# How to Put Text Behind a Person in a Video on Mac

Put text and other visual items behind a person with manual occlusion in ZXN Studio. Follow the layer setup and refine the depth boundary.

Canonical: https://zxn.studio/blog/how-to-put-text-behind-person-video-mac

Putting text behind a person in a video makes a flat shot feel layered. The title stays visible across the background while a head, shoulder or hand covers the overlapping letters. You can use the same idea for a graphic or another visual item, provided it lives on its own layer and has a suitable video underneath it.

**In ZXN Studio on Mac, add a title above your footage, select the title, then open AI tools > Occlusion. Choose the footage as Background video and click Enable AI occlusion.** This manual workflow gives you control over the item, its timing and the depth boundary. You do not need to apply Creator Text Wall first.

The screenshots below show a real editable project using [presenter footage by Engin_Akyurt on Pixabay](https://pixabay.com/videos/man-phone-camera-selfie-61298/). We added a Basic title, selected the original video as its depth source and enabled occlusion manually. The guide also explains how to carry that layer arrangement over to other items and how a comparable composition works in Final Cut Pro.

## What occlusion does to your layers

Normally, an item on an upper video track covers the layers beneath it. Occlusion selectively hides parts of that upper item, allowing the underlying footage to show through. The result looks as though the foreground person is standing in front of your lettering, even though the title remains on the upper track.

The manual controls use estimated depth from the source video. This is useful for people, but also for other foreground objects that are sufficiently separated from their surroundings. It is not a command to select every human perfectly. A depth boundary can include a nearby chair or miss a difficult piece of hair. Inspect the shot and adjust the result rather than assuming the subject’s identity determines the mask.

Occlusion and tracking solve different problems. Occlusion decides which portions of an item are hidden. Tracking controls how that item moves with the footage. A stationary headline behind a presenter can use occlusion without a tracked attachment. A label that follows a moving object may need both.

## 1. Put the original video on a lower track

Open ZXN Studio and switch to **Advanced**. Import the original video through the Media browser, then add it to the timeline. Start with a short section of the shot you intend to use. Keeping the first experiment brief makes it easier to inspect every head turn and hand movement.

Use footage with some clear distance between the foreground and background. A reasonably sharp presenter against a more distant wall is a useful starting point. Fast motion, transparent objects and fine hair can make the boundary harder to judge. Leave enough empty space around the subject for the text to remain readable.

The source must be a video clip on a lower timeline track, and it must overlap the upper item in time. A file sitting only in the Media browser is not enough. Neither is an unrelated video elsewhere on the timeline. Save a project copy before experimenting with an existing edit.

## 2. Create an ordinary title above the footage

In the Media browser, open the **+** menu and choose **Title**. This adds an editable Basic title. Keep its track above the source video and align its duration with the part of the shot where you want the words to appear.

Select the title and use **Text > Content** in the Inspector to replace YOUR TITLE. Choose a short phrase, a sturdy typeface and a size that allows some lettering to extend beyond the person’s silhouette. Our example uses BEHIND THE SCENE. Before occlusion is enabled, the entire line crosses the presenter’s face. That is a useful starting check: it confirms the title exists, overlaps the footage and is actually above it.

![Basic title crossing the presenter’s face, with the lower video track and Enable AI occlusion control visible in ZXN Studio](/media/articles/text-behind-person/manual-before-occlusion.png)

*The manual starting arrangement: a Basic title on V2 and the original video on V1. No Creator Text Wall template is involved.*

Avoid hiding most of an essential word. Decorative overlap works well for an introduction, but a keyboard shortcut, price or instruction should also appear somewhere fully readable. Check the phrase at the size of a phone player instead of judging only from the large desktop preview.

## 3. Choose the source and enable AI occlusion

With the **title** selected, open **AI tools** in the Inspector and find **Occlusion**. The controls belong to the item you want to place behind the foreground. Selecting the original footage instead would configure a different item.

Choose your footage in **Background video**, then click **Enable AI occlusion**. Afterwards, the source selector is labelled **Depth source**. In our example, it points to the same presenter clip visible on the lower track. Occlusion is marked as a Studio feature in the interface.

The manual feature uses depth-anything-v2-small locally. The interface explains that the first use downloads about 50 MB. Allow the depth result to appear before evaluating it. The screenshot below shows the actual result: we shortened the title to CREATE and moved it left. Only the edge of the final letter overlaps the presenter, keeping the word easier to read while retaining the depth cue.

![Manually enabled occlusion hiding the title behind the presenter, with Depth source, Foreground cutoff and Edge softness visible](/media/articles/text-behind-person/manual-occlusion-controls.png)

*The title has been shortened and repositioned after enabling occlusion. Its lower video source, Foreground cutoff of 0.48 and Edge softness of 0.08 are visible. Those values are an example, not universal settings.*

## 4. Refine the boundary and check several frames

Adjust **Foreground cutoff** to change the depth boundary that conceals the title. Make a small change, inspect the overlap and compare another frame before continuing. The right value depends on the shot’s depth estimate and where you want the lettering to sit.

