Someone films a page of a book and wants to mark one sentence as they read it aloud. Someone screen-records a contract and wants to draw a box around a clause. Both reach for the title tools, because the result they picture looks like a title, and both find the tools do not fit. The reason is worth understanding, because it decides which control you need.

Two different jobs that look the same

A title is text your editor draws. The editor knows exactly where every letter sits, so it can measure the words and fit a highlight to them automatically.

An annotation sits over text your editor cannot read. The words are pixels inside the video, whether that is a photographed page, a shared screen, or a slide. Nothing in the software knows where the sentence begins or ends, so nothing can measure it for you. You place the shape.

A video frame with captions burned in, one word highlighted as it is spoken.

That is the whole distinction, and it explains the frustration: people try to highlight footage with a title style, and the highlight lands on the title's own text, which is empty or somewhere else entirely.

Why a title style cannot do this

Title highlights derive their geometry from the text layout. Feed them no text and there is nothing to measure, so they either draw nothing or draw a shape unrelated to what you can see. This is not a bug to work around. It is the reason a separate annotation tool exists.

Annotations solve the opposite problem. They know nothing about content and everything about position, which is exactly right for marking up something the software cannot read.

Setting up the annotation

In Zella the annotation tools live under the Arrows tab, named that rather than Callouts because the word describes what is inside it.

  1. Open the recording and click Arrows in the right-hand inspector.
  2. Choose Highlight, Underline or Box, matching the same three gestures used for titles: a band for emphasis, a line for a term, a rectangle for something to copy.
  3. The shape lands on the canvas. Drag it over the line of text in your footage.
  4. Resize it to the line using the handles on its edges and corners. Each handle pins the opposite edge, so dragging the right edge leaves the left one where you put it.
  5. Trim its bar on the timeline so it appears when you start reading that line.

The sweep is already on. These three shapes default to a Draw On entrance, so the shape draws across the line instead of appearing whole. If you want it to appear some other way, the Animation dropdown in the shape's edit panel changes the entrance, and a matching Exit control decides how it leaves.

Matching the sweep to what you are saying

The sweep should track your voice. Start it as you begin reading the sentence, and size it so the draw finishes at or just before the end of the phrase.

Two adjustments do most of the work:

  • Move the start, not the length. If the highlight feels late, drag the whole shape earlier on the timeline before you start changing its duration.
  • Let short phrases finish early. A highlight that completes while you are still talking reads as confident. One still crawling after you stop reads as slow.

If you are annotating several lines in sequence, give each its own shape rather than one long one. Sequential shapes let the viewer's eye reset, and they let you re-time a single line later without touching the rest.

Working with a book page or a document

Filmed paper has two problems that a screen recording does not.

The page moves. Even on a tripod, a hand-held book drifts. If the shot has meaningful movement, either lock the camera down or accept that the highlight will sit approximately rather than exactly. A slightly oversized band forgives more drift than a tight one.

The page is not flat to the lens. Shooting at an angle means the text runs slightly diagonally across the frame while your annotation is a level rectangle. The fix is at capture time: shoot square to the page. A shape that is level and text that is not will look wrong no matter how well you time the sweep.

For screen recordings neither applies, and you can size shapes tightly. If the document contains anything private, handle that first, because blurring sensitive information is a separate pass and it is much easier before you have placed a dozen annotations over the top.

Keeping annotations readable

  • One shape at a time on screen. Two highlights in the same frame split attention.
  • Contrast against the page, not against your brand. A yellow marker over cream paper is nearly invisible. Check it against the actual footage rather than a preview background, and if you want a number to aim at, the WCAG contrast guidance is a reasonable floor even though it was written for interfaces rather than video.
  • Do not outline and fill. A box with a heavy border and a tinted interior is two annotations arguing.
  • Cut it when the point lands. Annotations left running into the next paragraph make the viewer re-read the old one.

Getting a shape to actually fit the line is its own skill, and resizing and shaping callouts covers the handles in detail. The same discipline that keeps arrows and callouts legible applies here, and if you also plan to caption the video, decide early where captions sit so a highlight and a caption are never fighting for the lower third.

Frequently asked questions

Why does the title highlight style not work on my footage? It measures the layout of text the editor drew. Text inside your video is pixels to the editor, so there is no layout to measure. Use an annotation shape instead.

Can the software find the text for me? Not for this. Text recognition can locate words in a frame, but a marker that snaps to detected text drifts whenever the page moves or the recognition wobbles, which looks worse than a shape you placed once.

How do I highlight a moving line of text? Either lock the shot so it does not move, or make the shape slightly larger than the line so small drift stays covered. Tracking a highlight to shaky footage rarely holds up.

What size should the highlight be relative to the line? Slightly taller than the letters and a little wider at both ends, the way a real marker overshoots. A band exactly the size of the text looks mechanical.

Can I change how the shape appears? Yes. The Animation dropdown sets the entrance, with Draw On as the default for these three shapes, and a separate Exit control sets how it leaves.

Should I annotate or just zoom in? If the viewer needs to read it, zoom. If the viewer needs to know which part of something they can already read matters, annotate. They solve different problems and stacking both at once is usually too much movement.