you can not only listen, but view the video as well. (though lacks the styles and karaoke stuff of SSA)

SW also allows an ADJUSTABLE WINDOW SIZE! SSA however, it likes to stay at full screen. Grrrrrr....

Sakaki-
2004-08-01, 17:32
I have just started to play around in the subbing and timing field and yes I do agree. SW is better in getting down the times of the subs and alot better interface for that as well as you can not only listen, but view the video as well. (though lacks the styles and karaoke stuff of SSA)

SW also allows an ADJUSTABLE WINDOW SIZE! SSA however, it likes to stay at full screen. Grrrrrr....

Lets see here now, Timing a 24 minutes episode of abut 300 to 600 dialogue lines takes about 27 minutes or so from a wave file, this is if you just concentrate on the timing and not anything else, this can also go faster depending on the layout of the episode.

Then fine timing "Scene timing" takes about 24 minutes, that gives a time of 51 minutes or so, though on this step you can do a little trick, during the time you wave time encode a lq rip of your raw and boost it up with more keyframes.
Then you can jump keyframes and listen to every line you see pop up on screen at keyframes if they continue or should be cut. (Note i find this style very borring and its not flawless, by not being flawless i mean you cnat check your timing during this pass wich i think is very important.)

So about 37 minutes would be a good speed if you optimised your timing using substation Alpha and Virtual dub to achive your goals.

d-hogan
2004-08-01, 18:49
I apprised Todd of this thread and this is his statement:

>From Todd:
A friend of mine first formats the script before I see it. She catches
stupid things and puts it in a format I can use to time. Then sends it to
me. All this while I work on other shows.

I time in real time using both video and audio. So a 25 minute show takes 25
minutes to time. I correct my human error factor of ears to fingers and run
a timing bar tape with the script running in real time and record it (25 min
more). I then go through line by line and fix grammar, timing errors, split
lines up, place text and do credits(usually cast only and song credits). The
editing takes anywhere from 2-3 hours for a 25 minute show depending on
quality and amount of dialogue. LOGH takes a lot longer to edit than, say,
Initial D. I then watch it without the timing bar with friends holding
paper and pencils to catch errors in grammar and timing(25 min more). I may
watch it twice.

25 + 25 + 2hr + 25 = 3hrs 15 min for a 25 min show. Add variable time and
it's 4hrs 40min.

Todd Perkins
Central Anime
(Currently deployed)

crumja
2004-08-11, 01:53
[QUOTE=Sakaki-]Lets see here now, Timing a 24 minutes episode of abut 300 to 600 dialogue lines takes about 27 minutes or so from a wave file, this is if you just concentrate on the timing and not anything else, this can also go faster depending on the layout of the episode.[\QUOTE]

I challenge that. For something with < 300 lines it might be possible if you click fast enough to catch the beginning and end of the line, but it won't be (unless your timing is off) for a 300-600 lines script. I'm sure most fansub timers will agree with me on that.

LytHka
2004-08-11, 09:23
Lets see here now, Timing a 24 minutes episode of abut 300 to 600 dialogue lines takes about 27 minutes or so from a wave file, this is if you just concentrate on the timing and not anything else, this can also go faster depending on the layout of the episode.

I'm not sure what you're talking about here. My biggest guess is that you're talking ideal conditions, meaning flawless script, all the line breaks done by the editor, ... etc. Also... 300 -> 600... There is quite a difference between a 300 and a 600 line script. Unless you have some nice timing tools besides SSA (I really can't imagine how they'd work other than voice recognition, maybe something similar to xzombie and even something like that won't help you with overlapping lines) or some special human ability to do this fast, there is probably no way to have a 300+ line script rough timed nicely under 30 minutes. I challenge you on that one too.

zalas
2004-08-11, 10:22
I'm not sure what you're talking about here. My biggest guess is that you're talking ideal conditions, meaning flawless script, all the line breaks done by the editor, ... etc. Also... 300 -> 600... There is quite a difference between a 300 and a 600 line script. Unless you have some nice timing tools besides SSA (I really can't imagine how they'd work other than voice recognition, maybe something similar to xzombie and even something like that won't help you with overlapping lines) or some special human ability to do this fast, there is probably no way to have a 300+ line script rough timed nicely under 30 minutes. I challenge you on that one too.
300 lines in 30 minutes is 6 seconds per line
600 lines in 30 minutes is 3 seconds per line
both are doable theoretically, as long as there's not much background noise and that the timer doesn't make a mistake in breaking the lines. If he has to play thru all the excerpts, then he has about 1 second per line of clicking/GUI time for 300 lines and 1/2 second per line for 600 lines. Thus, not much room for mistakes ^^;

LytHka
2004-08-11, 10:52
300 lines in 30 minutes is 6 seconds per line
600 lines in 30 minutes is 3 seconds per line
both are doable theoretically, as long as there's not much background noise and that the timer doesn't make a mistake in breaking the lines. If he has to play thru all the excerpts, then he has about 1 second per line of clicking/GUI time for 300 lines and 1/2 second per line for 600 lines. Thus, not much room for mistakes ^^;

Forgetting a couple of things here:

MOST anime episodes are 21-22 minutes long (excluding the OP & ED), so the lines would be even more crammed together in this example. More lines in a script doesn't really mean that the there's a lot of short lines usually, it means there are many overlapping lines, meaning there's A LOT of background noise when timing one line. Considering that distracting normal human senses, timing slows down. "theoretically" is another word here saying a human should have complete focus on the wav AT ALL TIMES. Personally I think that's impossible.

Another point I'd like to make. Most timers aren't experts in Japanese, I know I'm not, still I do have a good sense, knowledge, when a certain line starts and ends. I see "bad" timings from reputable groups all the time, not mentioning some names that are present in this thread as well and a good part of those timings has something incommon with reckless timing.

(Yes, I would like to see someone rough time a 300+ episode under 30 minutes with their own timing style not getting bad.)

SirCanealot
2004-08-11, 11:30
300 lines in 30 minutes is 6 seconds per line
600 lines in 30 minutes is 3 seconds per line


I'd just like to see a video of someone keep up an insane pace like that for 30 minutes, heh.

Heibi
2004-08-11, 11:54
I'd just like to see a video of someone keep up an insane pace like that for 30 minutes, heh.


Well, I keep up with the LOGH pace for 55 minutes at a time. And their scripts are in the 700+ lines category, which increases after editing via line splits. Of course I time in real time as we all know. When I get the overlapping dialog problem I make the lines invisible and manually change the times to reflect their true timing.

I love timing in real time.

SirCanealot
2004-08-11, 12:28
Timing in real time with a Genlock type thing (I assume that is what you mean) isn't as tiring as:
Left click (start time), right click (end time), left click on grab times - six times per second.

