What the double exposure look is
Double exposure started on film. The photographer opens the shutter twice on the same frame, so two images land on top of each other in one picture. Because the frame takes light twice, each exposure usually gets about one stop less light so the result does not blow out. The useful part for us is where each image shows up: wherever the first shot left the film dark, the second shot fills in almost completely. That is why the classic look is a subject against a plain background, with a whole landscape appearing inside the figure.
On screen it reads as memory, identity, or a person carrying a place inside them, which is why it shows up in title sequences, album art, book covers, and posters. It is one of the few styles where the meaning comes built in. Put a forest inside someone's profile and the viewer fills in the story without a word of narration.
It continues the club's style series alongside [AI film noir videos](/blog/how-to-make-ai-film-noir-videos), [AI vintage film videos](/blog/how-to-make-ai-vintage-film-videos), and [AI watercolor videos](/blog/how-to-make-ai-watercolor-videos). It gets confused most with an ordinary crossfade and with a green screen composite, so here is the difference.
Double exposure vs crossfade vs green screen
| Double exposure | Crossfade | Green screen composite | |
|---|---|---|---|
| What you see | Two images at once, one living inside the shape of the other | One shot fading out as the next fades in | A subject cut out and placed in front of a new background |
| How long it lasts | The whole shot | A second or two between shots | The whole shot |
| Where the second image goes | Inside the subject | Over the whole frame | Behind the subject |
| What it says | Memory, identity, a person and a place as one | Time passing, one scene ending | The subject is somewhere else |
Two ways to make AI double exposure videos
There are two routes, and the choice matters more than any prompt line.
- One prompt: describe the whole effect in a single generation, for example a double exposure of a woman's face in profile, her silhouette filled with a misty pine forest, plain off-white background, soft muted colors. It is fast and good for testing ideas. The weakness is control: the model decides where the forest shows through, the face tends to soften or shift, and you cannot change the inner world without regenerating everything.
- Two layers: generate the silhouette and the inner scene as separate clips, then blend them in your editor. This is the route I use for anything a client is paying for. The face stays exactly as you made it, you can try three inner worlds on the same silhouette, and you decide exactly where the second image appears.
For the silhouette layer, prompt a person or object in clear profile against a plain light background, with strong contrast and very little movement: a slow blink, hair moving slightly, a gentle turn of the head. For the inner layer, prompt a scene with clear shapes and slow motion: trees swaying, clouds drifting, waves rolling in, city lights coming on. The prompting basics are in [how to write Higgsfield video prompts that get results](/blog/how-to-write-higgsfield-video-prompts-that-get-results), and camera language is in [how to control camera motion in AI video](/blog/how-to-control-camera-motion-in-ai-video).
How to blend the two layers in your editor
- Put the silhouette on the bottom track and the inner scene above it.
- Set the inner scene's blend mode to screen or lighten. On a light background, that makes the scene show most strongly where the silhouette is dark, the same way the second exposure fills the dark parts of a film frame.
- Mask the inner scene to the shape of the silhouette if it is bleeding into the background, and feather the mask edge so it fades rather than cuts.
- Pull the inner scene's opacity down until the face still reads. A forest that hides the eyes and mouth completely loses the person, and the person is the point.
- Grade both layers to one palette, the method in [how to color match AI video clips](/blog/how-to-color-match-ai-video-clips). Muted, slightly desaturated tones are the safe default.
The signature move is a still silhouette with a moving world inside it. Keep the outer layer almost frozen and let the inner scene drift: clouds moving through a face, a city coming alive inside a profile at dusk. That contrast between stillness and motion is what makes the effect look intentional rather than accidental.
The Edge Check: the club's double exposure quality test
Most broken double exposures fail at the edges, not in the middle. Run every clip through four checks before it goes in the edit.
- Edge: pause at the start, middle, and end. Does the outline of the silhouette hold steady, or does it wobble, shimmer, or grow a halo? Shimmering edges are covered in [how to fix flickering in AI video](/blog/how-to-fix-flickering-in-ai-video).
