SRT Files Explained: How to Create, Edit and Use Them

SRT is one of the most widely supported subtitle file formats, and one of the easiest to break. Here is what is inside the file, how timestamps work, and how to avoid the classic errors.

What an SRT file is

SRT stands for SubRip Subtitle. It is one of the oldest and most widely supported subtitle file formats: a plain-text file with a sequence of timed cues. There is no video, no font, and no color. It records when each line appears, when it disappears, and what it says.

Because it is plain text, an SRT can be opened in any editor, diffed, translated, and regenerated by a script. That simplicity is why it is still a common exchange format for subtitles.

Anatomy of an SRT cue

A cue has four parts: an index number, a timestamp line, one or more lines of text, and a blank line before the next cue.

1
00:00:01,000 --> 00:00:04,200
Most people scroll past
a video without sound.

2
00:00:04,400 --> 00:00:07,800
That is why captions matter.

The timestamp format is HH:MM:SS,mmm. Hours, minutes, seconds, then milliseconds after a comma. The arrow is two hyphens and a greater-than sign. The blank line between cues is the standard convention; some parsers tolerate variations, but others will merge or drop cues, so keep it.

Save the file as UTF-8. Accented characters and non-Latin scripts depend on it.

The rules that actually matter

These are the safest common conventions, not a formal standard. Player behavior varies, so test the file in the destination you care about.

• Number cues in order. Sequential numbering is recommended for compatibility; some parsers tolerate gaps, but you should not rely on it.
• Use exactly --> between start and end.
• Use commas for milliseconds, not periods.
• Leave one blank line between cues, and no blank lines inside a cue. This is the convention most parsers expect.
• Aim for one or two lines, roughly 15 to 20 characters per second.
• Avoid overlapping cues unless you know the target player handles them.
• Keep styling minimal. Bold/italic tags are inconsistently supported.

Common errors and how to fix them

1. The comma/period mistake. 00:00:01.000 is VTT style; SRT expects 00:00:01,000.
2. A missing blank line. Two cues can run together, and the player may show the wrong text or none at all.
3. Wrong arrow. A Unicode arrow or a single hyphen is not the format.
4. Overlapping timestamps. A cue starts before the previous one ends; some editors silently drop text.
5. Encoding damage. Accents turn into symbols; re-save as UTF-8.
6. Reading speed. A perfect timestamp with 20 words on screen is still unreadable.
7. Drift. Converting from another format or a different frame rate can shift every cue. Spot-check the first, middle, and last.

Creating an SRT with Capsly

1. 01

   Upload the video or audio file

   The video stays on your device. Audio-only files work too.

2. 02

   Transcribe

   Only the extracted audio is uploaded for speech-to-text. After transcription, the audio is deleted, with a lifecycle backstop for failed requests. Language detection and word-level timing are automatic.

3. 03

   Correct names, numbers, and timing

   Word-level timing makes small nudges easy. Fix the text before polishing the look.

4. 04

   Export the SRT

   Choose SRT (and VTT) sidecars. Add the rendered MP4 if you also need burned-in captions.

The free beta includes 15 minutes of transcription per month and 10 uploads per day, with clips up to five minutes.

If you are starting from a video rather than an existing subtitle file, the upload-to-export workflow covers the full path from a raw clip to SRT, VTT or a burned-in MP4.

Using an SRT in the real world

• YouTube Studio: upload it as a subtitle track under the video’s Subtitles section.
• Editors: import it as a caption track; most NLEs accept SRT directly.
• Players: keep the SRT next to the video with the same base filename so it loads automatically.
• Translation: send the SRT to a translator; keep the timestamps intact.
• Social apps: Reels, TikTok, and Shorts generally do not accept a sidecar SRT upload. Burn the captions in instead, and verify current support before relying on a file. Here is the comparison.

SRT vs VTT

WebVTT is the web-native cousin. It starts with a WEBVTT header, uses periods for milliseconds, and supports more cue settings and styling. Browsers generally prefer VTT; many editors and platforms still accept SRT. Capsly exports both, so you can match the destination instead of converting by hand.

If a platform accepts only one, check before you export. If it accepts both, VTT is often the better fit for web players; SRT is often the safer bet for editing software.

Edit or regenerate?

If the words are right and only a timestamp or line break is off, edit the SRT. If the transcript is wrong, the language is wrong, or the timing drifts throughout, regenerate from the source. Keep the source video and the project so you can redo it without starting over.

A small fix should take a minute. A full regeneration should be rare.