A Checklist for Captions and On-Screen Text Before You Commit
Before you add a single caption to a highlights edit, run through these questions on accuracy, readability, timing and small screens so the words support the footage instead of fighting it.
Daniel Ortiz · · 6 minutes to read
Start with the audio you actually have
Captions begin with listening, not typing. Before deciding on a style, play the clips you plan to use with headphones and note where speech is clear, where it is muffled by room noise, and where a speaker trails off. If a sentence is hard for you to hear, it will be hard to caption honestly, and you may be tempted to guess. A guess that changes a speaker's meaning is worse than leaving a clip out.
I ask learners to mark every spoken line they intend to caption and write a rough transcript first. This takes longer than you expect, but it pays back later, because you can see which lines carry the point of the moment and which are filler. A transcript also shows you where one speaker interrupts another, which affects how you split text on screen.
Only once the transcript exists should you think about fonts, colors or placement. The words come first, and the design serves them.
- Listen to each clip once without looking at the picture
- Write a rough transcript of the lines you plan to use
- Flag any phrase you cannot hear with confidence
Check that the words match the speech
The first question on the checklist is simple: does the caption say what the person said? Light cleanup is reasonable. Removing a stray repeated word or a false start can make a line easier to read. Rewriting a sentence so it sounds smoother, or merging two separate remarks into one, is not cleanup. It changes the speaker's comment, and viewers who attended the event will notice.
When a trimmed clip begins mid-thought, read the caption and ask whether someone seeing only that fragment would understand the claim being made. If the answer is no, extend the clip or choose another line. Captions cannot repair a missing context, and adding explanatory text on screen puts your words in the speaker's mouth.
Keep names, titles and numbers exactly as given in your source notes. A misspelled name on a recap video is the kind of error people remember longer than the highlights themselves.
- Keep wording faithful, trimming only false starts and repeats
- Confirm names, titles and figures against your source notes
- Extend a clip rather than explain missing context with extra text
Readability on a small screen
Most people will watch your highlights on a phone, often with one hand and sometimes in a bright room. Test your captions at that size before committing to them. Hold your phone at arm's length, play ten seconds and ask whether you can read every line without squinting. If you cannot, the text is too small, too thin or too close to a busy part of the picture.
Aim for short lines, roughly two lines at most on screen at once, with a plain, sturdy typeface. Strong contrast matters more than style: light text on a soft dark band usually survives changing backgrounds better than colored text placed directly over the footage. Leave margins so nothing sits under the interface elements that many apps overlay near the bottom of vertical video.
If your video will be exported in horizontal, square and vertical versions, check the caption position in each one. A line that sits comfortably in the wide frame may be cut off or crowded in the tall one.
- Read a ten-second sample on an actual phone
- Keep to two short lines at most
- Check text placement again in every export format
Timing, rhythm and reading speed
A caption that appears late feels like a mistake, and one that lingers after the speaker has moved on feels sluggish. Line up each caption so it appears just as the words begin and leaves shortly after they end. Give viewers enough time to read: if you need to read a line twice to finish it, the viewer certainly does too.
Reading speed varies, so build in a little margin rather than cutting it fine. When a speaker talks quickly, split the line into two shorter captions instead of squeezing a long sentence into a short window. This also gives the edit a steadier rhythm, because the text changes in step with natural pauses.
Watch the finished clip once with the sound off. If the point still comes across and nothing feels rushed, your timing is in good shape. Then watch it with sound, looking for any moment where the text and voice drift apart.
- Appear with the first word, clear soon after the last
- Split fast speech into shorter consecutive captions
- Review once muted and once with sound
Titles and on-screen text beyond captions
Captions are not the only words in a recap. Opening titles, speaker labels and closing text each need a reason to be there. Ask whether a viewer would miss the information if you removed it. A speaker label helps when a stranger appears on screen; a decorative title that repeats what the footage already shows only takes space from the picture.
Keep these elements consistent. Use the same typeface, the same placement and the same style for similar items throughout the video, so the viewer learns the pattern after the first one. Inconsistent text makes an edit look assembled in a hurry, even when the clips themselves are good.
Be honest in any headline or summary line. If the video shows a short selection from a longer day, say so plainly rather than implying it covers everything. Plain wording builds trust with the audience and with the people who gave permission to be filmed.
- Use speaker labels only when the viewer needs them
- Keep typeface, size and position consistent throughout
- Describe the video as the selection it is
A final pass before you decide
Before you call the text finished, take a last pass with fresh eyes, ideally after a break. Read every caption aloud and compare it against the transcript. Look for typos, inconsistent capitalization and punctuation that wanders from line to line. Small slips are easy to miss when you have watched the same clip thirty times.
Ask someone who did not work on the edit to watch it on their own phone and tell you what they understood. Their questions are more useful than any rule. If they stumble over a line, rewrite or retime it. If they never noticed the captions at all because the text felt natural, you have probably done the job well.
Then decide: keep the captions as they are, adjust a few lines, or return to the transcript for another round. In our course, this is the kind of detail we review together in feedback rounds.
Good captions are quiet. They match the spoken words, stay readable on a small screen, arrive and leave at the right moment, and never change what a speaker meant. Work through the checklist on accuracy, legibility, timing and consistency before you commit, and you will catch most problems while they are still easy to fix. When in doubt, favor the plainer option and the clearer wording, and let the footage carry the moment.