Image ChatGPT Image 2
Three versions of a woman create surreal scale illusion in plaza
A hyper-realistic fashion ad shows the same woman at three scales: a blurred foreground figure, a giant behind a museum, and a crisp billboard portrait.
Prompt
Use the uploaded portrait only as the identity and hairstyle reference. Preserve the real facial identity, face shape, facial features, skin tone, hairstyle, hair volume, and natural real-life appearance. Do not copy the original portrait’s clothes, background, lighting, pose, or image quality. Create a hyper-realistic surreal fashion advertising photo in a modern architectural plaza. The same person appears in three different scales within one scene. In the foreground, show a normal-size version of the person close to the camera, intentionally out of focus with strong depth of field blur. She wears a clean white short-sleeve shirt and a dark striped tie, with a soft calm expression, slightly posing with one hand near the chin. In the background, a giant enlarged version of the same person is hidden behind a large modern white stone museum building, only the upper face, head, hand, and pointing finger visible. The giant version looks toward the camera and points at a huge billboard on the building wall. On the building facade, place a large clean vertical fashion billboard showing the same person clearly, wearing a white shirt and dark tie, posing elegantly like a school-uniform fashion portrait. The billboard should look like a real outdoor advertisement printed on the building, with realistic perspective, shadows, and material texture. Scene details: bright daytime, pale blue sky, clean modern plaza, white geometric architecture, glass windows, realistic sunlight, soft shadows, slightly cool color grading, cinematic commercial photography, surreal scale illusion, realistic skin texture, natural hair strands, high-end editorial fashion look, sharp architectural details, real camera depth of field, ultra-realistic, no watermark, no logo, no Chinese text, no random text Negative Prompt: watermark, logo, Chinese text, random text, AI-generated mark, Doubao watermark, blurry main face, distorted face, changed identity, different person, fake skin, plastic skin, over-smoothed skin, anime, cartoon, illustration, doll-like face, bad anatomy, deformed hands, extra fingers, missing fingers, broken fingers, unnatural arms, wrong perspective, floating body parts, duplicated messy face, low-resolution, overexposed, dirty background, messy billboard, unreadable text, poster artifacts, unrealistic architecture, warped building, bad composition
Published: May 21, 2026 by Shinning