Write and test captions for the smallest intended viewing screen. Use short meaningful lines, leave room for platform overlays and keep text away from the couple's faces and hands. Caption creation and export may require an external editor; do not assume a generation workspace provides every subtitle format.
Decide how the captions will travel
Burned-in captions become part of the picture and remain visible when a video is copied elsewhere. A separate subtitle file can remain selectable and editable where the player supports it. A clean master without either is useful for future translations. Keep these deliverables distinct instead of expecting one file to satisfy every destination.
Adobe documents caption export choices in its own editing workflow. Consult your chosen editor's current documentation for the formats it supports. That does not imply equivalent controls exist in Imagild Film Studio, which you can use for the approved source shots.
Edit spoken language into readable units
Do not split a sentence only because a timer reached a round number. Keep a person's name, a short phrase and the word that gives a joke its meaning together when possible. Read the line aloud and watch it once without sound. If the viewer cannot understand who is speaking or what changed, revise the wording or timing.
Caption example: the groom says, “Apparently, saving the world was the easy part.” Show that complete thought over the reaction, then use a separate caption for the bride's reply. Do not reveal her punchline while the first line is still being spoken.
Avoid an arbitrary universal word limit. Typeface width, frame shape, language and display size change how much fits. Use a concrete layout test: export a frame at the intended dimensions, view it at phone size and confirm that no line touches the reserved margin. Shortening a line is often more effective than shrinking the font.
Reserve space for faces and platform controls
The bottom of a vertical video can contain a description, navigation and engagement controls. Those overlays vary by platform and may change. Check the current preview in each target app rather than treating a single safe-area diagram as permanent.
| Problem | Diagnostic | Revision |
|---|---|---|
| Words overlap hands | Pause at the ring exchange | Move or shorten the caption |
| A name wraps awkwardly | Check the longest name at phone size | Rephrase around an intact name |
| White text disappears | Inspect the brightest background | Use a restrained readable backing |
| Joke arrives too early | Watch speech and reaction together | Delay the revealing words |
Keep a subtitle review sheet
Record speaker, approved wording, pronunciation notes and timing corrections. Check proper names manually. Automated transcription can mishear names even when the rest of the sentence is correct. Distinguish captions for spoken dialogue from decorative titles so viewers do not mistake a slogan for speech.
Keep the clean master and final transcript. If you later make a square or vertical cut, reassess the caption position instead of copying coordinates from the landscape file. Reframing changes the relationship between text, faces and the important action.
Are burned-in captions always better?
They are dependable for visual consistency, but viewers cannot turn them off or easily change their language. A selectable subtitle track can be more flexible in a supported player. Choose according to the destination, and keep an editable transcript either way.
Does English narration mean the captions are correct?
No. Speech language, spelling, speaker identification and synchronization are separate checks. Have someone review the final exported version, not just the text in the writing document.