Aegisub
February 09, 2010, 12:56:23 AM *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News:
 
   Home   Help Search Login Register  
Pages: [1]
  Print  
Author Topic: Timing Tutorial  (Read 5538 times)
Baka-dono
Comic Sans MS
*
Posts: 2


View Profile Email
« on: March 09, 2008, 11:33:16 PM »

----1. Introduction
----2. FAQ
----3. Timing Tutorial
--------3.1. Preparations
--------3.2. Shortcuts
--------3.3. Rough Timing
--------3.4. Post processing (Fine Timing)
--------3.5. Check it again
--------3.6. Shifting
--------3.7. Karaoke Timing
---------------3.7.1. Preparations
---------------3.7.2. Splitting a Line
---------------3.7.3. Timing a Syllable
---------------3.7.4. Kanji Timer
----4. End / Contact

1. Introduction

Okay, before we start I would like to explain this tutorial. You can find about 25 different timing tutorials using Google. I've read some of them and learned many things. But most of them are pretty old and don't use the new developed features of programs like Aegisub. I started two years ago as a timer and became pretty much experienced with time, so I thought I can write a tutorial myself. This tutorial should be seen as a small guide on your way of becoming a successful timer. The solution presented here is most likely not the easiest, fastest or best. It’s just one way of doing things. Please note that this is my first try writing a tutorial.  If you have a suggestion, you can contact me anytime. (See End / Contact)
I'm not a native english speaker either, so there are lots of grammar mistakes and misspellings. Would be great if someone could edit this tutorial Wink.

2. FAQ

Q: What is fansubbing?
A: Fansubbing means creating subtitles from a show or a movie that is not available in your language. Fansubs are made by fans for fans and are free. Fansubs are normally created by freaks that have too much time on their hands. They join forces in groups to be able to impose their favorite show on innocent people. Inside the group, a fansub is divided into different jobs, in most cases Raw providing, Translation, Timing, Editing, Typesetting, Karaoke-Fx, Encoding and Quality Check. This tutorial covers Timing and Karaoke Timing (needed to do flashy karaoke).

Q: What is Timing?
A: Timing is basically nothing more than determine the time a subtitle appears on the screen. Timing is divided into rough timing and fine timing. Rough timing captures the time a person talks, fine timing means adjusting the timing to the video to prevent scene bleeds. More on that later.

Q: What is scene timing?
A: See Post-Processing

Q: What are Lead-ins and Lead-outs?
A: Lead-ins/outs refer to an extra amount of time added to the normal time a subtitle stays on screen to give the viewer the chance to read a subtitle without pressing the pause button. More on that in the chapter Post-Processing.

Q: What programs do I need for timing?
A: There are many programs out there capable of creating subtitles. This tutorial is only covers Aegisub, since this is the only free program that is developed constantly. You can download Aegisub here.

Sabbu is a good alternative, loved by many timers, but its development is no longer continued. I can only advise you to stay away from programs like Substation Alpha, Subtitle Workshop and Medusa. Those programs have none of the features Aegisub has and are out-dated.

Q: How long do I need for one episode?
A: I would say you need 1-2 hours at the beginning. Later on you can handle an episode within 30-90 minutes. If you care about quality, you watch your timing again after you’re done to spot errors, which takes 20 more minutes. I like to take breaks between timing, since it gets really boring.

Q: Do I have to speak Japanese to become a timer?
A: Absolutely not! Even without knowledge of the Japanese Language you can tell when a person starts to talk and when he/she/it stops. However, Japanese is a really interesting language and it is helpful, be it timing or generally watching fansubs. Over time you are bound to learn Japanese Words anyway.

Q: Your tutorial sucks, where else can I learn timing? / I didn’t understand a part; can I look it up somewhere else? / How do I install Aegisub?
A: Aegisub Help <-- you should be able to find any answer there.

Q: Timing sounds boring. Is it worth to be a timer?
A: Timing is very boring and no one will thank you for sacrificing your free time. Many people praise great karaoke effects, a nice encode or beautiful typesetting. As a timer, you have to renounce getting credits. But timing is a necessary part for your favorite anime being released as a fansub. I can’t tell you if it’s worth the trouble, you have to decide that for yourself.

3. Timing Tutorial

3.1. Preparations

I assume you have installed Aegisub and are ready to get started. You need two things: a raw (workraw preferred) and a script with the translation. Hopefully it’s saved as .txt/.ssa/.ass. A perfect script is a .txt file that includes the actor, followed by a “:” and the translation.