I play RTSes - I know how much pressure that would put on a wrist :P

Kanna
2004-08-11, 12:40
Lets see here now, Timing a 24 minutes episode of abut 300 to 600 dialogue lines takes about 27 minutes or so from a wave file, this is if you just concentrate on the timing and not anything else, this can also go faster depending on the layout of the episode.
...
So about 37 minutes would be a good speed if you optimised your timing using substation Alpha and Virtual dub to achive your goals.
I'll also have to call BS on that. 27 minutes to time an episode in a neat and correct manner is impossible. Granted if the episode had no music, just straight speech and everyone's talking one at a time yes, it is likely 27 minutes will get the job done. If you want to do it right, as in checking the times and adding leadin/leadout times, it'll take much longer. Hell I've been timing for 4 years and I still spend 40 minutes per 230 line episode. Speed timing just leads to errors and inconsistencies, sure I can time an ep in 10 minutes flat, I just won't guarantee it's done with care or accuracy.

Heibi
2004-08-11, 13:03
Timing in real time with a Genlock type thing (I assume that is what you mean) isn't as tiring as:
Left click (start time), right click (end time), left click on grab times - six times per second.

I play RTSes - I know how much pressure that would put on a wrist :P

You're right. I up arrow key down arrow key for the titles. I right click to back up during my timing run to undo a title(handy little feature) if I jumped the gun.

Sylf
2004-08-11, 20:31
I think I've timed something just short of 200 lines (one of the LastEXILE) in just over 30 minutes. Since I translated the episode too, I knew line breaks etc pretty well. :) (rough time only)

Kawaii_tsunami
2004-08-24, 22:58
im a bit unclear about how the audio and the script goes with the video in the guide it doesnst really make a clear statement how and when...im sorry..but im a newbie to all this, can someone explain?

Darth_E_
2004-08-24, 23:09
im a bit unclear about how the audio and the script goes with the video in the guide it doesnst really make a clear statement how and when...im sorry..but im a newbie to all this, can someone explain?

?

It's simple. You have two methods to time scripts :

1.Rip a 8bit mono PCM audio file from the raw, load it on SSA and time with it
or
2.Load the video in subtitle workshop and time according to when they move their mouths.

Difference between boths? Method 1 is more accurate , but you have to go over the script twice and check for scene bleeds. Whereas Method 2 is less accurate, but you dont have to worry about scene bleeds alot since you can see when do the frames change in SW.

I personally use and recommend the first method.

Yours,
-Elly

Kawaii_tsunami
2004-08-24, 23:11
ok thank u

Sakaki-
2004-08-25, 07:35
I'm not sure what you're talking about here. My biggest guess is that you're talking ideal conditions, meaning flawless script, all the line breaks done by the editor, ... etc. Also... 300 -> 600... There is quite a difference between a 300 and a 600 line script. Unless you have some nice timing tools besides SSA (I really can't imagine how they'd work other than voice recognition, maybe something similar to xzombie and even something like that won't help you with overlapping lines) or some special human ability to do this fast, there is probably no way to have a 300+ line script rough timed nicely under 30 minutes. I challenge you on that one too.


Yeah i guess its ideal conditions when no one botheres you and no that is accounted for with some splits, and i tend to think that its better with a script where all lines are clumpred together then one where its not.

And i beg to differ 30 minutes is more then enuff if you sit down and seriously do the rough timing.

Kawaii_tsunami
2004-08-25, 15:40
i dont understand what a bleed or a stutter is quite clearly..and i cant seem to find any...is there a certain way to watch the sub??

this is a great guide too^^

Sylf
2004-08-25, 16:55
Bleed : When a character speaks a word, the sub for it appears on the screen, then the scene changes quickly, but the sub remains on screen. On some subs, this happens more often, because the subtitle creator sticks to the policy that the sub should remain longer than (x) seconds, no matter what. Some other people can't stand that kind of bleed over...

zalas
2004-08-25, 17:12
Bleed : When a character speaks a word, the sub for it appears on the screen, then the scene changes quickly, but the sub remains on screen. On some subs, this happens more often, because the subtitle creator sticks to the policy that the sub should remain longer than (x) seconds, no matter what. Some other people can't stand that kind of bleed over...
Generally, bleeding over is a problem if the sub is running over into the next scene for a really small amount of time. This gives certain people a weird feeling because it's almost close enough for it to be considered one event, but the scene change the subtitle end are far enough to be a bit annoying. If something _has_ to bleed over (like if the speaker almost carries over into the next scene), usually it's better to extend his/her lines so that there is at least a good 1/2 second intrusion into the next scene.

Kawaii_tsunami
2004-08-25, 18:35
thank you zalas and sylf!!

Ramza
2004-08-29, 09:26
Hi guys, i've got a strange problem. I subbed FMA ep.44 , saved the ssa file, opened vdub, raw file and added the subs using filter. And when i checked compressed file it was totally crapped. Quality went way down, the whole movie was playing as in slow-mo and everything just sucked. I tried divx and xvid compression and nothing. Could you help me in any way?

zalas
2004-08-29, 09:51
Hi guys, i've got a strange problem. I subbed FMA ep.44 , saved the ssa file, opened vdub, raw file and added the subs using filter. And when i checked compressed file it was totally crapped. Quality went way down, the whole movie was playing as in slow-mo and everything just sucked. I tried divx and xvid compression and nothing. Could you help me in any way?
Looks like you might have left compression settings to some default, which included a low bitrate. I would recommend checking that by going to the select compression dialog box and configuring your encoder. For more information, visiting www.doom9.org is a good first step. Check out their guides on encoding.

Kawaii_tsunami
2004-08-30, 13:38
I have another question....when i play the file i just timed it turns kinda black and the sound kinda stops and starts and stops and starts..etc..
does anyone know y it happens?? or how to fix it??

zalas
2004-08-30, 14:10
I have another question....when i play the file i just timed it turns kinda black and the sound kinda stops and starts and stops and starts..etc..
does anyone know y it happens?? or how to fix it??
I don't know if you're playing this thing in a media player with directvobsub, or in Virtualdub using filters or playing the rendered output. If you're in Virtualdub, you might have problems playing things that are encoded in VBR MP3 audio. You might also have problems if your computer simply can't render the video fast enough.

Sylf
2004-08-30, 14:48
You must be using the playback feature of Sub Station Alpha. And if that's true, then that's not a bug, but it's a feature. The audio is audible only while there are subtitles showing. Use VirtualDub(mod) and TextSub filter, so you can actually see the sub in the picture.

Kawaii_tsunami
2004-08-31, 12:06
ok im going to try it.thank u

Kawaii_tsunami
2004-09-06, 12:02
I have another question...
when i finish timing with vdub, and saving it, it goes at 10 fps
is that good or bad??
is that y my video is sometimes scratchy?

and is 10 fps suitable for a timer??
and the bitrate of the video is somewhere near 840.
is 840 too slow?
thank u!! :help:
:hmm:
:stupid:

Ryo-oh-ki
2004-12-21, 18:28
I just thought I'd throw another perspective on this subject.

