Video Captions: Keep the Words Clear of the Action

An illustrator records a tutorial on choosing the right brush in a drawing app. The caption arrives exactly as she names the brush. It is easy to read and spells the name correctly. But it covers the bottom of the menu, just where the brush appears. Someone watching without sound now has to choose between reading the instruction and seeing where to act. The caption itself is right. The combination of words and picture is harder to use than either would be on its own.

That problem calls for a different review from proofreading. Whoever checks the captions needs to watch the picture at the same time, in the version the audience will receive. The question to ask is simple: what does the viewer need to see while reading this line? Sometimes the answer calls for moving the caption. Sometimes it means changing the shot so that the text and the action can stay in predictable places.

Find the moments where the picture carries the instruction

Start with a few points at which the viewer needs to make a choice. In the brush tutorial, they need the name of the tool and its place in the menu. In an interview, they might need to read a speaker’s name while following the introduction. In a product demonstration, they might need to see a hand move a catch while the caption describes the move. Mark these moments in the rough cut. It is easier to leave room for them while editing than to fix each collision on export day.

Sort the visual details by purpose. A caption over a decorative stripe in the background may be harmless. A caption over the one button that makes the procedure work is a problem. If your graphic is a speaker’s name, you could show it a moment earlier and leave it on screen long enough to read. If the viewer needs the instruction and the image together, separating them in time may make the tutorial less useful. Try shifting the action higher in the shot or giving that part of the demonstration a closer view.

Work with the options your published video actually has

Open captions are part of the video picture and stay visible, as the W3C’s guide to open captions explains. That gives the editor direct control over the composition of that export, but the text becomes part of the picture. A separate caption track can keep the words editable and let the viewer turn captions on or off where the player supports it. The May 2026 WebVTT Candidate Recommendation Draft describes cue settings that can place and size a caption box. A specification tells you what can be expressed in the file; see what your actual publishing route does with it before assuming the result will match the editor’s preview.

For example, you might raise one caption so that a menu at the bottom remains visible. Test the uploaded version with captions enabled and check whether the player keeps that cue high enough. Try the smallest screen and the most constrained layout you expect people to use. If text wraps differently there, the caption box may cover another part of the picture. If you publish a vertical edit as well, review it separately: the frame has changed, and space that was empty in the wide version may now hold the action.

Avoid moving each caption simply because a little more space is available somewhere else. Familiar placement can make the clip easier to follow. Reserve a move for a real collision, check that the new position stays readable against the background and return to the usual place when the conflict has passed. If every other line needs to jump, reconsider the layout of the video itself.

Do a viewing check after the final upload

• Mute the clip and turn captions on. Follow the actual instruction. Stop at the moment you have to choose an option and check that you can read the words and locate the relevant control.
• Review any title or label that appears alongside speech. Check that the viewer has time to read both. Where possible, remove redundant graphics or give the essential label its own space.
• Try the dimensions you will actually publish. Check the wide and vertical cuts in their intended player or preview. Look for captions that wrap or compete with other text at the edges.
• Keep the correction with the right version. Update the editable cues and any open-caption export that reused the affected cut. Recheck the corrected moment after delivery.

MaryChuks.com’s guide to keeping the editable caption track with the final cut can help a team find and fix the right version later. The placement review serves an earlier purpose: it gives the editor a way to catch moments when two useful elements are fighting for the same space. For the brush tutorial, the test is whether a viewer can read the brush name and find the brush while the instruction is happening. If they can do both without guessing, the words and picture are working together.

For a recurring series, keep a short note alongside the edit template about where important details tend to appear. A screen tutorial might leave space under a menu, while a craft demonstration might reserve room above the work surface. Revisit that choice when the framing changes. The template should reduce the collisions an editor has to fix, while leaving room for exceptions.

Sources

• W3C Web Accessibility Initiative, Providing open captions.
• W3C, WebVTT: The Web Video Text Tracks Format, Candidate Recommendation Draft (20 May 2026).