For example:

Code:
Ren:Excuse me.
Ren:Shinra-sama, it's morning.
Shinra:Ren?
Ren:Yes.
Shinra:I had too much to drink and my head hurts.

You start with opening both files. Let’s open the script first. You can click on the Open File symbol or choose File -> Open Subtitles….
A Text Import options menu appears. In the Actor Separator box you type in the symbol you used to separate actor from translation. In the example above, we need to choose “:”. Comment Starter is the same, you enter a symbol you used in front of a comment line. Comment lines are visible within the program, but are not visible in the video.

Next you have load the video. Choose Video -> Open Video…



Simply choose your raw and click on open. Should your raw be in a different format than .avi/.avs (like .mkv), you have to choose “Other supported formats…”. To begin with timing, you still have to load the audio. Choose Audio -> Open Audio from Video. This can take a while, but soon your screen should look like this:



The green bar is the so called waveform, a way to make audio visible and the most important help you can get. Over time you can see on this form when someone starts talking and when the talking ends.
To the right you can see three control units. The first two controller scale the waveform horizontal and vertical. The last one adjusts the sound volume. Play with those settings a bit.
You could start timing now, but the make everything a bit more arranged, choose View -> Audio+Subs View. You don’t need to see the video while timing, so it only gets in the way.

3.2. Shortcuts

Great, finally we can get to the point. Or do we? Before that I want to explain some shortcuts. You can click around with your mouse and finish timing a script, but it will only end in chaos and requires the double expenditure of time. You should try to use shortcuts (the keyboard) from the beginning. Now I’ll list the most important shortcuts you use while timing. (Please note: I use SSA Shortcuts. If you want to use the new ones or make your own shortcuts, choose View -> Options… and the category Hotkeys)

Code:
”S” : Plays the line from the start marker to the end marker.
“T” : Plays the line from the start marker to the end of the video.
“H” : Stops playing the audio.
“D” : Plays 500ms before the end marker.
“W” : Plays 500ms after the end marker.
“Q” : Plays 500ms before the start marker.
“E” : Plays 500ms after the start marker.
“G “ : Saves your selection and jumps to the next line. (You’ll press this a lot!)
“A” : Moves the waveform to the left. (If you have a mouse with a wheel, simply scroll to move the waveform, I prefer this method)
"F" : Moves the waveform to the right. (If you have a mouse with a wheel, simply scroll to move the waveform, I prefer this method)

Those are the shortcuts you should remember. It can’t hurt to know more than that, but remembering the ones mentioned above is sufficient.

3.3. Rough Timing

Finally we can start!
One can even explain it with one sentence:
Press T, listen when a person starts to talk, Left click on the waveform where the person starts to talk and right click where the person stops. You’re done. Simple, isn’t it?

A bit more detailed:

Start playing the audio by hitting “T”. You will see a white marker running. This marker indicates the position you’re hearing right now. Continue until you hear the first spoken sentence. Left click on the position the white marker was when the person started speaking. Right click on the position you assume the person ends speaking. You’ll notice a red marker appearing when left clicking and a yellow marker appearing when right clicking. If you set both markers, the area between those two is dyed red. Furthermore you can see the Start and End Time boxes are now blue. This indicates you have changed the time specification for this line.

But still, nothing is final. Most of the times you won’t be done with two simple clicks. It’s your goal to catch the phrase. (Gotta catch them all, right?)
The red marker should be positioned at the beginning of the sentence, the yellow marker at the end. Shortcuts will help you, so hopefully you remembered them.

To see whether or not you’re at the beginning, press “Q”. Do you hear the actor talking? If true, left click a bit further left on the waveform. If you’re not hearing anything, press “E” to hear whether or not the actor starts talking directly. If not, you positioned the marker too much to the left.  Adjust the marker and you’re done with the start time. Now you have to do the same with the end time. By pressing “D” you should hear the end of the sentence; with “W” you should hear nothing. To continue press “G” and Aegisub saves this line and jumps automatically to the next line.



After you pressed “G”, you will see your markers changed their positions and your old selection is darkened.  Aegisub uses the end time of your previous line as the start time of a new line and adds two seconds to the end marker.
You have to repeat this process about 300 times and you’re timing script is almost done.

By the way, if several person talk at the same time, you don’t have to worry about it. Simply set your markers, even if the markers overlap with the previous line. It’s the job of a typesetter to straighten it out. We already have enough to do.

When you’re done with the rough timing, SAVE! This is very important. Save this file as Some_show[rough_timing].ass, we might use it later.


