The project
Use a short spoken sequence with a settled picture edit. Decide whether the destination needs styled burn-in captions, a subtitle sidecar, or both as separate deliverables. If you need a portrait version, prepare that canvas before the final typography pass.
1. Build the correct transcript
Open Edit Captions and generate from the intended dialogue track and range, with the correct spoken language and available model. Alternatively, import a subtitle file that matches this exact cut.
Read along with the audio. Correct names, specialist vocabulary, punctuation, and missing words. Preserve what was actually said rather than rewriting the speaker’s meaning for a punchier caption.
2. Refine cue boundaries
Break phrases where they are naturally read. Avoid a single very long line or a series of one-word flashes when a stable phrase would be clearer. Review any timing and readability notices.
For word-by-word presentation, check individual highlights against the speech. Correct cue-level timing first, then inspect the smaller word intervals.
3. Style for the image
Choose a readable scale and strong contrast. Play through the moving background and use safe-area guides. Reserve room for the destination’s interface where necessary. Inspect the portrait composition separately from the landscape version.
Use animated styling to support emphasis, not to obscure each word as it arrives.
4. Deliver the intended form
Choose the correct caption track and language in Deliver. Burn-in becomes part of the picture. A sidecar remains a separate file, and its exact font and animation do not travel like rendered pixels.
Review your result
Watch the exported movie at a small size, with sound and then muted. For sidecars, test in the intended receiving workflow. Check the first cue, speaker changes, and the final caption against the film’s ending.