Preface:
This document assumes the following:
Section 1: Obtaining a translation
When creating a timed script for dvd playback on the pc, you will have two main considerations, the source of your translation and its timing. Sources can be many, whether it be a vhs tape, a vcd, a newly done translation by yourself or someone else, or a dvd that already features subtitles. In the case where your source is a vhs tape, vcd, or a new translation you can skip ahead to section 4 of this document that deals with the usage of substation alpha to manually time and transcribe a script.
If your source is a dvd that features subtitles, we need to determine some further details. If the dvd has subtitles which are not optically burned into the source print, we can extract the timing information along with the text of the subtitles using a program called subrip (visit doom9.org 's download section to obtain a copy, as it would appear the main page is down). DVD subtitles come in several flavors. They can be removable, forced, or in closed caption format. All are extractable with subrip, but with some caveats:
1) Closed caption subtitles come in several colors, which causes several headaches for subrip. Subrip works via OCR, like the software that is often bundled with scanners. It works on principal by recognizing characters of a certain color, against a certain background color.
Subrip can only work with one set of "colors" at a time, meaning you would have to rip the subtitles several times, each time defining a new color for subrip to work with in order to extract closed caption subtitles. This creates a massive headache as subrip for some reason messes up the timing offsets for the different sets of subtitles, meaning you have to shift the timings around. These types of subtitles can be extracted, but beware they are a bit of a headache. US and French studios seem to favor using these sort of subtitles.
2) The other caveat to be considered is the presence of CSS encryption. Subrip generally only works with dvds that are not CSS encrypted. If your dvd is CSS encrypted, you will have to download a program like DeCSS (again visit doom9.org for such programs) or the like and remove the CSS encryption while copying all the relevant files to your hard drive. This takes a bit of time but compared to manually timing the subtitles it is nothing much to worry about.
Section 2: Using Subrip
Subrip is a fairly intuitive program to utilize. Most of the default settings should prove sufficient for most occasions. The first menu you'll want to look at is located under File -> Open Vobs:
In this section, you start by choosing Open IFO. This may present you with a list of choices, just choose any one to start with. Upon opening the ifo file, the Vob Files: section should now list a bunch of .VOB files. The main movie vobs are what we are concerned with. Generally the main movie will be the set of files that contain a bunch of files sized at 1048404 KB. They can be grouped by the VTS_0X_X numbers. The movie will start with a 1048404 KB file and end with a smaller file. In the example above you can see that the main movie files are in VTS_03, sections 1-5. Now that you know which VTS_0X corresponds to the movie files, you can select the correct IFO file. The matching IFO file will bear the same VTS_0X number, so in this case our IFO file would be VTS_03_0.IFO. Use the Open IFO button to select the correct file. You should now see several options under the Language Stream section. Choose the English stream, or which ever stream you want to use.
Handy tip: If the DVD doesn't feature English subtitles, you can still rip the other set of subtitles to generate the timing for the script. Choose Save sub pictures as bmp under the Action section. When Subrip is done with the bmp extraction, go into the results section, change the output format to subrip, and there you have your timing.
When you have your movie files and language stream selected hit start. The first after running for a while, subrip will prompt you to select color to see white text on background. Scroll through the options until you have achieved this. Subrip should shortly then bring up the following menu:
You will be required to define each character that it highlights, so it knows which character in the subtitles corresponds to which subtitle. If you find that it tends to recognize large groups of letters, you can play with the settings in Options -> OCR advanced setup to increase or decrease it's sensitivity. After a bit of tedious work to define the characters, subrip should finish with the subtitle extraction.
The results are written to a window that we see in the above picture marked Subtitles (bottom left corner). Bring this menu into view:
What you now need to do is some post OCR spelling correction if your quotes have been messed up, or if I and l have been confused. Under the corrections menu there are options to automate this. When you are done go to the File menu and save the output as a .SRT file. I recommend running SRT2SSA.EXE to convert the .SRT to a substation alpha file format. By doing so you can easily load this file into substation alpha for further editing or spell checking. Now that you have a timed script, you can skip ahead to the dvdsubber section, and put it to use.
Section 3: Creating a WAV file for timing in Substation Alpha
If you have to create a new timed script, substation alpha (www.eswat.demon.co.uk) is the program to use. Substation Alpha supports several methods of timing, but I am only going to recommend using wav timing. In order to wav time, you need a wav file to time from. Substation alpha unfortunately limits your wav format to a 8-bit mono file, with any sampling rate you choose. Now this normally isn't a bad option, but when you record such a wav file with most low end soundcard's inputs, you get a wav file that lacks clarity. Low passages can be extremely difficult to time this way. Feel free to use whatever program you want to record your wav file, but I recommend doing this all digitally, and using a clever little program called total recorder (www.highcriteria.com). This is a clever little program which allows you to redirect any sound from your PC to total recorder. Total recorder's drivers can then save the resulting sound to a wav file with the settings you choose.
The main program view of total recorder looks like this:
The first thing you want to do is go into the options menu and have a look at the settings. Normally you should just leave the default options on and just choose where you want your temporary and finished files to be saved. When this is done, pop into recording and source parameters. Here are the settings I use. Of note should be the recording parameters. Make sure you alter these to 8 bit mono or else you wav file will not be accepted by substation alpha!
