Convert VTT captions to SRT

What a VTT file contains that SRT cannot

WebVTT was designed for the web and carries more than SRT does. A real-world file can include a header with metadata, comment blocks, a stylesheet, named regions of the screen, and per-cue settings that position and align each caption. Here is a small one.

WEBVTT

NOTE Exported from the editing suite

intro
00:00:01.000 --> 00:00:03.500 align:start position:10%
<v Anna>Welcome to the show.</v>

The SRT you get from it keeps only what SRT can express.

1
00:00:01,000 --> 00:00:03,500
Anna: Welcome to the show.

The header, the note, the cue identifier and the settings are gone, the cue is numbered, the dot became a comma, and the voice tag became a plain speaker label.

When to convert VTT to SRT

Most desktop video editors, many TV and set-top box players, and a lot of older subtitle tools read SRT and nothing else. If a web platform gave you VTT and your editor wants SRT, this is the step in between. It is also the better format to archive in if you want one that nearly everything can open, since SRT is plain text with almost no rules to break.

What is lost, and what is not

• Positioning and alignment (the line, position, align and size settings) cannot be represented in SRT and are dropped. If your captions rely on being at the top of the frame to avoid on-screen text, check them after converting.
• Styling from a STYLE block is dropped. Inline <i>, <b> and <u> are kept.
• Karaoke-style timestamps inside a cue (a time in angle brackets partway through the text) are removed, leaving the whole line.
• Entities such as & and < are turned back into the characters they stand for, because an SRT player will not decode them.

Header, notes and the first cue

The first block of a VTT file is the header, and it can run on for several lines of metadata until the first blank line. The converter skips that whole block, so a file that starts with WEBVTT - Title and a Kind: captions line converts the same as a bare one. If your file has no WEBVTT header at all, it is read as SRT-style text and a note tells you so. A cue identifier above a timestamp is used as the cue's name in VTT and is not needed in SRT, where the cues are simply numbered from 1 in order.

Related fixes

If the converted subtitles are consistently early or late, use the subtitle time shifter. If you only want the words without any timing, the SRT to text tool reads VTT as well. To go the other way, use the SRT to VTT converter. For what each format is for, see transcript file formats explained, and for making SRT files from a recording see how to make SRT subtitles from a video.