You can download text srt subs for many movies and tv-series from
if you can't find it you must convert manually:
DVD to srt text subtitles:
Use subrip or dvdsubedit. (See guides below)
idx/sub to srt text subtitles:
Use Subtitle Edit or subrip or subresync (included in vobsub) or AviDemux(Tools->OCR(sub to srt)).
Rip the DVD to your HDD using DVDFab Decrypter first.
Select File->Open full domain and select the first vob from the main movie, usually the first VTS_01_1.vob or the first 1 GB file
Select the subtitle stream you want to convert under Subpic selection at top right
Select OCR -> OCR selected subpics
Last save to a srt at the bottom right under Save as srt.
Edit the text in for example notepad to fix OCR errors.
1. Select File->Open Vob(s)
2. Click Open IFO and select the .idx or the VTS_XX_X.vob from the VIDEO_TS folder(if a commercial dvd rip it first to your HD with DVD Decrypter)
3. Select the Language stream
4. Select Subpictures to Text via OCR
Identify all characters manually.
Last save text file to a srt
1. File->Save as
If subrip wont work try an older version like subrip 1.16 or the latest beta.
+ Reply to Thread
Results 1 to 30 of 71
Last edited by Baldrick; 16th Mar 2016 at 03:11. Reason: Removed dead subtitle link sites
Thanks for this, I was pulling my hairs out wondering how to use the sub/idx files. Question is, do I need to go through this processes for each of the languages in the idx, or is there a way that AVI2DVD will accept all the subtitle streams?
[EDIT] Never mind, I simply went ahead with the process for each of the languages, and used the individual subtitle streams. A pain, but not a tragedy. However, I do still wonder whether it is possible to work directly with sub/idx files. Perhaps after enconding to DVD?
I'll give that a go after work...
I am aware of this feature. In fact I don't want to use it, as I want an exact match.
I would expect that the ripped subtitle timing should match with the movie it is ripped from, and as it isn't I suppose I'm doing something wrong.
Get my idea?
i hope this is the right place to ask. i downloaded SubRip1.50beta and VobSub2.23 to convert *.idx and *.sub files to the *.srt format. i fired up subrip and loaded the files and did all the steps necessary to convert them. but however it kept asking me to fill each individual line in there instead of just running it through some process and giving me the final output in the *.srt file (at least that was my idea how it should work ). well then i tried the SubResync included in the vobsub program (which looked more easier and user friendly than the previous) but it was the same. it showed the text assigned to each dialog in the movie but i had to fill it in there... and when i just clicked OK through the all file to the end then it showed in the movie some [skipped line n.65] or something like that...
my question is. do i have to go through the whole movie and fill every single line there or is there another way to do it? thnx.
Originally Posted by daniel123
I did manage to find a sub (.srt) for a French movie I have at www.divxsubtitles.net so you may want to keep searching.
lol. so much for 'converting' the subtitles. its really kinda absurd... why do i have to fill it in there when the lines are already included in the *.idx and *.sub files. if i knew the language i might aswell do my own subs and it would be the same. i hope someone is gona write quick some plug-in or whatever to fix this.
thanks for the link anyways. will try..
For OSX users, there's a Universal Binary version of Miyu available here:
While ffmpegX seems to work with most ".srt" files (if you use the mencoder from version 9s), sometimes there's just something about that file that ffmpegX doesn't like. Miyu will load in an AVI and a ".srt" file and permit you to export it to a few different formats. As I write this I'm exporting to DV. Assuming the subtitle is embedded in that DV, I'll use ffmpegX to convert it to whatever other format I'll require.
I'll post with results when encoding is done; should be this evening.
why do i have to fill it in there when the lines are already included in the *.idx and *.sub files.
That's what the OCR process is all about. However, if you're lucky, you'll only have to fill in each letter once. Once for caps, smalls, italic caps (if any), and italics small (if any). The whole process takes maybe 10-15 minutes. Once you get most of the letters typed once, it really starts to go fast.
As stated, though, if you have the IDX/SUB files already, convert to SUP format subs using Sub2SUP, and then use an authoring app that accepts SUP files. Saves doing the OCR.
Partial Miyu results: I underestimated the GB necessary. The app seems to require twice as much space as you would think. Apparently, it takes the first-pass results (to DV) and then wants to "flatten" to another file which, presumably, would also be around 12+GB/hr.