To do the recording, you can manually start and stop the process by starting your dvd software player, starting the movie, and hitting record. However this is really cumbersome and I recommend using the scheduling option to automate things a bit, so that you don't have to worry about stopping the movie in time (Options -> scheduling):
Click on new, which should bring up this menu:
Simply fill in the required information for starting and stopping the job. Look at the run time of the movie, and create a job which matches its run time with a few extra minutes to be safe. Choose the start time for a minute or two ahead of when you want to start recording. When finished hit ok, which should bring you back to the scheduling menu. Make sure the task is enabled and go back to the main menu.
One last step to check before recording. Go into the Windows control panel and select multimedia, check and make sure that your window looks like this, with playback and record through total recorder selected:
When you wish to record, first start up your software dvd player. Hit play and then pause at the very start of the movie. When your scheduled start time is about to begin or has just begun, hit play again. This should start total recorder's recording of the dvd movie's sound. This is a bit tricky to setup at first, but just play with it till you have the hang of things. At the end you should have a clean as possible wav file. Now I should note there are many vob to wav type programs, which purport to convert the file digitally to a wav file. I would advise caution before using these as I have had massive synch problems when using them in the past. I highly recommend sticking with total recorder as it produces good output and doesn't require you to mess about with vob files. One last note here: if you find your wav file is noisy or distorted, go back and check your sound card's mixer settings. You may have to mute some of the settings to get clean audio. This varies by sound card, so if the file sounds fine, don't worry.
Section 4: Timing with Substation Alpha
Substation Alpha can be a bit intimidating, so lets start with the main window:
Now what we are concerned with here is wav timing, so don't worry about the other menus and settings. Simply choose Timing and then Time from WAV file. You should now see this:
Click on open and select the wav file you created with total recorder. You should now see something like this:
You should see the wav form of the wave file you created. This linear wav form represents the audio of the movie, including the speech, sounds, etc, from start to finish, left to right. What we want to do here is isolate a line of dialogue to subtitle. This is done with the yellow and red markers you see above. Left clicking the mouse sets the "in" yellow marker and right clicking selects the "out" red marker. The box above the wav form is where you type in the corresponding text. It is important here to get your markers set fairly accurately so that you don't start or stop the dialogue too early or too late. I recommend clicking the zoom button all the way to the right to make the visible wav form portion as zoomed in as possible. The Y-scale setting is handy as most sections of actual spoken dialogue have peaks that differ from the background ambience. Adjusting this settings sometimes helps to more readily identify the audio peaks that are associated with dialogue. When you have set your markers and typed in your text, hit the Grab Times button. This will automatically make an entry in the grid below with the start and stop times, and the subtitle line. Very accurate, and fairly simple to do. Advance along the wav form and repeat for your dialogue. To do this you must have some sort of a video or audio reference so you know what to dialogue corresponds to the speech. Now if you know the movie's original language and are doing you own translation, the wav form is all you need to reference. However, most of us are going to require a visual reference.
Chances are your visual reference is going to be a vcd, vhs tape, or dvd with non-removable burned in subtitles. Now there isn't really a way to automatically extract the subtitles on these sources. Unfortunately you are going to have to manually type out the subtitle lines. Now you can use your dvd player or vcr and manually pause and unpause and you go along transcribing, but this is going to put wear and tear on your decks. I highly recommend by a video capture device of some sort to digitize your source. If you have a VCD, simply copy the files to your computer and playback from there. I utilize an inexpensive tv tuner that has video/audio capture abilities. I digitize my source to an mpeg 1 file, and then run it through an editing program while I subtitle. For this I use m1-edit (www.mediawaresolutions.com), as it has frame by frame advance capabilities which are very useful for shifting back and forth to isolate lines. However, Windows Media Player is free, and you can use it although it has cumbersome forward and reverse capability. When you have your file and player sorted out, what you do is have both substation alpha and the mpeg player running. Go to a line of dialogue in the mpeg file, and listen to the speech and associated line. Now find the corresponding section in the wav form in substation alpha and subtitle that line. Very tiresome but it beats learning another language. Repeat ad nauseam until your movie is finished. Periodically save your progress in substation alpha, and then when you are done you should have a nice .SSA file ready to import into dvdsubber.
Section 5: Using DVDSUBBER
At its most basic level, dvdsubber (www.dvdsubber.com) is a program with two main components. The first component is a software dvd player shell, or front end. It is a shell because it lacks its own filters to decode dvd audio and dvd video. These filters come with other software dvd players such as Windvd or Powerdvd. Each has its own audio and video filter, which dvdsubber can utilize to playback dvds. Nothing revolutionary so far, but it does have a few nifty features like aspect manipulation and the sort. The revolutionary parts comes with the second component, the subtitling engine. DVDsubber contains a subtitling engine which is capable of synching a timed script to a dvd. By matching the start and stop of your script with the corresponding locations on the dvd, dvdsubber can make sure that your movie and subtitles stay in synch over the course of a film. This is VERY important as due to differences in time codes and sampling rates, your timed script would normally loose synch over the course of the film, and dvdsubber prevents this, provided your script was timed with a reasonable degree of accuracy. You would think this is a very simple thing, but we had nothing like it for years, and it has truly made subtitled playback a convenience. Gersen has made a decent user guide at dvdsubber.com, so please consult that guide for further info on dvdsubber.