I also time show on-the-fly and have done so for a number of years. (a few hundred eps)

I personally use a minimum time of 0.70 seconds for each subtitle (except for some strange circumstances) and try to add another 0.20 or 0.30 to the end of normal subtitle lines. I also try to not leave less than a 0.50 sec gap between subtitles. I've gone back and forth on whether or not to cut lines before a scene change. Nowadays I think it's more effort than it's worth, but it does indeed look nicer when it's done.

One thing I've noticed is that the smallest difference that can actually be noticed in timing is about 2 frames worth, unless you're talking about captions. 2 frames is about 0.06666 seconds or 6 2/3 hundredths of a second. And even then, most "normal" (not fansubbers) people can't tell if a subtitle is wrong unless it's almost 0.20 seconds off.

And since I work doing subtitles with a friend of mine, I've also noticed one other thing that is somewhat obvious if you think about it. The beginning and ending times of spoken dialogue in a show is judgemental. We sometimes have disagreements about when a line should begin or end. Sometimes it comes down to what someone can hear versus what someone else can hear. Other times it comes down to preference as to if you want it to begin with a breath-like sound or with a harder sound or whatnot. And finally, different audio systems may make lines sound early or late depending on how it was timed.

Anyway, I've found that for me, the thing that takes the most amount of time is fixing the script to sound nice compared to what was said in japanese. I spend more time doing that than the timing effort I think.

For a normal TV series episode that's about 25 mins, I typically:

1) Edit the script ahead of time to fix any glaring mistakes, break up lines (guessing), make changes to the script to conform to my own style.

2) Time the episode "on-the-fly" 3 times, saving the last time. The first and second times through, I'll pause my source video and write down any lines which I want to change before doing the last timing and fix it in the script. There's not typically that many, but bad scripts can generate a lot of changes.

3) I run off the timed script with my subbing program and have it display a clock on the screen to work from when editing.

4) I sit down with a friend and watch the show, pausing, rewinding, and fixing any problems we encounter. I call this the "edit." Again, we typically end up spending more time fixing the script words than the script times.

5) Repeat step 3.

6) Repeat step 4. I call it the "proof."

7) I add in any credits or anything else that is necessary.


I think the method I use to time is just as accurate as someone who times with wave timing. I had arguements about this when .wav timing was first time, and I'm still sticking to my guns. Someone can time very poorly using .wav timing, just as with on-the-fly timing, and someone can time well with both methods. It just depends on what suits you more. In both cases, to get accurate times, you have to go and edit/proof the timings. I just can't imagine looking at some .wav file for as long as some people do. It suits me more to just watch and enjoy the show a few times in a row.


Incidentally (sp?), the fastest I've timed something on-the-fly was just timing every episode once. I did this for an LD box set I own for my own viewing purposes, and it worked out well enough to watch. It had 70-some episodes in it and I occassionally screwed up big time, but in those cases, I just fixed the time the next time I watched it. I still think I did a better time than some of the old Arctic subs :) But you can't beat that for quickness in timing, if you just want a rough set of times.

Kyle

FATa|
2004-12-22, 11:32
Hi guys, i've got a strange problem. I subbed FMA ep.44 , saved the ssa file, opened vdub, raw file and added the subs using filter. And when i checked compressed file it was totally crapped. Quality went way down, the whole movie was playing as in slow-mo and everything just sucked. I tried divx and xvid compression and nothing. Could you help me in any way?
The most major reason for this happening (it happens to me and makes sorting out scene bleeds a bitch) is because of your computer specs. Its probably tooo slow, or too overloaded with other programs. I checked this theory by running a subtitling filter on vdub on my computer with regular programs running. then I went on my mums new computer and di the same thing with same files and it was very much faster (like i said new comp, better specs) then i went back to my comp and got rid of everything i didnt need off the system control manager; sysreset.exe, msnmsgr.exe, xxxtools.exe (lol - damn bug) and everything else, you could probably even get rid of explorer if you needed to. this helps speed htings up. then you need to optimize the settings in vdub. Here is the process i go through for optimum performance:

1) minimize input video to 25%

2) video has to stay on full proccessing mode but audio can be set to direct stream copy

3) stop input video from playing (f9 or options.display input video)

4) Instead of using source audio, use a .wav file (taken earlier for use i SSA i presume) that is of extremely low quality (mono, 8bit)

5) options> sync to audio this is default. make sure you dont select t odro frames when behind as this will make it impossible to scene bleed check. better to see it all slowly than not at all.

That should help considerably in minimising CPU usage.
I personally use SSA and vdub. I time off the wav in SSA (takes a couple hours; i've never done it without distractions, i think i would explode, and i always play each line twice over before i select the times so as to check for errors) then once I've done that I watch it in Vdub and right down (by hand yes) all the times that need changing due to scenen bleeds or errors. watch the areas which i knew were off and then send .ssa file off for qc. I usually try and rely on the editor to do his job properly. If, when I mature and my brain leanrs to concentrate more, perhaps i could time an average episode in.... 40 minutes being careless (thats probably testement to my perfectionist mentality)

Kanji13
2004-12-22, 12:33
Thanks Sasaki for:
"The Art of putting subtitles
On a video and audio stream"

I already using SSA to do the timing, it's a very useful tool, but don't you think although it is the best but it lack somthing, cause .wav alone not enough to complete the translation, you need raw to time other words like episode,letters, signs.
did you hear of a program called Sabbu? it's still beta, but i think it will be better than SSA, you can choose from timing using audio or video.
SSA contains KARAOKE thingy... i don't make karaoke with it to be honest, cause Arabic language read words from right to left, i usually karoke manually and it's a tough job i assure you that ;) . so how about Karaoke with SSA, does it looks OK?
i'll show you one sentence using Arabic language to make the KARAOKE:
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0400 ,0000,0000,,{\alpha&HC0&\t(0,300,1,\alpha&H00&)}{\c&H0044ff}فِـيـل
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0225 ,0000,0000,,{\alpha&HC0&\t(300,350,1,\alpha&H00&)}{\c&H0044ff}نـو
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0000,0000,,{\alpha&HC0&\t(350,650,1,\alpha&H00&)}{\c&H0044ff}شَيــم
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0240,0000,,{\alpha&HC0&\t(650,750,1,\alpha&H00&)}{\c&H0044ff}أبَاوتْ
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0410,0000,,{\alpha&HC0&\t(750,1000,1,\alpha&H00&)}{\c&H0044ff}شَيبْ
All the above just to make one sentence !! tough isn't it?
Thanks

ultrataro
2004-12-22, 16:07
I used to use SSA & wave. I've switched to SW and take script into SSA for final positioning & style editing. I find having the video in addition to wav audio to be better. I find the most of the time shifting features of SW hard to use though, but don't usually use it. I use just the grab start time & end time buttons & set the step interval to match as closely as the frame rate as can. Need a reasonably fast machine Note that I'm pretty much a one man operation, I do the translation & initial timing as one step. Go back to do signs & written text. At this point already, good enough that I generate a rough video that I hand off to friend for proofing & editing. While he's doing that I'll fine tune some timing, but usually not a whole lot is needed.
I try to avoid subtitles that bleed over scene changes. I find when I'm watching & bleed happens, there's a tendancy to think a new title has started & the eye automatically starts to read the title again. But no rule is concrete with me. I try to avoid longer than two line subs, but sometimes it's hard to avoid. Super short titles I do really try to avoid. .5 seconds is pretty close to too short. I'll start a event early or let it bleed if it'll otherwise get shorter. Too short get's distracting, when it pops up, your eye automatically goes to read it, but by that time it's gone. & a lot of times the event is pretty close to not needing translation anyway, so I'd almost rather leave it out. These kind of events it doesn't make sense to make the viewer have to go back & pause. If on the other hand it's just plain manic I'll do it, as even if you know the language you'd probably have to go back & rewatch it to catch everything anyway.
Anyway, that's how I like to do things, but the stuff i do is pretty much just for myself & friends.

