Skip to content

Export captions

Export Captions writes one track of captions out as a separate subtitle file, for a platform or player that takes one. Burned-in captions need none of this: they’re part of the picture in any movie you export.

Menu
File ▸ Export Captions…
Agent tools
export_captions
  1. Choose File ▸ Export Captions…. It has no keyboard shortcut, and it’s greyed out until some track holds only text clips, with caption text on at least one of them.

  2. In the save panel, choose the Track and the Format. The name starts as your project’s name with .srt on the end.

  3. Choose Save.

Exporting changes nothing in the project, so there’s no undo step. If the file can’t be written, an alert says why.

The Track pop-up lists the tracks you can write, front-most first, with the front-most chosen. Each row reads like “Captions — 12 cues (track 4)”. The pop-up is greyed when there’s only one to choose.

Nothing marks a track as a caption track, so any track made only of text clips is listed, a titles track included. A track with any other kind of clip on it isn’t listed, and neither is one whose text clips are all empty.

With a nest open, the list shows that nest’s tracks.

The Format pop-up offers SubRip (.srt), the default, and WebVTT (.vtt). Choose whichever your player or platform asks for. Changing it changes the extension in the name field and keeps the name you’ve typed.

The words and times are the same in both. The files differ like this:

SubRip WebVTT
First line of the file The first cue WEBVTT
A number above each cue Yes No
Timestamps 00:00:04,133 00:00:04.133
&, < and > in the text Written as they are Escaped

Both use Unix (LF) line endings and end with one newline.

Each text clip on the track becomes one cue, at the time and with the words it has now. A caption you dragged exports where you dragged it. Styling plays no part, so restyling a caption doesn’t change the file.

  • A caption with no text is left out.
  • Line endings become LF. Blank lines inside a caption are removed, and so are spaces at the ends of its lines.
  • A word caption track is written one cue per page, with the page’s words as written. The style’s UPPERCASE, the word timings and the styling don’t go in the file.

Export a track and import it again, and every cue goes back on the frame it came from, at every ordinary frame rate.

The file itself may not match the one you first imported, because import moves each time onto a whole frame. At 30 fps, 00:00:04,120 comes back as 00:00:04,133. A file only comes back byte for byte when its cue times are already on whole frames and a frame lasts a whole number of milliseconds, as it does at 25 or 50 fps.

export_captions takes a path, and can be given a track and a format. With no format it follows the path’s .srt or .vtt, and a format you name wins over the extension.

It never guesses the track. With no track to write, it writes nothing and says so. With more than one and none named, it writes nothing and lists the candidates. A track it writes comes back through import_captions with every cue where it was. See export_captions.