3.4. Post-Processing (Fine Timing) or Help, what are those pink lines doing in my beautiful green Waveform o_O!

Leads:

Why are movies and shows more successful in today’s life than books? Because you don’t have to read when watching a movie! Reading requires more time than listening and you have to consider this fact while timing. An actor talks sometimes faster than a viewer can read the subtitle and we as timers are responsible that the average viewer can read everything without pressing pause. How do you do that? You change the start and the end time of every subtitle a little bit.

If you’re thinking: “Damn it, couldn’t you tell me earlier? I already timed the whole script!” rest assured, everything can be added automatically. Aegisub has a wonderful feature called Post-Processing which covers this boring job. More on that later. To give the viewer more time, you add lead-ins and lead-outs. The difference between in and out is simply where you add them. Lead-ins before a sentence starts and lead-outs after a sentence is spoken. Most of the times lead-ins last less milliseconds than lead-outs (about half of the time).

Example:


Scenetiming:

You probably noticed those pink lines on the waveform while timing. If you wonder what they do, they indicate keyframes. Aegisub has a great feature that shows all keyframes in the waveform when you load a video. So what are keyframes?
Just imagine keyframes as the first new picture of a new scene. A good timer knows that it helps the human brain to easily read subtitles when they don’t appear after a scene changes. This is called “bleeding”. After a scene bleed, the brain wonders “What happened right now?” and lost the plot. To prevent scene bleeds a timer has to scene time. Post-Processing helps you with scene timing as well.

Scene bleed (animated .gif)

No scene bleed (animated .gif)


Linking:

Linking of subtitles can be seen as good style. You can argue about whether it makes sense or not but I prefer linked subtitles. Sometimes end time of one line are very close to the start time of the following line. Linking adjusts the end time of the first line and equates it with the start time of the following line. Sounds confusing but the following pictures should make things clear:

No linking (animated .gif)

Linking (animated .gif)


Post-Processing covers this part as well.


Post-Processing:

So how can you use this god-like tool and the fulfillment of every timers dream? Choose Timing -> Timing Post-Processor… and you open up this legendary menu.
Several options can be set. All of them are self-explanatory so I won’t go into details.

In the Lead in and Lead out box you enter yours leads in milliseconds you’d like to add to the subtitle. Lead-ins range from 40-250 ms, lead-outs from 200-500. Personally I like using 200 ms lead-in and 400 ms lead-out, but that’s preference.

The next option sets the amount of linking. The Threshold box indicates how many milliseconds subtitles have to be near each other before they are linked by the Post-Processor. Standard is 300 ms, I like to use 350 ms. The Bias slider doesn’t work the way I want it to, so I’ll explain a trick later. Basically, if you put it to the right, the end time of the first line is set to be the same as the start time of the first line. When the slider is in the middle, the subtitles will “meet” in the middle as well. Just push it to the right for the sake of this tutorial Wink

The last option contains the settings for Keyframe snapping (only supported when you have a video). Keyframe snapping shortens or lengthens the start or end time when a keyframe (that pink line) is near, so the line ends or starts exactly when the scene changes. If you move your cursor over the several boxes, you get a short explanation of what they’re for. Be aware that those settings are not measured in milliseconds but in frames. (A frame lasts about 40-50 milliseconds)
How you set the number of frames is your choice. Play a bit with the settings, you’ll find what you’re looking for.

Remember the trick I mentioned earlier? It’s really simple, you have to use the Post-Processor two times. If you hit OK right now, Aegisub will sometimes add no lead-ins to lines while linking. I don’t want that, I want all lines to have a small lead-in. To do that I use the Post-Processor first with only the lead-in box checked. In this box I set half the amount of the lead-ins I want to use. Hit OK and open the Post-Processor once more. Check on all other options and hit OK again.



Your timing script is done!!!

Congratulations! You’ve done it! Good job! (>^_^)> (>^_^<) <(^_^<)

Now you should save your script. Remember to save it as a new file, for example Some_show[timing].ass.
It was a long way but a new timer is born.


3.5. Check it again

Your script is done but it might be wise to check your work once more. Every human makes mistakes and machines make even more. Though the Post-Processor took care of most things, it’s not perfect. For example there are two keyframes close to each other and the program is confused which one to choose, resulting in a scene bleed. Choose View -> Full View and watch the episode again. (Or if you’re like me, mux the script with the raw and get a .mkv to watch)

Should you spot an error select the line and correct it. To correct a scene bleed you have to position the red or yellow marker exactly on the pink line. I hope your scripts won’t have any errors. If there are some, you can fix them with no problem. Congratulations, you have finally a perfectly timed script.