Sergejack
2004-12-22, 17:02
Thanks Sasaki for:
"The Art of putting subtitles
On a video and audio stream"

I already using SSA to do the timing, it's a very useful tool, but don't you think although it is the best but it lack somthing, cause .wav alone not enough to complete the translation, you need raw to time other words like episode,letters, signs.
did you hear of a program called Sabbu? it's still beta, but i think it will be better than SSA, you can choose from timing using audio or video.
SSA contains KARAOKE thingy... i don't make karaoke with it to be honest, cause Arabic language read words from right to left, i usually karoke manually and it's a tough job i assure you that ;) . so how about Karaoke with SSA, does it looks OK?
i'll show you one sentence using Arabic language to make the KARAOKE:
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0400 ,0000,0000,,{\alpha&HC0&\t(0,300,1,\alpha&H00&)}{\c&H0044ff}فِـيـل
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0225 ,0000,0000,,{\alpha&HC0&\t(300,350,1,\alpha&H00&)}{\c&H0044ff}نـو
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0000,0000,,{\alpha&HC0&\t(350,650,1,\alpha&H00&)}{\c&H0044ff}شَيــم
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0240,0000,,{\alpha&HC0&\t(650,750,1,\alpha&H00&)}{\c&H0044ff}أبَاوتْ
Dialogue: Marked=0,0:01:28.46,0:01:30.45,Style1,Comment,0000 ,0410,0000,,{\alpha&HC0&\t(750,1000,1,\alpha&H00&)}{\c&H0044ff}شَيبْ
All the above just to make one sentence !! tough isn't it?
Thanks

Ask some programmer to write you some lines of code.

Kanji13
2004-12-23, 04:48
Ask some programmer to write you some lines of code.
Yes you are right, since the Filter can't read "RTL". I think if i do some changes inside the filter source code it may works.
So.. can you give me the source code for textsub.vdf .

vivideyecrystal
2007-02-03, 12:18
Just playing around on some old music videos because I've got nothing better to do and have run into some trouble.

"
Start VDub
Go to "Video" ---> Filters.
Click the Add button.
Select TextSub
Click the Open button and select SSA file you saved.
Exit the filtering area and make sure the TextSub filter is loaded. "

I've followed the instructions up to here: I can't load TextSub in VDub, it's nowhere to be seen. It's not even an option. How do I do this?

I know this thread is old, but if someone could help a newbie that'd be great.

martino
2007-02-03, 12:34
Download the latest VSFilter from HERE (http://sourceforge.net/project/showfiles.php?group_id=82303&package_id=84359). Extract the VSFilter.dll from the package, rename the extension to "vdf" and copy/move it into the VDub plugins directory and now it should work.

edogawaconan
2007-02-03, 23:46
no. for now, get Aegisub (http://www.aegisub.net). It makes your life much easier

GokuSoul
2007-02-08, 21:49
Okay help me here i converted the AC3 audio from my dvd to wav pcm mono so i can open it in Aegisub. Then I timed it now when I attach the AC3 audio to the video everything is like 1 second too early now but when I put it to my wav its good. Problem is I want to use AC3 audio so how can i convert the wav pcm to AC3 so it can work?

the.Merines
2007-02-08, 22:02
Okay help me here i converted the AC3 audio from my dvd to wav pcm mono so i can open it in Aegisub. Then I timed it now when I attach the AC3 audio to the video everything is like 1 second too early now but when I put it to my wav its good. Problem is I want to use AC3 audio so how can i convert the wav pcm to AC3 so it can work?

So your WAV->AC3 converter somehow added a second of nothing before the audio? That's weird. In any case, try this page.

http://www.videohelp.com/guides.php?guideid=180#180

Hopefully that will help. I'm barely a noob at this stuff so it might not be what you're looking for.

Neobody
2007-02-11, 08:33
Just playing around on some old music videos because I've got nothing better to do and have run into some trouble.

"
Start VDub
Go to "Video" ---> Filters.
Click the Add button.
Select TextSub
Click the Open button and select SSA file you saved.
Exit the filtering area and make sure the TextSub filter is loaded. "

I've followed the instructions up to here: I can't load TextSub in VDub, it's nowhere to be seen. It's not even an option. How do I do this?

I know this thread is old, but if someone could help a newbie that'd be great.

lol! I am a Noob too... You need to install VobSub... when you are installing, maybe the second page, they ask you what do you want to install.
Go click more Option "+" for Plugins... and remember to check Plugins for VitualDub. Then u are done. Just complete the installation and you will find your plugins on VitualDub ^_^.

Hope this helps

TheFluff
2007-02-11, 09:46
DO NOT USE VOBSUB
NO REALLY I MEAN IT, DO NOT INSTALL IT FOR HEAVEN'S SAKE

What is known as "VobSub" is in fact an ancient and extremely buggy version of VSFilter. It does NOT have proper unicode support and you should NOT be using it under ANY circumstances.

Instead, do what martino said a few posts back. That will do you a lot more good.

Neobody
2007-02-12, 09:13
Oh.. Thanks on that info. Anyway.. I am still trying to learn timing skills. A noob here XD. Er... I have a Noob Question... what are unicodes?

Sylf
2007-02-12, 16:50
http://en.wikipedia.org/wiki/UNICODE :
Unicode is an industry standard designed to allow text and symbols from all of the writing systems of the world to be consistently represented and manipulated by computers.

Blue Storm
2007-02-27, 04:45
This may be a stupid question but it is driving me insane. I downloaded all the files specified in the opening post and then downloaded One Piece ep 298.avi from Saiyaman raws. I opened the file in VDub as the guide said but the next step refers to an 'AVI' -> save wav menu option.

I have looked everywhere and there is no audio or avi option in my menu bar. I have file, edit, video, streams, options, tools and help and non of them contain any option I can see for getting the audio stream and saving it as a stand alone file.

Any help would be appreciated.

Mitomi
2007-02-27, 06:13
I opened the file in VDub as the guide said but the next step refers to an 'AVI' -> save wav menu option.

In VirtualDubMod it's under Streams -> Stream list -> Save WAV

In VirtualDub, File -> Save WAV.

Blue Storm
2007-02-27, 08:00
Hello all.