- Inside: does the inner scene stay inside the figure, or is it leaking across the background in a way you did not plan?
- Face: do the eye, nose, and lips stay the same person the whole way through? A face that slowly changes is the fastest way to make the piece look cheap. For a recurring person across several pieces, see [Higgsfield character consistency](/blog/higgsfield-character-consistency).
- Clean: is the background still plain? Stray textures, text, or a second faint face appearing in the background mean the blend or the generation went wrong. Trim or regenerate.
Short clips hold the effect better than one long take. Build longer pieces from several short ones, the approach in [how to make AI videos longer than the clip limit](/blog/how-to-make-ai-videos-longer-than-the-clip-limit).
How to make money with AI double exposure videos
- Musicians: the artist's profile with the world of the song inside it is a natural visualizer or lyric video background, and a short loop of it fits a streaming canvas. See [how to make an AI music video](/blog/how-to-make-an-ai-music-video) and [how to make Spotify Canvas videos with AI](/blog/how-to-make-spotify-canvas-videos-with-ai).
- Authors: a book trailer with the main character's silhouette filled with the story's setting says what the book is about in five seconds.
- Travel and outdoor brands: a traveler's profile filled with the destination, the style-forward version of [how to make AI travel videos](/blog/how-to-make-ai-travel-videos).
- Personal brands, coaches, and podcasts: an intro with the host's own photo and their world inside it. Use the client's real photo for their face, never a generated stand-in.
- Gentle tributes: a family photo with a meaningful place inside it, handled with the care described in [how to make AI memorial tribute videos](/blog/how-to-make-ai-memorial-tribute-videos).
- Presets and prompt packs: once your silhouette prompts, inner-scene prompts, and blend settings hold the look, package them for other creators, the path in [sell AI video templates and presets](/blog/sell-ai-video-templates-and-presets). Price client work with [how to price AI video services for clients](/blog/how-to-price-ai-video-services-for-clients).
Make one this week: your own profile with the place you grew up inside it, built as two layers and blended in your editor. Post it. It shows a client you can control a style instead of hoping a prompt lands, the approach in [how to build an AI video portfolio with no clients](/blog/how-to-build-an-ai-video-portfolio-with-no-clients).
Inside the Higgsfield Income Club at higgsfieldincomeclub.com, $9/month, creators share the style prompts, edit templates, and client pitches behind every look in the style series, double exposure included. See the [pricing page](/pricing) for what is included. Two layers, one still silhouette, and check every edge.
Short, practical drops on income paths, prompts that work, packaging for views, and getting paid. No spam, unsubscribe anytime.
Frequently asked questions
How do I make an AI double exposure video?
Make two layers: a silhouette (often a face in profile) on a plain light background, and a second scene like a forest, city, or ocean. Animate each with small, slow motion, then put the scene on top of the silhouette in your editor with a screen or lighten blend mode, mask it to the figure, and lower its opacity until the face still reads.
Can I make a double exposure video with one AI prompt?
Yes. Describe the effect directly, such as a face in profile filled with a misty forest on a plain background. It is fast for testing, but you get less control over where the second image shows and the face can shift, so for client work building two layers and blending them is more reliable.
What blend mode makes a double exposure?
Screen or lighten is the usual starting point when the silhouette is dark on a light background. The second scene shows most strongly inside the dark figure, which mimics how a second exposure fills the dark parts of a film frame.
What is the Edge Check?
It is the club's quality test for double exposure clips: the silhouette edge stays steady, the inner scene stays inside the figure, the face does not morph, and the background stays clean. A clip that fails any of the four gets trimmed or regenerated.
What is the difference between a double exposure and a crossfade?
A crossfade is a quick transition where one shot fades into the next. A double exposure holds two images together for the whole shot, with one living inside the shape of the other.
Who pays for AI double exposure videos?
Musicians, authors, travel and outdoor brands, personal brands, coaches, and podcasters, families who want a gentle tribute, and other creators who buy prompt packs and presets.
Last reviewed by David on October 30, 2026