The app was hanging after the first pass so I force-quit it and the "myfile.mov.tmp" file remained. It was complete and played as 640x480 in QT Pro. I should mention that the original project in Miyu was set up as 640x480 so this "DV" export is not "DV-Stream" but, instead, simply a "DV" file. A small distinction but it's worth noting if you're intending to transcode further.
By the way, the subtitle is now embedded in the movie. The original movie was letterbox but, as a 640x480 Miyu project, the export is 640x480 (really letterboxed) with the subtitle appearing in the lower black letterboxed area. If I play with the font, size, and style in Miyu, I'll really be able to tweak it nicely.
I've dropped this file into ffmpegX for a final conversion to H264 for iPod. Miyu has an H264 export capability but it uses Apple's QT engine for this so it's glacially slow. I use ffmpegX (2-pass, Qmin:3, Qmax:28, Exaustive, 20% add'l bitrate) and can connect my iPod to my TV.
So this is just one more tool in our toolboxes.
Is it possible for the IDX/SUB to have two subtitle streams, one in Chinese and one in English? After I run SubtoSup, I get 2 SUP files, one is ENG and one is CHI. That was a surprise. But how do I check to make sure it is what it is? It does not open with Notepad.
Subrip says the file is not recognized. Oh well, in SubResync it seems the sub times are off compared to the movie anyway. This is really annoying. All the subtitles I find online are mistimed to the actual DVD. I don't see why companies cannot put Chinese subtitles in their DVDs, it will make it all the more convenient.
Originally Posted by Sakuya
Originally Posted by Coolpplse
Originally Posted by manono
One other thing, on many of the lines (I'd say 2/3 at least) SubRip can't figure out where each letter starts and ends. I've tried using the "extend selection" button but then it starts skipping letters. So I went to the "enter manually" button so I could just type the whole line at once and get it over with. Is there any reason I shouldn't be using that?
Do you mean that it will start doing some of the text on its own?
Yes, not only some of it, but all of it. This assumes that each letter is the exact same every time it's repeated. That's not always true. But you type everything in the very beginning, and then when it comes to those letters again, you don't have to type them again. The letters you have to type get fewer and fewer, until before long it's racing along.
One other thing, on many of the lines (I'd say 2/3 at least) SubRip can't figure out where each letter starts and ends. I've tried using the "extend selection" button but then it starts skipping letters.
Meaning it'll show you 2 or 3 letters kind of joined together? You either type in those 2 or 3 letter combinations, or you go back and adjust the spacing in configure. I don't do that often, because I don't have a good understanding of how it works, and find it more trouble than it's worth. If it gets too annoying I usually quit SubRip and OCR using Sub Resynch (comes in the VobSub package). It requires IDX/SUB input (not IFO/VOB like SubRip), but that's easy to get. And it seems to have a better method of showing the individual letters.
For those who may not be aware of this, ai4spam (the current developer of SubRip) has made his character matrix set available. If you use this it will save you about 90% of the drudgery of OCRing using SubRip. It's basically a compilation of probably hundreds of OCR'ed DVD character matrices combined into one file. On most DVDs (or idx/sub pairs), you only have to enter a few characters, all the rest are recognized for you. On some DVDs, it works 100% and you don't have to enter anything. On a few others with really weird fonts, it may not work at all.
It still will make the odd mistake so be sure to run "Post OCR Spelling Correction" before you save, then spot-check it in Notepad or whatever after you're done.
To use: unrar the download and copy the file "ai4spam.sum" to the "ChMatrix" folder in SubRip. You want to load the matrix file before you start ripping: in the "What to Do?" window, in the box labeled "Characters matrix file" click "..." to browse to the ai4spam.sum file. It will take a few seconds to load. Click "start" and off you go.
If you're using this matrix, do not click on the "Search for an existing Matrix that has a match for this Character" box in the "New Character(s)" window that pops up when SubRip doesn't recognize a character. I've never done it myself, but I've read a couple posts saying that this can crash SubRip if the a14spam matrix is present.
The matrix is located at:
I've been using it for over a year and I love it. I have about 120 individual matrix files I made character by character before I discovered this jewel. No fun at all.
Hope you find it useful.
please don't hijack topics. Post a new topic instead. /Admin Baldrick