After watching fansubs for a few years I got it into my head that I should look into what is involved so that I might one day be able to put back into the fansubs which have given me countless hours of entertainment.

I read around this site and on Google and kinda learnt how do do timing and put simple sub titles onto a piece of video.

I have no idea what the actual translation was so I kinda just made it up myself, it is my first go after spending about 2 hours learning. If you could just look it over and tell me if I'm on the right track then I'd be grateful.

http://www.mowensworld.com/timing/OnePieceTimerTest.wmv

(In wmv format as I haven't reached encoding processes yet).


Also I would like to know if there is anywhere you can find completed translations of anime in text form and their raw so I can have a shot at timing an entire episode of something.

Quarkboy
2007-02-27, 12:32
Not too bad for a first try. There's only one really big error I caught, at 0:28 the sub disappears before she's done talking.

One a side note, you should get a job at 4kids, your "dialogue adaptation" skills are great! j/k

(Note to technical people: There's no point in criticizing fine timing at this level, so I won't berate him on that).


There are a couple of good timing guides out there you should read if you haven't already:

There's my guide http://www.fansubbers.org/index.php/KB/TheArtOfTiming (it's a little outdated 'cause it was designed for sabbu, which is not being developed anymore). But most of it is independant of the program you use.

There's this mammoth thread spanning 5 years here http://forums.animesuki.com/showthread.php?t=573&highlight=timing with an outdated (but still basic) timing guide.

And there are probably other's out there. I think people generally should learn how to use aesgisub to time with, now (if you're just learning).

martino
2007-02-27, 12:40
One note; When you have an audio file (and it's something else than WAV) in an AVI container and use VDub to demux it, just like Mitomi mentioned in the second line using the Save WAV option I have come across a problem when trying to mux it into MKV which I explained to myself as VDub automatically assuming that the file is WAVE and automatically writes an identification header to it even though in reality it is something else. So I would personally recommend you using VDubMod for this operation to avoid any issues.

Blue Storm
2007-02-28, 00:13
Where is a good place to get some experience? Apart from making up my own subtitles for raws? I would like to try time a full episode but Im not sure where to find some scripts.

Quarkboy
2007-02-28, 00:54
Where is a good place to get some experience? Apart from making up my own subtitles for raws? I would like to try time a full episode but Im not sure where to find some scripts.

Try joining a group willing to train you. You might have to look around a bit, but the smaller groups are often willing to help people learn if they are interested.

Generally finding scripts lying around for recent shows is pretty hard.

Although, if you know how, you could demux a script from a softsubbed show, strip out the times of the ssa (export it to text using aegisub), and then retime it from scratch to practice, I suppose...

endy
2007-03-01, 05:52
Where is a good place to get some experience? Apart from making up my own subtitles for raws? I would like to try time a full episode but Im not sure where to find some scripts.

you could try applying to some groups and try your hand at their timing tests, if they don't accept you, well at least you have a little more experience. If they do accept you, then good for you, you'll be getting a script to time often and you'll pick up on the subtleties quickly.

Kaonashi
2007-03-07, 09:49
What can a newbie use for timing in Mac OS X?

Sylf
2007-03-07, 11:02
I've heard of people using Sabbu in OS X.

Kaonashi
2007-03-07, 12:18
I've heard of people using Sabbu in OS X.

I'll try to compile Sabbu but there isn't a native solution?

Sylf
2007-03-07, 16:36
There's also a program called Miyu (http://www.fluffalopefactory.com/miyu/index.html). The group who's making this program, fluffalopefactory, released a few episodes of Cat's Eye using this program as a proof to show subbing in OS X can be done couple of years ago. I really have no idea what the maturity of this program is, or how similar the interface is compared to aegisub/sabbu/substation alpha.

TheFluff
2007-03-08, 05:38
Aegisub (the 2.0 prereleases in particular) kinda work(ed) on OS X, but I think compilation on non-windows systems is broken at the moment since noone has bothered to fix FFMPEG. The program itself does work under UNIX-based systems, but it has a number of glitches and quirks that aren't present on Windows. It can also be rather difficult to get it working at all... But the plan is to fix all of these problems before 2.0 is released.

Kaonashi
2007-03-08, 09:15
Aegisub (the 2.0 prereleases in particular) kinda work(ed) on OS X, but I think compilation on non-windows systems is broken at the moment since noone has bothered to fix FFMPEG. The program itself does work under UNIX-based systems, but it has a number of glitches and quirks that aren't present on Windows. It can also be rather difficult to get it working at all... But the plan is to fix all of these problems before 2.0 is released.

AFAIK, Aegisub on OS X relies on GTK which doesn't make it a native application.

TheFluff
2007-03-08, 09:52
No, it relies on wxWidgets (which may or may not rely on GTK), but AFAIK it doesn't require X or anything like that. What's your definition of a native application?

And you asked for something that a newbie can time with in OS X; I answered. In your original post you didn't state any requirements that it had to use a certain GUI toolkit.

Kaonashi
2007-03-08, 10:25
No, it relies on wxWidgets (which may or may not rely on GTK), but AFAIK it doesn't require X or anything like that. What's your definition of a native application?

And you asked for something that a newbie can time with in OS X; I answered. In your original post you didn't state any requirements that it had to use a certain GUI toolkit.

When I say a native application, I meant something that just worked: easy install, seamless integration with OS X, etc.

Right now, sabbu doesn't compile so that's why I'm looking for something "native".

Narutofan302
2008-11-15, 19:44
Where can you get a script of the translation and the cartoon that needs the timeing. i have all the requirments. do i have to ask anime suki or shinsen subs or something?

Arm
2008-11-15, 19:49
Where can you get a script of the translation and the cartoon that needs the timeing. i have all the requirments. do i have to ask anime suki or shinsen subs or something?

Indeed, you have to ask Shinsen-subs. Go to http://www.shinsen-subs.org/ and ask on their forums. :)

Kristen
2008-11-15, 20:35
Where can you get a script of the translation and the cartoon that needs the timeing. i have all the requirments. do i have to ask anime suki or shinsen subs or something?

Here's the easy method:

Install mkvmergeGUI
Install mkvextractGUI
Download a mkv sub of your choice
Extract the subtitle, video, and audio streams using mkvextract
Merge the video and audio streams in mkvmergegui

Bingo. You have a script and a raw. Now, just open the subtitles, go Timing->Shift times, and shift all times backwards 20 hours. This'll reset all times at 0:00.00 (Unless you're timing something longer than 20 hours). Time away.

timgorden11
2008-11-20, 07:45
Howdy all,

What do my fellow timers think of TPP (Timing Post Processor) in Aegisub?

Personally i don't nor shall i ever use it, because i think virtually every line needs a different lead in/out.

TPP doesn't allow for the smooth lead out that i try to use when timing eg. if someone walks past or does something big enough on screen, not to distract people from the subs but to make it look like the subs "flow" with the video (its hard to describe)

Plus mouth timing can only be done manually.

What do other timers think about TPP?

False Dawn
2008-11-20, 07:55
What do other timers think about TPP?


