Window Peeper AI Prompt: The Burger and Curtain Scene
Window peeper AI prompt instructions: create the diner, burger, unblinking stare, and closing curtain video using precise timing and framing.
The Seadanse Team7 min readThe team behind Seadanse. We run the models we write about, and we publish the specs, prices and limits the marketing leaves out.
The window peeper scene relies on a simple recipe: one diner sits inside a window eating a burger, notices an outside camera watching them, and slowly pulls a curtain shut.
If you don't want to type custom text commands, you can try the Window Peeper AI workflow to transform an image directly. For reference, an existing CapCut template shows an 8.2-second version where a diner eating near glass spots the lens and closes the curtain.
Copy this Window Peeper AI prompt
These three copy-ready prompts are DIY creative variants for third-party tools with text boxes, not Seadanse's built-in generation settings. Details like a restaurant booth, a large bite, paused chewing, and a dark ending are creative instructions for manual text generation. When you're building the clip in a text tool, clear descriptive beats set up the three core actions: eating, staring, and closing the curtain.
A continuous vertical 9:16 shot filmed from the dark exterior street looking into a warm, brightly lit ground-floor diner window at night. Inside sits the person from the reference photo, keeping their exact face, hair, and clothing, while ignoring the original photo background and pose.
The diner holds an unbranded plain burger in generic paper packaging. First, the diner takes a bite of the burger.
Next, they stop chewing mid-bite. They slowly turn their head to lock eyes directly with the exterior camera outside the glass in a long, uncomfortable, unblinking stare.
Finally, while maintaining deadpan eye contact, they lower the burger to the table. They reach for a dark fabric window curtain and pull it smoothly across the frame until the window is completely covered and the scene ends in darkness.
Natural ambient diner and street audio, no background music.Use this prompt alongside your own reference portrait. The instruction tells the generator to take your subject's face, hair, and clothing while ignoring the original photo background and pose. For a deeper breakdown of setting up your project, read our how-to walkthrough.
Give the stare and curtain their own time
The joke works best when the timing feels deliberate. If the diner looks up too fast or whips the curtain shut immediately, you'll lose the awkward comedy of the moment.
Structuring the clip across ten seconds gives each action room to land. Here is the recommended beat breakdown:
| Time | Action | Framing & Lighting |
|---|---|---|
| 0–2s | Exterior push toward lit window | Cool dark street pushing toward warm interior |
| 2–4s | Diner takes one burger bite | Subject seated inside, holding plain burger |
| 4–6.5s | Eyes lock into camera while chewing | Slow head turn, deadpan stare through glass |
| 6.5–8s | Put burger down, grasp curtain | Eye contact stays unbroken, hand reaches up |
| 8–10s | Pull curtain across until fully closed | Fabric sweeps across frame, ending in black |
If you're using a generation platform that supports timestamped directions, you can paste this timed variation directly into your text input:
A single continuous vertical 9:16 video filmed from outside a window at night.
0-2s: The camera slowly glides forward from the dark, cool street toward a brightly lit restaurant window.
2-4s: Inside, the person matching the uploaded reference image takes a large bite from a plain burger with unbranded paper packaging.
4-6.5s: The diner pauses chewing mid-bite. They slowly turn their eyes directly toward the camera outside and hold an intense, unblinking stare through the glass.
6.5-8s: Without breaking direct eye contact, the diner slowly sets the burger down on the table. They reach up to grab the edge of an opaque curtain.
8-10s: The diner pulls the curtain steadily across the glass until the window is fully obstructed, leaving the frame dark.
Ambient room tone and curtain slide sounds, no music.Giving the eye contact two and a half full seconds builds the awkwardness. It's important to put the burger down before pulling the curtain so the physical movement stays natural.
Keep the camera outside the window
To keep the sense of an unwanted observer out on the sidewalk, frame the shot strictly from outside. Set up a cool, dark street exterior looking into a warm, lit interior, and state clearly that the camera never crosses the glass.
A static vertical 9:16 camera position placed strictly outdoors on a dark, cool sidewalk at night, looking straight through a clear diner window. Inside, the diner matches the identity, facial features, and outfit of the reference photo, framed in warm indoor lighting.
No other patrons are visible. The subject eats a generic burger with no visible logos or text.
Halfway through eating, the diner freezes. They turn their head toward the exterior lens, locking eyes in a deadpan gaze.
Still staring directly into the outdoor camera, the diner places the burger on the table. They grasp the edge of a thick cloth curtain and pull it entirely across the windowpane until the view inside is blocked.
Continuous shot from outside only, never cutting or moving indoors. Natural outdoor atmosphere and quiet room tone.Avoid adding extra elements like branded fast-food wrappers or crowded dining rooms. Keep the scene to one diner and plain props so the prompt stays focused on the core actions.
Make the scene from one photo on Seadanse
The prompts above are designed for DIY tools where you manually control a full text box. If you want the clip without writing complex scene directions, Seadanse gives dedicated steps that handle the sequence for you.
You don't need to write a full prompt or set camera paths. Follow these simple steps:
- Upload one clear photo of a person, cartoon character, or game character, and confirm photo consent.
- Choose your desired output resolution, with options for 480p, 720p, or 1080p.
- Leave the optional opening action collapsed to use the classic burger bite, or expand it if you want to specify a different opening activity before the stare and curtain.
- Use the Window Peeper AI generator by clicking Create.
The system processes the clip as a fixed 10-second vertical 9:16 video. You can access the tool via guest checkout using an existing credit pack, or check pricing to pick up credits before you start. Once processing finishes, your video is ready for direct download.
If you want to create other character clips from a single picture, explore the wider catalog of AI video effects, including the AI Arrest Video template.
Window Peeper prompt questions
Can I use my own clothes?
Yes. The prompt instructs the system to carry over the facial features, hair, and clothing from your input photo, just as the Window Peeper AI effect intends to reflect what you're wearing. The original background and pose are discarded so the subject can sit by the window.
Can I change what the person is doing at the start?
Yes. On Seadanse, an optional opening-action field sits collapsed by default. You can leave it alone to keep the traditional burger bite, or open it to specify another opening activity before the person notices the camera and pulls the curtain.
Is a song required?
A song isn't required for this video. The scene works best with realistic environmental sounds, such as street noise, chewing, and the curtain moving along its rod.
Which person does the photo become?
Your uploaded picture becomes the diner sitting inside the window at night. It's that diner who looks out and reacts, while the camera represents the outside observer watching through the windowpane from the street.
