How to make your captions move out of the way when a pop-up appears

Last updated 2026-09-15 · 8 min read

What is the short version?

A vertical clip has exactly one comfortable place for captions, and every other element wants that space too. The pop-ups that make your point appear there, the captions sit straight on top of them, and the viewer ends up reading neither the words nor the evidence properly.

1. Tell Vizard Agent what the pop-ups are and why they matter.
2. Ask for captions that move rather than a fixed position.
3. Ask for the panels' own edges checked too.

What do you need before you start?

The clip and a ranking. Vizard Agent needs to know what wins when two things want the same part of the frame — usually the evidence beats the captions, but not always, and stating it once saves several rounds.

• The clip. With its pop-ups in place.
• What must not be covered. Screenshots, panels, banners.
• What may be covered. Background, empty space.
• Where captions normally sit. Their resting position.
• The frame. Vertical changes everything.

What do you type into Vizard Agent?

Name the elements and rank them against each other. "Keep the captions readable" and "keep the pop-ups fully visible" are both perfectly reasonable requests that happen to conflict in a vertical frame, so tell Vizard Agent which of the two gives way when they collide.

Prompt

Variants worth knowing:

• "Have to stay fully visible." Ranks the two.
• "Move them out of the way whenever." A rule, not a position.
• "Leftover edges around them." The other common fault.

What does Vizard Agent actually do?

Here is the Vizard Agent sequence on a real set of vertical clips built around on-screen evidence. The caption work all comes after the panels are settled, because where the captions are allowed to go depends entirely on what the panels have already taken.

1. Reviews every pop-up in the vertical layout.
2. Checks a promo banner's outline against the original.
3. Checks the comment panel for a leftover edge.
4. Removes the layers sitting behind the panels so they read as clean.
5. Finds where subtitles can sit without covering faces or pop-ups.
6. Checks the subtitle-safe space in each selected moment.
7. Adds headlines and word-timed subtitles that move when a pop-up appears.
8. Tunes the size and spacing so the text stays readable without covering the evidence.
9. Gives the captions more room below the speaker's mouth.
10. Verifies the spacing across every clip in the set.

Step four fixes something people see but cannot name. A panel composited over an earlier layer leaves a faint rim of the old graphic around it, and the clip looks slightly cheap without anybody being able to say why.

Step five is the one that has to happen before any text is placed. Subtitle-safe space is not a fixed band in a vertical frame — it is whatever is left after the face and the pop-ups have taken what they need, and it changes through the clip.

Step nine is a small thing that matters. Captions pushed up tight under the chin read as crowded even when nothing overlaps, so the gap below the mouth is set deliberately rather than by default.

What does the result look like?

Captions that behave. They sit in their usual place most of the time, step aside when a panel appears, and come back afterwards — and Vizard Agent shows you the subtitle-safe space it found in each moment rather than asking you to trust it.

When does this not work well?

Some frames have no spare room. If a pop-up fills the width and the speaker fills the rest, the captions have nowhere to go, and the honest choices are a smaller panel, a shorter caption, or no caption over that moment.

• Full-width pop-ups. No space left to move into.
• Constant pop-ups. The captions never settle.
• Very long caption lines. They need width wherever they sit.
• Pop-ups that move. The safe space changes every frame.
• Accessibility requirements. Captions may have to stay put.

How do you fix a result that came back wrong?

Say which moment it is and which of the two things is covered. Vizard Agent measured the safe space moment by moment, so a note about one second of the clip changes that moment's placement only and leaves the caption positions everywhere else exactly as they are.

• "The caption covers the screenshot." That moment's text moves.
• "It jumps around too much." The captions hold one position longer.
• "There is a grey edge round the panel." The layer behind it is removed.
• "It is too close to my chin." The resting gap is increased.

How does Vizard Agent compare to doing it yourself?

By hand you set a caption position, spot two collisions, and keyframe the caption track upward for those moments. Then you change a pop-up's timing and all the keyframes are in the wrong place, so most people give up and shrink the captions instead.

Common questions

Can it move captions up as well as down? Yes. Vizard Agent places them wherever the safe space actually is.

Will they jump about? Not if you say so. Vizard Agent can hold one position across a run of moments.

What if the pop-up fills the frame? Then Vizard Agent tells you rather than putting the caption on top of it.

Can it shrink the pop-up instead? Yes, if you prefer that trade. Say which one gives way.

What is the leftover edge? A rim of an older layer behind the panel. Vizard Agent removes it.

Does this work with word-by-word captions? Yes, and Vizard Agent keeps the timing while moving the position.

Can I keep captions off entirely over the pop-ups? Yes. That is a different rule and Vizard Agent will follow it instead.

How much gap below the mouth? Enough that it does not read as crowded. Vizard Agent sets it deliberately.

Will it work across several clips? Yes, and Vizard Agent keeps the caption treatment consistent across the set.

What about the platform's own buttons? Tell Vizard Agent and it treats those as occupied space too.

Can it check the evidence is legible? Yes. Vizard Agent inspects each panel rather than only the captions.

Does moving captions affect the timing? No. Only position changes; the words land at the same moments.

What if a pop-up moves during the clip? Harder, and Vizard Agent will say when the safe space is changing too fast.

Can I see the safe space it found? Yes. Ask Vizard Agent for the frames it measured.