Lawful Miniature: Nineteenth-Floor Window — AI Portrait Prompt

Use the attached reference photo as the identity reference: keep the same face, skin tone, hair and features as the person in that photo, without beautifying, aging or altering them. Create a square 1:1 museum-catalog frame with a strong first read and a slower second read at exactly 2880x2880 pixels. Location: the nineteenth-floor window of an unfinished tower, floor swept bare. Show the subject in a slate flannel shirt jacket buttoned once over a charcoal tee, measuring the window's sill height with a pocket tape. Treat this mechanism as the frame’s governing rule: a wooden view camera whose tilted lens lays a sloping wedge of focus that renders the scene beyond as a tabletop model achieves its entire effect without altering a single object it shows. Explore why the eye trusts context more than it trusts the thing itself; direct expression and posture toward composed surprise, never melodrama or ad-like confidence. Frame from one marked standing position so the alignment lands exactly and their face stays readable. Make the composition align their gaze with the exact line the illusion depends on. Render it as conceptual scale-study realism combining architectural miniature exactness, truthful shadow geometry, unretouched photographic optics, and quiet human-centered observational framing, using one collimated key with crisp parallel shadows and soft neutral fill and a palette of dust beige, asphalt grey, and a single brick red. Photorealistic: this must read as a real photograph from a physical camera, never an illustration, painting, or 3D render. Resolve real surface detail: paper fibre visible at cut edges, plastic that shows mould lines, and paint thicker where a brush turned. Keep one identity-locked person, credible hands and fingers, natural fabric, and coherent scale, contact, reflections, shadows, airflow, deformation, and materials. Background people only when essential, always secondary and clearly different. Avoid surreal scale collage, floating figures, painted anamorphic pavement, and dollhouse cartoon styling. No text, labels, logos, signatures, watermarks, random symbols, plastic skin, excessive bloom, or glossy AI finish.