I have a .srt file that looks like this :

00:00:06,862 --> 00:00:09,408
It's the first sense you use when you're born.

00:00:09,408 --> 00:00:11,283
One out of every fifty of your genes

00:00:11,283 --> 00:00:12,625
is dedicated to it.

These are subtitles for a TEDed video that I'm translating. I'd like to check the spelling and grammar of what I've written, but I can't find how. Most browsers and subtitles editors have some sort of spell check, but all they do is check if a group of letters exists in a dictionary. I have software to check both the spelling and grammar, taking into account the context, but it doesn't know how to read a .srt file.

So I guess I need a way to ignore all the "non text" part of the file, removing the carriage returns too, AND be able to go back to the srt file, once I'm done.

I'm thinking I could use some sort of invisible character that will be ignore by my spell checker, but be able to mark where the lines have to be cut.

Any ideas ?

  • rename srt to .docx and open it in office word, or .odt for open office. – MAKZ Feb 15 '14 at 10:44
  • or, what extensions is supported by the software you are using ? rename .srt to that extension – MAKZ Feb 15 '14 at 10:45
  • My question is not "how to open a text file" – Manu Feb 15 '14 at 11:06
  • 1
    I think you should try to reword the question a bit, like "How can I use spell check on sentences spaning multiple subtitle lines" or something similar. Just to avoid being disambiguous. – Mario Feb 15 '14 at 11:16
  • 1
    @MAKZ That's true, but Word doesn't know how to "skip" the timing lines. It can/will ignore them, but it won't be able to link multiple segments into one sentence. This will, however, only diminish some basic grammar checks, not the overall spell checking functionality. – Mario Feb 15 '14 at 11:24

While this isn't a automated or programmatic solution to the problem, this is how I usually handle such files when translating:

  • If a subtitle line doesn't end mid-sentence, I keep everything as is.
  • If a sentence continues in the next subtitle line, I merge all parts (or at least the rest of the sentence) into the very first line (which will obviously become too long).
  • Then, I'm using just default spelling tools provided, e.g. in Microsoft Word. It's pretty easy to simply skip over the lines with timing information.
  • Once this is done, I go through the lines and split subtitles that are too long once again.

Obviously, this is far from being perfect, and based on the length of your actual lines it might be very hard or impossible to do.

  • Yes, it doesn't look like there's a tool to do that. I'm hacking something in javascript, the question may have been a better fit for stackoverflow :) – Manu Feb 15 '14 at 11:22
  • 1
    Keep in mind that SO isn't "do the work for me", but in case you run into problems, you could ask a question there and then link this mention as reference (plus update/edit this question with any related stuff you're asking). – Mario Feb 15 '14 at 11:25
  • Yes, I was thinking of posting the question and my solution on SO, so that someone else may use or improve it. – Manu Feb 15 '14 at 11:27

Use this freeware program Subtitle Workshop like I do. Program uses Word's spelling checker. So you simply click on "Replace" or "Replace all" when you are in Spell Checker dialog. Till now I checked about 2000 subtitles. Sometimes I open serbian subtitle and with Subtitle Workshop program I create croatian subtitle.


You haven't included anything about your operating system environment. For Linux, it dosen't rely on extensions so any text editor can import it. For Windows.. rename the extension to .txt if notepad et al. is too stupid to open it.

  • My question is not "how to open a text file", it's "how to edit only the text portion of a .srt, while still being able to return to the .srt format". Thanks for the down vote. – Manu Feb 15 '14 at 11:07
  • 1
    You can open any file in Notepad and other text editors. E.g. Word will interpret it as a text file and ask you which encoding to use etc. From what I understood, the actual problem is the fact that your standard word processor won't be able to pick up the full sentences, since they're separated by meta data. – Mario Feb 15 '14 at 11:09
  • sort of, it's not really metadata, but I need one line out of 3. I'm trying to hack something in javascript. – Manu Feb 15 '14 at 11:15
  • Ok, I understand the problem now. What you are asking is not something that a program can "just do". Besides that, withdrawing all the non-text lines just to run it through a spell checker sounds like an inefficient way to go about things. Regardless, I would recommend you take up bash or python scripting and devise some code to parse the output as you intend (good luck doing any of this on a non-linux machine). As for trying to do this in javascript, I can only laugh. Javascript is not a scripting language, so have fun. – Peter Feb 15 '14 at 11:52
  • 1
    And that's exactly the issue desribed above - the goal is to automate that part. – Mario Feb 21 '14 at 11:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.