Subtitles are one of the easiest ways to make a video easier to follow. They help viewers who are deaf or hard of hearing, people watching with the sound off on a train or in an office, and anyone listening in their second language. Most video tools ask for them as a small text file, and that file is almost always one of two kinds: SRT or VTT.
The two formats look nearly the same and do nearly the same job, which is why people mix them up. This guide shows exactly how they differ, which one to pick for common situations, and how to produce a subtitle file from a recording without uploading the video anywhere.
What a subtitle file is
A subtitle file is plain text. It holds a list of cues, and each cue is a start time, an end time and the words to show between them. The video itself is untouched: a player reads the video and the subtitle file side by side and draws the words over the picture at the right moments. Because the file is separate, you can add, swap or remove subtitles without re-exporting the video, and you can offer several languages from one video.
That is different from “burned-in” or “open” captions, which are drawn into the pictures of the video itself and cannot be turned off. Burned-in captions need a video editor. SRT and VTT are the switchable kind.
SRT: the format everything understands
SRT, short for SubRip, is the older and more widely accepted of the two. Each cue has a running number, a time range and one or more lines of text, with a blank line between cues. Times are written as hours, minutes, seconds, a comma and milliseconds.
1 00:00:12,400 --> 00:00:15,100 How does a cell decide 2 00:00:15,100 --> 00:00:17,800 what gets in?
Nearly every video editor, desktop player and video platform can read SRT, which makes it the safe default. If a form simply says “upload subtitles” and does not say which kind, SRT is the format to try first.
Try it on your own file
VTT: the web’s own format
VTT, or WebVTT, was designed for the web. A VTT file starts with the line WEBVTT, the numbers are optional, and times use a dot before the milliseconds instead of a comma. The same two cues look like this.
WEBVTT 00:00:12.400 --> 00:00:15.100 How does a cell decide 00:00:15.100 --> 00:00:17.800 what gets in?
VTT is the format that browsers understand natively. If you put a video on your own web page with the HTML video element, the track element that attaches subtitles expects a VTT file. VTT can also carry extras such as positioning and styling that SRT cannot, though plain text cues are all most videos need.
Which one should you choose?
The honest answer is that it rarely matters, because the content is the same and you can export both from one transcript. A simple rule covers most cases.
- Uploading to a video platform, a social network or a learning system: try SRT first. Most of them also accept VTT, so use whichever the upload form lists.
- Adding subtitles in a video editor, or opening a video in a desktop player: SRT.
- Playing a video on your own web page: VTT.
- Not sure: export both. They are tiny files.
Making a subtitle file from your video
Writing cues by hand is slow, so most people generate a first draft. In OmniaScribe, pick your video on the subtitle generator page, or on the video to text page, and the transcription runs in your browser. Only the sound track is used, and the video is never uploaded. If you are on the VTT subtitle generator, the VTT export is highlighted on the result instead of SRT.
When the transcript is ready, use the SRT or VTT export button. Long lines are split into short one-line cues of about 42 characters, which is comfortable to read on a screen, and the time of each line is shared out by its length so a cue stays up in step with the speech. Cues never overlap, and the file is plain text.
Before you export, read through the transcript once. Names, brand terms and numbers are the words speech models get wrong most often, and the Edit view lets you fix them. Editing changes the words only, so the timing stays matched to the speech and the subtitles you export carry your corrections.
Adding the file to your video
The exact buttons differ by service, but the pattern is the same almost everywhere: open the video’s settings, find the subtitles or captions section, choose to add a language, upload the file, and publish. Check the result by watching a minute or two with subtitles turned on.
- On a video platform, look in the video’s details or editor for “Subtitles” or “Captions”, then “Upload file”.
- In a desktop player, drop the subtitle file onto the playing video, or give it the same name as the video and keep both in one folder.
- On your own web page, add a track element inside the video element, point it at the VTT file and set its language.
- In a video editor, import the SRT as a captions track and style it there. You can also export the video with the captions burned in.
If the words and the speech drift apart, the subtitle file and video probably do not match, for example when the video was trimmed after the transcript was made. Transcribe the trimmed version instead.
Making subtitles people actually read
A subtitle can be correct and still hard to read. Viewers need time to take in each cue, so very fast speech produces cues that flash by. When a stretch feels rushed, shorten the wording in the Edit view: fewer words in the same time are easier to follow, and the meaning usually survives.
Keep an eye on punctuation and capitalization, since automatic transcripts handle both well but not perfectly. Spell names consistently. If the video has speakers who matter to the story, add a short label such as the person’s name at the start of the cue. The engine does not tell voices apart yet, so labels are yours to add.
Finally, remember that subtitles are also text. Keep the transcript with the video’s page where it helps readers, and keep your original files, because you can re-export both formats whenever a platform asks for something different.