3.6. Shifting

Oh those damned encoders, theyalways find a new raw! In your career as timer someone will definitely ask you to shift your timing to a new raw. Aegisub helps you with that as well. Load the new video into Aegisub and open your script. You will notice that your timing is not in the right place. It’s your job now to correct this. Select a line and move the red marker to the position it should be. The Start time box dyes blue, indicating the new time, but the original start time is still shown in the start time row of the line.
Now you have to calculate the difference in time. Write down this difference or try to remember it. Without editing the line, go to your timing menu and choose Shift Times…. In the Time box, enter your just calculated time. Beneath this you can choose between Backward and Forward. Move your cursor over one of them to see what they stand for. This and the other options should be obvious, so I won’t go into detail (this is becoming a trend, sorry).

Hit OK and your script is now adjusted to the new video. However! There is a possibility that scene bleeds occur. To correct this we will rely on the Post-Processor once more. Activate Keyframe snapping only and use as settings 1/1/1/1. Aegisub will now straighten out all unwanted scene bleeds.
Logged
Baka-dono
Comic Sans MS
*
Posts: 2


View Profile Email
« Reply #1 on: March 09, 2008, 11:33:41 PM »

3.7. Karaoke Timing

3.7.1 Preparations

Karaoke timing starts like normal timing. Load your script and video and time all lines. You probably want to time 3 times at the same time, kanji, romaji and the translation. Just select all three lines before you hit “G”. Fine timing is not necessary, rough timing is enough.

3.7.2. Splitting a Line

To achieve a karaoke effect, timing one line isn’t enough. You have to time every single syllable in this line. To make things easier Aegisub comes with a karaoke mode. Choose your first romaji line (we don’t care about kanji and translation lines for now) and press the button called “Karaoke” under the waveform. Right next to this button you’ll see a now active “Split” button. Press this button and move your cursor over those white boxes. Everywhere you left click a blue marker appears. Between two markers should be one syllable. If you have a text like:

Code:
kono gaido wa sugoi

You want to split it into:

Code:
ko|no |ga|i|do |wa |su|go|i

If you made a mistake, you can click again to remove the marker. After you’re done, hit enter or “G”. You’ll notice a symbol in front of every syllable when you look at your line. If you look at the text, each symbol stands for {\k...}. The number after the \k indicates the duration of this syllable. And you have to adjust this duration.

3.7.3. Timing a Syllable

Timing a syllable is the same as timing a normal line. Though you won’t left/right click but drag the yellow lines to the left or right. All shortcuts still work, so use the shortcuts to hear when a syllable starts and when it stops. Drag the dashed line of the first syllable by holding down you left mouse key. Move its position into the right place and skip to the next syllable by clicking on the syllable in the waveform or clicking on it in the box next to the Split button. After some back and forth everything should be in place and you can hit enter or “G” to accept the changes.

If there is a pause within the line you have to add another {\k…} tag. Click the Split button again and set a marker where the pause starts. Hit enter and you have added a pause that you can drag like any other syllable. Easy, isn’t it?



3.7.4. Kanji Timer

A fairly new Aegisub function is called the Kanji Timer. It’s a feature that helps you transfer your romaji timing to kanji lines. Before you can use it, you have to create new styles though.

Styles are for typesetters and could fill another 10 pages of tutorial.  I’ll just explain how to create new styles. Choose Subtitles -> Styles Manager… and you have a new window pop up. In this window you’ll see most likely a style called “Default”. Click the button called “Copy” below this style on the right side of the window. Another window pops up and you can rename the style to whatever you want (for example “Kanji”). Repeat this once more and create a new style for your romaji lines. Close the Style Manager after you’re done.

Now select all romaji line (hold down Ctrl and click the lines Tongue).  Under the “Karaoke”, next to the Comment Checkbox you find a dropdown menu. Choose your romaji style. Repeat this with your kanji lines.
Open the Kanji Timer (Timing -> Kanji Timer…). As Source select your romaji style, as Dest your kanji style. Hit Start! and you can select your syllables and the matching kanji with the arrow keys. After you’re done close the window. Aegisub has once again done the dirty work for you.



4. End / Contact

Well, that’s it. I’m sure there are many open questions (like alpha timing which isn’t covered).
If you have questions feel free to write me a PM or send me an e-mail (bakadono.2@googlemail.com).
If you have suggestions (be it misspelling, flaws or inconsistency) notify me as well. I’m open to criticism Wink
Thanks to the Aegisub developers who made this great program!
I also want to thank Devil Doll for his suggestions and corrections!

Thanks for reading; I hope I was able to help you a bit.
Logged
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!