Subtitle Tools Pro

SRT vs WebVTT: What Actually Differs

SRT and WebVTT look almost identical in a text editor. The differences that matter are the ones that make a file silently fail in the wrong player.

8 min read · Updated 2026-11-16

Open an .srt and a .vtt side by side and they look like the same format with different punctuation. Both have a header of sorts, both have timestamp lines with an arrow, both have text underneath. Convert one to the other and the diff is a few characters per line.

That similarity is misleading. WebVTT was written as a replacement for SRT, by people who had spent years dealing with malformed SRT files, and almost every difference exists because SRT had no answer for a specific problem. Knowing which differences are cosmetic and which will break playback is the difference between a conversion that works and one that produces a file the player ignores.

The short answer

Use SRT when you are handing subtitles to something old or something that only accepts SRT: most desktop players, most standalone subtitle editors, a large number of broadcast and DVD workflows, and many translation agencies. Use WebVTT when the target is a web browser, an HTML5 video element, or a modern streaming platform that states VTT in its documentation.

The filename extension is not a reliable signal. Plenty of files named .vtt contain SRT content with a WEBVTT line bolted onto the top, and plenty of .srt files contain WebVTT cue settings that the SRT parser will choke on. Check the content, not the name.

SRT has no header at all. The file starts with a cue number. WebVTT requires the first line to begin with the five characters WEBVTT, and the spec allows that line to carry a description after a space:

WEBVTT - Chapter 1
WEBVTT Kind: captions
WEBVTT Language: en

All three are valid. This is worth spelling out because it is a common source of failure in conversion tools: a parser that tests for the exact string "WEBVTT" and nothing else will reject the other two forms, which appear in real files produced by real encoders. If a converter tells you a VTT file is missing its header when you can plainly see the header, this is why.

In the other direction, converting VTT to SRT means the header line has to go. SRT has nowhere to put it, and leaving it in produces a file whose first block is not a cue.

2. The decimal separator

This is the difference everyone knows: SRT separates seconds from milliseconds with a comma, WebVTT with a period.

SRT: 00:00:01,000 --> 00:00:03,500
VTT: 00:00:01.000 --> 00:00:03.500

It is a single character and it is the most common cause of a subtitle file being rejected outright. It is also the least interesting difference, because every conversion tool handles it.

3. Optional hours

WebVTT allows the hour field to be omitted when it is zero, so 00:01:23.000 can be written 01:23.000. SRT has no such shorthand. A parser that assumes the three-group form will misread or reject the short one, and a parser that only accepts the short form will fail on anything longer than an hour.

4. Cue identifiers are optional, and can be text

In SRT, the number above each cue is mandatory and sequential. In WebVTT the identifier is optional and, when present, is an arbitrary string that never appears on screen:

WEBVTT

intro
00:00:01.000 --> 00:00:02.000
Hello

scene-3
00:00:03.000 --> 00:00:04.000
Goodbye

Converting to SRT means discarding those identifiers and writing sequential numbers instead. Discarding them is not optional: an SRT file whose cue blocks start with the word "intro" is malformed, and players vary in whether they skip the block, drop the whole file, or show "intro" on screen.

5. Cue settings

This is the largest structural difference. WebVTT lets a cue carry display settings on the timestamp line itself:

00:00:01.000 --> 00:00:03.500 line:90% align:center position:10% size:80%

These control where on the video the text appears, how it is aligned, and how wide the text box is. SRT has no equivalent — position in SRT is whatever the player decides. So a VTT to SRT conversion has to drop them, and that drop is lossy: the text survives, the layout does not.

The reverse is also worth stating. Converting SRT to WebVTT cannot invent cue settings, because SRT never had them. If your converted VTT file looks badly positioned, that is not a bug in the conversion — the information was never in the source.

A parser that validates the timestamp line against a pattern anchored at the end of the line will reject every cue that has settings. That looks like a corrupt file and is actually a stricter-than-spec parser.

