Original document
GPT Image 2 · en
Scene:
A square overhead product photograph of a rolled parchment neighborhood map partly unfurled on a walnut desk, with a brass house key, blank seal tag, and soft afternoon light. The paper edges are curled, the ink is brown and faded, and the desk casts realistic shadows under the scroll rods. The scene should feel like a premium real-estate closing gift rather than a fantasy poster.
Subject:
Convert a modern suburban neighborhood route into an antique treasure-map artifact. Show a plausible loop road, three cul-de-sacs, a small park pond, and a dotted path leading from a gate-like entrance icon to a hilltop house symbol. Preserve map logic: roads connect cleanly, the dotted path follows accessible streets, and the house destination sits at the end of the route. Use no street labels or readable sales copy.
Important details:
The visual language should prove data-format transformation: modern property-access topology becomes old parchment without losing practical route structure. Include a wind rose, small sketched trees, contour-like ink rings around the hill, subtle fold creases, tea stains, and wax-colored smudges. Materials must be physically credible: parchment translucency at thin edges, ink bleed in fibers, brass key tarnish, and shadows matching the scroll height.
Use case:
Maps to Treasure Map single study 5/8, testing a real-estate neighborhood map converted into a tactile treasure-map closing gift.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no real estate brand, no address labels, no price tag, no impossible disconnected roads, no random island coastline, no route that crosses buildings, no glossy poster paper.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.