It has its merits, it has its flaws. Some timers swear by it, some timers (myself included) never use it. I personally think it doesn't suit my timing style and that it would ruin my timings if I were to use it.

Quarkboy
2008-11-20, 08:05
I use it only to fix 2-3 frame bleeds before doing a final qc check.

I think it's more useful for people who have experience using it and know how to time in the first place in a way that works with the post processor.

jfs
2008-11-20, 08:34
Remember you don't have to use all the functions in the TPP. You can eg. use it just to fix sloppy timings, making your own initial work easier. (Be sloppy around scene changes and let TPP fix them for you, be sloppy around adjoining adjacent lines and let TPP fix it for you, but still create your own lead-ins and outs and not let TPP touch that.)

the.Merines
2008-11-20, 10:07
Only time I use the TPP is when adjusting to a new raw or fixing someone else's rough time. Beyond that, I don't touch it.

sangofe
2008-11-20, 12:48
Howdy all,

What do my fellow timers think of TPP (Timing Post Processor) in Aegisub?

Personally i don't nor shall i ever use it, because i think virtually every line needs a different lead in/out.

TPP doesn't allow for the smooth lead out that i try to use when timing eg. if someone walks past or does something big enough on screen, not to distract people from the subs but to make it look like the subs "flow" with the video (its hard to describe)

Plus mouth timing can only be done manually.

What do other timers think about TPP?

TPP is nice, but has some flaws. Personally I always use Sabbu to time. Good 'ol Sabbu.

Tofusensei
2008-11-20, 13:57
What do other timers think about TPP?

I am not timer but I am routinely the last person going over the script before it's done. I don't know how I lived without it. It turns lousy scripts into good scripts. Assuming you watch through once and fix the random times it overshoots it, it makes life so much easier.

Fills gaps between lines, auto scene times... More what do you want? :)

I'd say on average it knocks off about 15 minutes of work per episode for me personally. This is assuming the script coming in was not scene timed nor was close attention paid to fixing short gaps between lines. It makes both mine and the timer's lives easier. Not sure why everyone doesn't use it.

-Tofu

TheFluff
2008-11-20, 14:54
I always use the TPP and keep its effects in mind when audio timing. I never do anything manually that the TPP can do automatically, if I had to do what the TPP does by hand that I would have burned out and gone fuck fansubbing a long time ago.

As Quarkboy mentioned though, if you want to rely completely on it for your scenetiming/leadin-leadout/linking needs you need to tweak its parameters carefully and take its quirks into account when audio timing.

Daiz
2008-11-21, 14:50
TPP is bliss, and I time with TPP's effects in mind. I just fix any possible little things afterwards manually if there's anything to fix.

Kristen
2008-11-21, 15:25
Howdy all,

What do my fellow timers think of TPP (Timing Post Processor) in Aegisub?

Personally i don't nor shall i ever use it, because i think virtually every line needs a different lead in/out.

TPP doesn't allow for the smooth lead out that i try to use when timing eg. if someone walks past or does something big enough on screen, not to distract people from the subs but to make it look like the subs "flow" with the video (its hard to describe)

Plus mouth timing can only be done manually.

What do other timers think about TPP?

I personally use TPP. It is a crutch of sorts right now. It is great if you want fairly consistent lead ins/lead outs, and it is something that any new timer should start using. But, in the end, it is fundamentally flawed in that it requires two passes of the script, and sometimes it will identify incorrect keyframes to snap to. For instance, as is usually the case in ef, the screen may turn black for about 0.1 seconds and then back to a picture so as to create a "flash" effect. You would not want the black to have subtitles in it, but if you use a lead out of 240, it would snap to the second keyframe instead of the first. This is what should be cause on the second pass of timing.

I normally tell timers the exact same thing. Use TPP until you know exactly what it does and what should/should not be there, and then learn manual, because manual is more accurate.

dj_tjerk
2008-11-21, 17:26
For instance, as is usually the case in ef, the screen may turn black for about 0.1 seconds and then back to a picture so as to create a "flash" effect. You would not want the black to have subtitles in it, but if you use a lead out of 240, it would snap to the second keyframe instead of the first. This is what should be cause on the second pass of timing.

Did you even read TheFluff's and Diaz's posts? Just learn to compensate for what TPP does in those (few?) cases. If it happens a lot, there's probably a pattern, so it should be easy to compensate enough to make it snap to the first keyframe. If it doesn't occur a lot, I seriously doubt scenetiming manually will actually be faster >_>

Tofusensei
2008-11-21, 19:06
Yeah, if it takes on average 10 seconds to fix those instances where it picked the wrong key frame, and it happens 10 times per episode... That is 1 minute and 40 seconds extra of your time. Sure beats doing it manually.

Instruct your editor or typesetter to fix those and you don't even need to go through the ep again, saving an extra pass.

IMO scene timing should be the job of the editor or typesetter anyway. Forcing anyone to watch the same episode twice is cruel and unusual punishment :)

-Tofu

getfresh
2008-11-21, 19:35
Did you even read TheFluff's and Diaz's posts? Just learn to compensate for what TPP does in those (few?) cases. If it happens a lot, there's probably a pattern, so it should be easy to compensate enough to make it snap to the first keyframe. If it doesn't occur a lot, I seriously doubt scenetiming manually will actually be faster >_>

TPP is a great start and takes a lot of the work out but you should always use less rather than more with TPP settings. After running TPP doing a second pass aka "fine timing" to fix the few lines that were just over the limits of your settings and would look better snapped, extended, missed, joined, split, etc... No automation is ever perfect for something that deals with A/V multimedia since what is numerically correct doesn't always "look" the best. But if used correctly they can save you a lot of tedious work. People should also be aware that you must have a good understanding of timing/scene timing to even use things like TPP and other automation tools in the first place. If you are a total beginner using these automation tools isn't going to make bad timing wonderful looking, and they can only work with the variables that you give them, i.e. they will not polish a turd or act as a substitute for people who are too lazy to putting in the time and effort required produce good work.

the.Merines
2008-11-21, 21:01
Using "snap to keyframe" and a couple hotkeys, I normally time an episode in under 35 minutes, with a record of 17 minutes on an episode of D.Gray-Man (with only 2 minor errors), so TPP is not necessarily needed for speed. [/epeen]

Kristen
2008-11-21, 21:48
Yeah, if it takes on average 10 seconds to fix those instances where it picked the wrong key frame, and it happens 10 times per episode... That is 1 minute and 40 seconds extra of your time. Sure beats doing it manually.

Instruct your editor or typesetter to fix those and you don't even need to go through the ep again, saving an extra pass.

IMO scene timing should be the job of the editor or typesetter anyway. Forcing anyone to watch the same episode twice is cruel and unusual punishment :)

-Tofu

Heh. 10 minutes. :p Whether you do manual or TPP, it's going to take approximately the same time. I've heard from some people that manual is actually slower.

In the end, unless you're trying to speed sub an episode, the extra 10 minutes or so won't hurt.