6. STYLE and REGION blocks

WebVTT can carry CSS at the top of the file, in a STYLE block, which applies to all cues:

WEBVTT

STYLE
::cue {
  color: yellow;
  background: rgba(0,0,0,0.8);
}

There is also a REGION block for defining named display areas that cues can be assigned to. SRT has neither. Both have to be stripped on conversion to SRT, and the visual result will differ from what the VTT author saw.

7. Comments

WebVTT has a NOTE block for comments that should not be displayed. It is used for translator notes, timing reminders, and sometimes to disable a cue temporarily without deleting it. SRT has no comment syntax at all — anything in an SRT file is either a cue or a parse error.

8. Inline markup

Both formats support italic, bold, and underline, and both use angle-bracket tags, which is why the two are so often confused. WebVTT goes further with voice spans and language spans:

<v Anna>Where are we going?</v>
<lang fr>Bonjour</lang>

Voice spans matter for accessibility — a screen reader can announce who is speaking. They have no SRT equivalent, so converting to SRT means either dropping the markup or flattening it into the text, depending on the tool. Flattening is usually worse: it puts a speaker name into the subtitle line where a viewer will read it as dialogue.

9. Encoding expectations

WebVTT is defined as UTF-8. SRT predates that consensus and files circulate in Windows-1252, ISO-8859-1, Shift-JIS, GBK, and UTF-8 with and without a byte order mark. In practice this is the second most common cause of garbled subtitles after the decimal separator, and it is the one a conversion tool can least help with, because the tool has to guess the source encoding.

If converted subtitles show mojibake — accented letters replaced by pairs of odd symbols — the file was read in the wrong encoding, not mistranslated. Re-save the source as UTF-8 before converting.

10. Line endings

WebVTT specifies that lines end with a newline or a carriage return plus newline. SRT has no specification at all, so files exist with LF, CRLF, and occasionally CR alone. In practice every parser handles both common forms, but a file with mixed line endings inside a single file will produce cues that appear to lose their text. Normalising to one style before converting avoids the problem entirely.

11. Metadata and chapters

WebVTT files distributed with HLS streams can carry metadata cues and chapter information that are not subtitles at all — they are consumed by the player to build a chapter menu. Converting such a file to SRT produces something that looks like subtitles but is really a list of chapter titles with timestamps.

12. What actually supports what

TargetSRTWebVTT
HTML5 <video> with <track>NoYes
VLC, mpv, most desktop playersYesYes
YouTube uploadYesYes
Netflix and most broadcast deliveriesVia specific specNo
Premiere, DaVinci Resolve, AegisubYesPartial
HLS and DASH streamingNoYes
DVD and Blu-ray authoringYesNo

The row that surprises people is the first one. A browser will not load an SRT file through a track element, no matter how correct the file is — it will fail silently, with no error in the console, and the video simply plays without subtitles. This is the single most common reason someone needs to convert SRT to VTT.

Converting without losing what matters

A conversion between these two formats is not symmetric. SRT to WebVTT adds a header and changes a comma to a period; nothing is lost, because WebVTT can express everything SRT can. WebVTT to SRT is lossy: cue settings, styles, regions, comments, voice spans, and non-numeric cue identifiers all have to go.

  1. Convert from the original file, not from a previously converted one. Round-tripping VTT to SRT and back will not restore the styling.
  2. Check the encoding of the source before converting, and re-save as UTF-8 if it is anything else.
  3. After converting to SRT, open the result and confirm every block starts with a number, not a leftover cue identifier.
  4. After converting to VTT, confirm the first line starts with WEBVTT and that there is a blank line before the first cue.
  5. Play the result in the player you actually intend to use. A file can be spec-correct and still be rejected by a strict player.

Try it on your own file — no upload, nothing leaves your browser.

Open the SRT to VTT Converter (and VTT to SRT) - Free, Online →

If the result still looks wrong, the next question is usually whether the file is valid at all rather than whether the conversion worked — which is a different check, and a different tool.