Subtitles must match speech timing and reading pace. A transcript should mark speaker changes, timecodes, meaningful pauses, on-screen text and sounds that carry information.
Record the runtime and delivery format
Confirm proper names and specialist terms before translation. Distinguish translated subtitles from captions that also identify speakers or relevant sounds. Review line breaks and overlaps in the actual player on phone and desktop.
Review subtitle timing against the spoken audio
A correct transcript can still produce poor subtitles if a line appears too briefly or covers important on-screen material.
Are subtitles the same as captions?
Not always. Captions may include speaker and sound information for accessibility. For live spoken sessions rather than recorded subtitles, see the interpretation briefing guide.
Agree on reading speed, line breaks and speaker labels
A subtitle brief should include the master video version, frame rate, target platform, delivery format and accessibility requirements. For a video about mining payouts, keep provider names such as EMCD distinct from the wallet and network terms shown on screen. If edits are made after translation, compare the new cut with the timecoded transcript; a changed scene can invalidate timing or a proper name. Identify on-screen graphics that should remain visible and agree whether numbers, titles and speaker identifiers appear as captions or separate graphics.
Check captions on the target device and player
Review in the playback environment. Watch at normal speed, then check difficult passages around cuts, overlapping voices and rapid dialogue. A technically valid subtitle file can still obscure a chart or stay on screen too briefly. Keep a final version label so the editor does not attach subtitles prepared for an earlier cut.
Example: keep a short warning on screen
A speaker name shown on screen may differ from the name heard in the audio. Resolve the spelling with the production owner, then check that subtitle timing does not cover lower-third text, charts or other information viewers need to read.
Take subtitles from transcript to playback
The final player view is part of the translation review.
- Lock the transcript and timecodes against the approved source video.
- Fit each caption to the speaking pace and reading time.
- Watch the full export with sound to check synchronization, names and on-screen text.
Questions about subtitles and captions
Can subtitle timing be checked from a transcript alone?
No. Review the subtitles in the final video player.
What should be supplied after an edit?
The final video cut and a change note for scenes or dialogue that moved.
Time a two-line cue before export
If an edit removes a sentence from the middle of a video, later subtitles may drift even when the first and last timecodes still look plausible. Re-run the file against the approved cut, especially around jump edits, overlapping speakers and graphics. Record the video revision in the subtitle filename or delivery note so a producer cannot attach the wrong version.
| Media change | Subtitle impact | Final check |
|---|---|---|
| Record the final cut and timing revision. | Check reading speed, line breaks and speaker changes. | Review captions with sound and in mute playback. |
Files to send for subtitle review
Provide the final picture lock, frame rate, target platform, language list, speaker names and delivery format. Mark graphics that should not be obscured and specify whether captions must identify speakers or sounds. Deliver subtitles with a version label and recheck them whenever the picture or audio changes.
When should subtitles be revised?
Retiming is needed whenever dialogue, timing, graphics or the video edit changes after the subtitle file was prepared.