Still, manual tends to be more accurate, since you have more control over what you do. Like, "Yes" obviously does not need as much lead out as a line like "I met him in the park and danced around and was happy because he danced with me." TPP applies the same to both. Manual does not.

Koroku
2008-11-21, 21:59
I use TPP on most every script, though I limit it quite a bit.


The thing that bugs me the most is when it snaps beyond the next line to a keyframe. (Ex, the audio on line 2 starts just before the keyframe, and TPP puts line 1's end time at the keyframe, so they overlap, even though the audio doesn't)

Other than quirks like ^, it's a godsend. I remember trying to manually check each and every scene......

getfresh
2008-11-22, 00:33
I use TPP on most every script, though I limit it quite a bit.


The thing that bugs me the most is when it snaps beyond the next line to a keyframe. (Ex, the audio on line 2 starts just before the keyframe, and TPP puts line 1's end time at the keyframe, so they overlap, even though the audio doesn't)

Other than quirks like ^, it's a godsend. I remember trying to manually check each and every scene......

one annoying thing with TPP is it does not notice when it creates a collision for some reason. So you end up with .02 time overlaps from time to time causing collision detection to shift the line up in avoidance of the first line which looks nasty. It should have a setting for overlapping times that you can set it to use the start time of the second line if the threshold is lower than a certain value.

Tofusensei
2008-11-22, 01:24
Heh. 10 minutes. :p Whether you do manual or TPP, it's going to take approximately the same time. I've heard from some people that manual is actually slower.

In my experience it saves time and I've been fixing scripts other people have timed since 2001.

But to each his own :)

I run the post-processor before I ever go through the script. It certainly saves me time than not running it at all before I go through the script.

As already stated, this is all contingent on the fact that I'd be going through the script regardless and that the episode has already been loosely timed. I do not recommend running the post-processor if you do not plan to watch the entire episode afterwards. There's no reason the editor can't be the one doing this, thus negating the "but you have to make an extra pass" argument, IMO.

-Tofu

Daiz
2008-11-22, 05:40
I usually put the "fine time checking" as part of the editing/QC process as well.

getfresh
2008-11-22, 06:56
I usually put the "fine time checking" as part of the editing/QC process as well.

The only ppl I let anywhere near timing QC are timers, stylers and typesetters for the most part. Editors should focus 100% of their attention on the dialog and not be distracted with other tasks. Most QCers these days are pretty much just a 3rd pass edit imo. To be able to QC for a certain task you gotta know wth the fuck ups look like, you need to know what guidelines/standards/style the person who did the work is using, and just like with TC you should be as good as they are if not better. You can't have a 3rd grader checking a 6th graders homework, it doesn't work like that. Sadly it is becoming more and more standard to hire ppl to QC who have next to or no experience at all with fansubbing, which I attribute to some of the decline in quality we have seen over the last few years in relationship to the actual content/display of the subtitles. Oh well, hopefully more experienced fansubbers will get sick of doing the hard part and switch to QCing since it is less demanding of time. And hopefully they will come and apply at my group so I can lazy around more~

Daiz
2008-11-22, 08:33
Well gee, of course the people that will actually do things need to possess abilities required for doing those things properly. Also, nobody stops anyone from being multitalented.

getfresh
2008-11-22, 18:32
Well gee, of course the people that will actually do things need to possess abilities required for doing those things properly. Also, nobody stops anyone from being multitalented.

ya-huh~ they still do these things is separate passes if they are experienced at least in my experience with working in groups. And in case you have never heard this before, "A jack of all trades is a master of none." Think about it.

dj_tjerk
2008-11-22, 19:16
And in case you have never heard this before, "A jack of all trades is a master of none." Think about it.
Not really what he meant with multi-talented.. And being able to do more doesnt necessarily mean you're bad at all those things.. (Hell, I guess most multitalented people are actually better at the things they do than the people being able to do just one of those things.)

But leaving that aside.. TPP is just something that makes timing easier, and if you take 5 minutes to study it's behaviour, you won't have any 'snapping to the wrong keyframe' problems. I usually time with it's behaviour in mind, and then apply it. Then I usually watch the episode in aegisub whilst styling/fixing overlapping lines. (I want a hotkey for changing the style of a line plz. I hate moving my mouse :P).

getfresh
2008-11-22, 21:37
Not really what he meant with multi-talented.. And being able to do more doesnt necessarily mean you're bad at all those things.. (Hell, I guess most multitalented people are actually better at the things they do than the people being able to do just one of those things.)

But leaving that aside.. TPP is just something that makes timing easier, and if you take 5 minutes to study it's behaviour, you won't have any 'snapping to the wrong keyframe' problems. I usually time with it's behaviour in mind, and then apply it. Then I usually watch the episode in aegisub whilst styling/fixing overlapping lines. (I want a hotkey for changing the style of a line plz. I hate moving my mouse :P).

No what I was getting at is the fact that doing many different checks at once on your pass where you are doing work from scratch is a bad idea. Someone can have many talents and do many passes. But to split your attention on subs imo if asking for increased error accumulation. If an editor is a competent timer as well they should take care of checking during the QC not while they are editing. Yes some people are great multi-taskers, blah blah, I've heard the whole argument about it, and I still see the fuck ups that happen due to dividing your attention. Of course anyone is going to notice a whole sub is missing, if it starts a full second late, or it there is a 3-liner, but they shouldn't be sitting there looking for bleeds, missing periods, wrong styles, capitalization errors, etc... unless it directly pertains to their current task to be doing it. A jack of all trades never masters one because they are not directing their full attention in any one direction.

for the hotkey thing, you can get a standalone macro-program that you can record macros into for pretty much any program in windows. If you don't have windows, I don't know what they offer but I'm sure there is something, may be worth a look. Maybe set it up to use your number pad as triggers for each diff style.

pichu
2008-11-23, 17:41
not really. you can know all of the skills but excel at one or two skills. The word "master" is way too subjective anyways. Usually, when you get really good with one skill, you'll realize that you have much to learn, so you'll trap yourself into thinking that you still have a long way to go before becoming a master -- even though your skill level is way above millions of people who are in that very same field. The only people who can say and claim that they're masters are inexperienced and ignorant people. It's like a first-year student giving you a lecture on your expertise.

Why are you people so serious? Fansubbing is meant to be funsubbing; let's have some fun!

Kristen
2008-11-24, 08:46
one annoying thing with TPP is it does not notice when it creates a collision for some reason. So you end up with .02 time overlaps from time to time causing collision detection to shift the line up in avoidance of the first line which looks nasty. It should have a setting for overlapping times that you can set it to use the start time of the second line if the threshold is lower than a certain value.

What annoys me is not the fact that it makes these mistakes, but that they are really hard to fix when doing your second pass. Usually, when it makes those overlaps, it means that there was a keyframe that was within the ends after threshold, but outside of the starts before threshold for the two lines. And since line linking is applied first, it will line up the lines and then snap.

The issue is that you can either move the start time, which usually will make the text of the line start after the character starts speaking, or you can move the end time, which make a particularly nasty inverse bleed.

