What Is an SRT File? (Format Explained + How to Open It)

If you’ve ever downloaded captions for a video, or exported them from one, there’s a good chance the file ended in .srt. It’s not the only caption format, but it’s the one nearly everything else measures itself against.

The short version

An SRT file is a plain-text file that holds three things for every caption on screen: a sequence number, a timecode telling the video player when to show and hide the line, and the caption text itself. Nothing else, no audio, no video, no styling beyond a couple of basic tags. That simplicity is the entire reason it’s still the default format almost 25 years after it first showed up.

What’s actually inside one

Here’s a real SRT block, annotated:

14
00:02:17,440 --> 00:02:20,375
Alright, let's get into
the next section.

• 14 — the sequence number. Just tells the video player this is the 14th caption in the file, in order.
• 00:02:17,440 --> 00:02:20,375 — the timecode, in hours:minutes:seconds,milliseconds. This caption appears at 2 minutes 17.44 seconds into the video and disappears at 2 minutes 20.375 seconds.
• Alright, let's get into the next section. — the caption text. One or two lines, kept short on purpose so it’s readable at a glance.

A blank line separates this block from the next one, and the whole file is just that pattern repeated for every line spoken in the video. A ten-minute video with steady dialogue might produce 80-150 of these blocks back to back.

The formatting it does support

SRT isn’t purely plain text, it has a handful of basic HTML-style tags baked in:

2
00:00:05,000 --> 00:00:07,500
<i>Off-screen voice:</i> Wait, is that <b>everyone</b>?

• <i>...</i> for italics, usually reserved for off-screen speech, thoughts, or song lyrics
• <b>...</b> for bold, used sparingly to emphasize a single word
• <u>...</u> for underline
• <font color="...">...</font> to change text color on players that support it

That’s the whole list. You can nest them (italic and colored text together, for instance) and most modern players render that correctly, but plenty of platforms strip formatting entirely on upload. Don’t build a captioning workflow that depends on it looking the same everywhere, test on the platform you’re publishing to.

What it can’t do

Beyond those four tags, an SRT file has no idea what font you’re using, where on screen the text should sit beyond a rough default, or what your brand’s colors are. There’s no field for language metadata, no way to size text differently per line, and no support for the placeholder or pluralization logic that translation workflows sometimes need. If you’ve seen fancier animated captions with custom fonts and word-by-word highlighting, that styling is coming from the app rendering the video, not from anything stored in the SRT file itself. The format is deliberately dumb, in a good way: dumb formats are the ones that still open correctly in software that didn’t exist when the format was created.

Why the file’s encoding matters more than people expect

This is the part that quietly breaks more SRT files than anything else. An SRT file is text, but how that text is saved (its character encoding) determines whether special characters, accented letters, or non-English scripts show up correctly or turn into question marks and garbled symbols.

You’ll see SRT files saved in Windows-1252, Windows-1251, plain Unicode, UTF-8, or UTF-16. UTF-8 is the safe default: it handles the widest range of characters and is what nearly every modern platform expects. If you’re captioning anything outside basic English, or translating a track into another language, save (or export) as UTF-8. This is also the single most common reason a perfectly well-formatted SRT file “doesn’t work” on a platform, the timing and text are fine, the encoding just wasn’t what the platform expected.

Where SRT files show up

It’s not just a YouTube thing. The same .srt file works, unmodified, across most places video lives:

• Social platforms — YouTube, Instagram, LinkedIn, and Facebook all accept SRT uploads directly
• Video editors — Premiere Pro, Final Cut Pro, and DaVinci Resolve all import and export SRT natively, so captions survive a round trip through your edit
• Streaming and course platforms — most smaller streaming services and course platforms take SRT as their default caption format, since building support for anything else isn’t worth it when SRT already covers the case
• Other captioning tools — SRT is the common export format almost every captioning tool, CentClip included, offers, so a file made in one tool opens fine in another

That portability is the whole reason it survived. A format that only worked in one app would’ve died with that app.

How to open an SRT file

Two different questions hide inside “how do I open this,” and they have different answers:

To read or edit the raw file: any plain text editor works. Notepad on Windows, TextEdit on Mac, or a code editor if you have one installed. You’ll see the numbers, timecodes, and text exactly as written above, not captions playing over video.

To watch it synced with your video: you need a video player that supports loading an external subtitle file. In VLC, that’s Media → Open File for the video, then Subtitle → Add Subtitle File to load the matching .srt. Most other desktop players (and every major video editor) work the same way: load the video, then separately point it at the caption file.

To upload it to a platform: YouTube, Facebook, and most social platforms have a dedicated captions/subtitles upload step, separate from the video upload itself, usually under a “Subtitles” or “CC” tab once your video is live or in the upload flow.

Writing captions people can actually read

The format is simple; writing captions that don’t frustrate anyone reading them takes a bit more care. A few rules that hold up across almost every platform:

• Keep each line to roughly 32-40 characters. Much longer than that and the line has to be read faster than most people comfortably read, especially on a small screen.
• Cap it at two lines per block. A third line pushes text further up the frame and covers more of the video than it should.
• Break at natural pauses, not mid-sentence. Splitting after a comma or at the end of a thought reads far more naturally than a mechanical every-N-words split.
• Give each line enough time on screen. As a rough guide, if a line takes longer to read out loud than it stays visible, viewers won’t finish it. Don’t cut a caption’s display time just to keep pace with fast dialogue.
• Proofread before publishing. Auto-generated captions get names, technical terms, and homophones wrong often enough that a quick pass matters, especially since a caption typo is more visible and more embarrassing than a typo almost anywhere else in your content.

None of this is enforced by the SRT format itself, nothing stops you from writing a 90-character single line that flashes on screen for half a second. It’s just what separates captions that get read from captions that get skipped.

Making your own

You can type an SRT file by hand in a text editor if you want to, the structure above is all there is to learn. For anything longer than a 30-second clip, though, hand-timing every line against the audio gets tedious fast, and a single miscounted timecode throws captions out of sync for the rest of the video.

The faster path is auto-transcription: upload your video, the audio gets converted to text with timing attached automatically, and you download the finished .srt. That’s what CentClip’s SRT generator does, 5 cents a minute, no subscription. If your platform prefers a different format, the same transcript exports as VTT too, since the underlying data (text plus timecodes) is identical either way.

The honest limitation

SRT is a captions format, not a video editor. It won’t add animated word-by-word highlighting, custom fonts, or brand colors, some competitors handle that natively in their own proprietary export format, which locks your captions into their tool. Plain SRT/VTT is the tradeoff in the other direction: less flashy out of the box, but it opens in literally everything, forever, because it’s just text.