**Edge softness** changes the transition along that boundary. A little softness can help an edge blend, but too much can make letters appear to fade through the person. Compare a shoulder, the side of the head and any moving hand. A setting that improves one area can weaken another.

**Reverse depth direction** switches which side of the depth boundary hides the item. For a typical title behind a nearby presenter, start with the normal direction. Reverse it only when the result calls for the opposite relationship, and verify the change visually.

Play through the entire title section. Watch for text leaking across a face, abrupt changes around the outline and a phrase that becomes unreadable as the subject moves. Moving the title to a clearer part of the frame can be more effective than repeatedly adjusting the mask.

## Apply the same arrangement to other visual items

The useful pattern is an independently selectable upper item and a lower video source. The Inspector exposes Occlusion for supported visual clips, including images and motion graphics. Select that visual item, choose the overlapping video below it and refine its own depth boundary. Audio is not an occlusion layer.

Keep an effect and its intended source separate when the effect should sit behind something in the footage. Applying a visual effect directly to the source video does not automatically create an independent layer to mask. It changes that video’s appearance. The item being hidden needs its own place in the timeline so you can select it and assign the lower footage as its source.

## Example: put Matrix Rain behind the presenter

You can use the same manual controls for a separate motion graphic. Our Matrix Rain example draws falling green characters over a dark, opaque background. It creates a digital backdrop behind the retained foreground; it is not a transparent sprinkle of characters over the original room.

1. Download the [Matrix Rain source file](/media/articles/text-behind-person/matrix-rain.tsx). Open it in a text editor and copy its complete contents. This sample is prepared for the new graphic created in the next step, so keep its film identifier unchanged.
2. Move the playhead to the start of the intended section. Open **Graphics** in the browser and click **Create motion graphic**. In the code editor that opens, replace the starter source with the downloaded file’s contents. Wait for validation, then click **Apply changes**. You now have an editable graphic on a separate upper track.
3. Rename its clip **Matrix rain** by clicking the clip name in the Inspector. Adjust the clip’s timeline duration to cover your chosen section. Our test used five seconds. Hide the earlier title layer if you want to inspect the rain on its own.
4. Fit the graphic to the video frame using **Transform > Scale**. The supplied graphic is 1920 × 1080. In our 3840 × 2160 project, 200% fills the frame; in a 1920 × 1080 project, 100% is sufficient. Keep it centred unless you want a deliberately offset backdrop.
5. Select the **Matrix rain clip**, open **AI tools > Occlusion**, choose the original presenter footage as **Background video**, and click **Enable AI occlusion**. Adjust **Foreground cutoff** and **Edge softness** while checking the head, shoulders, hands and nearby objects.

![Editable Matrix rain graphic behind the presenter, hand and phone, with its separate timeline layer and manual occlusion settings visible](/media/articles/text-behind-person/matrix-rain-occlusion.png)

*The real project at four seconds. Matrix rain is on V3, the earlier title is hidden on V2, and the source video remains on V1. This shot uses Foreground cutoff 0.30 and Edge softness 0.04.*

The lower cutoff in this example keeps more of the shoulders in front of the graphic. It also illustrates why manual depth can include other nearby objects: the phone and hand remain in front too. These values were chosen for this footage. Scrub several positions and play the sequence before using them for a different shot.

## Creator Text Wall is an optional shortcut

If you want a prepared typography layout, **Creator Text Wall** in the Effects browser can create several editable title tracks for a selected video. That preset uses person segmentation and shows **Person cutoff**, whereas the manual depth workflow above shows **Foreground cutoff**. The two controls should not be treated as interchangeable descriptions of the same mask.

Use the preset when its layout suits the shot. Use the manual method when you want to choose your own title or configure another supported visual item. You can learn the layer relationship with a single plain title before adding a more elaborate design.

## Check the export and compare Final Cut Pro

Export a short sample and watch the resulting file outside the editor. Check the whole movement, not only the frame used in a screenshot. The overlap should establish depth while leaving the message understandable on a first viewing. For related compositions, see our [camera pop-out guide](/blog/camera-pop-out-effect) and the separate [background removal guide](/docs/background-removal).

In Final Cut Pro, a comparable arrangement uses the original shot below a title and an aligned duplicate above it. Apply **Magnetic Mask** to the upper copy, isolate the person, refine the selection and analyze the clip. That masked foreground covers the title while the lower copy keeps the background. Keep the video copies synchronized and avoid duplicating their audio. Apple’s [Magnetic Mask guide](https://support.apple.com/guide/final-cut-pro/ver1d67e3a53/mac) explains the selection and analysis controls.

If ZXN Studio shows no eligible background video, check track order and timing first. If every letter stays in front, check the upper item’s occlusion switch and source. If the lettering disappears too extensively, inspect the cutoff, direction and placement. Begin with a clearly readable composition, then add only enough overlap to make the depth convincing.

## Sources and further reading

- [Apple: Add Magnetic Masks in Final Cut Pro](https://support.apple.com/guide/final-cut-pro/ver1d67e3a53/mac)
- [Example footage: Engin_Akyurt on Pixabay](https://pixabay.com/videos/man-phone-camera-selfie-61298/)