False Dawn
2008-11-25, 09:07
The issue is that you can either move the start time, which usually will make the text of the line start after the character starts speaking, or you can move the end time, which make a particularly nasty inverse bleed.



In my experience as a timer, that's actually generally down to a dodgy raw encode or workraw (audio/video desync mostly) because the idea of having a scene change fractionally after someone starts speaking is visually disarming to the viewer, so I can't think why any series would have it. It's distracting. I've timed several episodes like it, all of which I've brought to the attention of the encoder, and they've usually been able to shift the audio to make it more in time with the video.

Of course, this is just with episodes I've timed - there are probably some series where there are odd scene changes in relation to audio changes, but they're few and far between. They'd be like one per episode, if that - if they're substantially more, I'd suggest it's a raw/workraw encode issue.

Kristen
2008-11-25, 11:28
In my experience as a timer, that's actually generally down to a dodgy raw encode or workraw (audio/video desync mostly) because the idea of having a scene change fractionally after someone starts speaking is visually disarming to the viewer, so I can't think why any series would have it. It's distracting. I've timed several episodes like it, all of which I've brought to the attention of the encoder, and they've usually been able to shift the audio to make it more in time with the video.

Of course, this is just with episodes I've timed - there are probably some series where there are odd scene changes in relation to audio changes, but they're few and far between. They'd be like one per episode, if that - if they're substantially more, I'd suggest it's a raw/workraw encode issue.

It's usually 1 or 2 per episode, nothing more. When there's multiple weird scene changes, it normally means that the audio isn't shifted correctly. But there'll be 1 case every 2-3 episodes where someone starts talking right before the scene change, and that's what drive me insane.

getfresh
2008-12-01, 19:08
It's usually 1 or 2 per episode, nothing more. When there's multiple weird scene changes, it normally means that the audio isn't shifted correctly. But there'll be 1 case every 2-3 episodes where someone starts talking right before the scene change, and that's what drive me insane.

try subbing older anime, and you will run into it all the time.

guest0815
2008-12-03, 14:36
I'm just trying to get a little into timing and I've got some questions regarding this post about using the tpp:

Yes, and one solution is to set the "ends after" threshold to at least greater than (leadout+link_threshold*0.5)/(1000/framerate) assuming you have the bias slider in the exact center; I prefer to have the bias slider around 3/4ths to the right though so I'd use 0.75 instead of 0.5. Another solution is to account for what the TPP will do when rough timing and simply place the line end marker a bit before the keyframe and let the TPP's lineout plus keyframe snapping take care of extending it to the keyframe.

I know, it's because I'm lazy.
I dunno what exactly amz had in mind when designing it but the larger the thresholds are the harder it is to avoid mistakes.

@ChrissieXD: your thresholds are way too high.

This is a bit off-topic, but some TPP tips and related maths:
- Always set "ends after" to greater than leadout/(1000/framerate). Since keyframe snapping happens last, this allows it to undo the leadout and avoid a bleed caused by you timing the line end right at the keyframe.
- By the same logic, "starts before" should be greater than leadin/(1000/framerate).
- As mentioned above, "ends after" should also be greater than (leadout+link_threshold*0.5)/(1000/framerate).
- Remember that the maximum possible autoadded total leadout in milliseconds is leadout+ends_before*(1000/framerate), since keyframe snapping is applied after leadout.
- Likewise, the maximum possible autoadded total leadin is leadin+starts_after*(1000/framerate).
- To some extent it is possible to compensate for imperfectly calibrated settings by considering the effects of the TPP when rough timing; see above. In fact you should always consider the effects of the TPP when rough timing.

My own settings for 23.976fps are leadin 120, leadout 240, linking 320, starts before 3, starts after 3, ends before 5, ends after 8. This doesn't satisfy the line linking requirement outlined above but I try to compensate for it by timing line ends near keyframes a bit earlier.

If you used a line joining bias slider that's not completely to the right then the same calculations would apply to the starts before snapping that were used on the ends after snapping right?

For example:

Line start is snapped to keyframe during rough timing. Tpp lead In extends it 120ms beyond that. This would normally be compensated for by a starts before snapping of 3 frames (~125ms) but the start is now inside the line joining threshold of the preceding line's lead out so you'd have to add some frames to the starts before snapping if you wanted to compensate for that just like it was described in the case of lead outs getting extending by line joining in the other direction. Is that so?

So, how many people are actually using a bias that's not totally to the right since that would get rid of this source of error and do they compensate via larger keysnapping? I'm asking because the Fluff uses a 0.75 bias but no keysnapping compensation at the start beyond the 3 frames for the 120ms lead in whereas he used 2 additional frames at the end beyond the 6 frames that are needed to compensate for the lead out.

Also what is the importance of ends before and starts after?

thanks :)

TheFluff
2008-12-03, 14:47
Line start is snapped to keyframe during rough timing. Tpp lead In extends it 120ms beyond that. This would normally be compensated for by a starts before snapping of 3 frames (~125ms) but the start is now inside the line joining threshold of the preceding line's lead out so you'd have to add some frames to the starts before snapping if you wanted to compensate for that just like it was described in the case of lead outs getting extending by line joining in the other direction. Is that so?

Yes, that is correct. I set my TPP like that because I don't snap to keyframes while rough timing, I place the start time a little bit after the keyframe intentionally because of the above TPP behavior.

Bananawaffles
2009-03-03, 15:28
I'm interested in helping to sub anime and I'm curious if having to know Japanese is a must for timing (J->E subs). I want to try this out, but I don't want to waste my time to find out in the end that I can't do it >_<

Tofusensei
2009-03-03, 17:01
Bananawaffles, it is not required.

tun
2009-03-03, 17:11
I'm interested in helping to sub anime and I'm curious if having to know Japanese is a must for timing (J->E subs). I want to try this out, but I don't want to waste my time to find out in the end that I can't do it >_<

It's not necessary, but it's helpful in making sure that you are timing the correct line to the spoken line. And in the case of long sentences that span over 2 lines, you'd know where to split them.

False Dawn
2009-03-03, 17:48
It's not necessary, but it's helpful in making sure that you are timing the correct line to the spoken line. And in the case of long sentences that span over 2 lines, you'd know where to split them.


Though most timers get by on knowing less than a hundred Japanese words in total. That's all you really need, to be honest.

sangofe
2009-03-07, 14:17
Though most timers get by on knowing less than a hundred Japanese words in total. That's all you really need, to be honest.

Not with "good" old Joe, no.

aeonsky
2009-03-16, 12:23
Just want to point out that Substation Alpha does not work on a 64-bit OS. Thus, I had to find something else to time with.

TheFluff
2009-03-18, 09:19
I'm pretty sure not even the hardcore oldskool timers use SSA anymore, note that the OP of this thread is a bit over five years old

False Dawn
2009-03-18, 16:27
I'm pretty sure not even the hardcore oldskool timers use SSA anymore.


You'd be amazed what some timers still use. I know a number of timers who still haven't made the leap to Aegisub :D