Prompt

BaldieZee

0Likes0ViewsAug 23, 2026

Use the reference person as the only identity benchmark and maintain a high degree of consistency in the five senses, face shape, eyes and overall temperament. An adult Asian woman with a clear and cold face, natural and delicate skin, long black hair was wet by rain, and the wet hair was cluttered on her cheeks and neck. She wore a slim black long-sleeved backless dress with minimalist silver earrings, thin necklace and rings. Urban viaduct scenes on rainy nights, wet roads, metal guardrails, distant high-rise buildings and traffic flow, street lights form a continuous warm yellow bokeh, and the ground reflects red, orange and cold blue lights. The characters are illuminated by warm color street lights and cool urban ambient light. The skin and wet hair have real water light, a strong sense of film, a shallow depth of field, a 35mm lens, slight film grains, high contrast but retaining dark details, real rain, real water droplets, fashionable photography of urban night scenes, emotional restraint, loneliness, ambiguity, and lasting for a long time. The character looked back at the camera, his eyes were quiet and melancholy, and his lips were slightly parted, as if he had taken a last look before parting on a rainy night. The movements are natural, the shoulders and neck are relaxed, the body proportion is true, the hand structure is accurate, and the clap is not stiff. Add advanced magazine-style English typography, use elegant high-contrast serif fonts, with separated word spacing, beige white and dark gold typography, the text is arranged in the negative space around the characters, not blocking the face, the typography is simple, breathable, and not messy. English text: MIDNIGHT MEMORY I STILL REMEMBER A night that never left me. Vertical version 3:4, single complete picture, no puzzles, no watermarks, no platform logo, no account text, and high-quality movie poster texture. Four action changes The character was walking on the viaduct, facing forward, and suddenly looked back at the camera. The character sat by the guardrail, looked up and closed his eyes, feeling the rain falling on his face. The character is sitting in front of a bus or tram, with his hands naturally placed on his knees, and his side face looks into the distance. The character held his face in one hand, his wet hair scattered on the side of his face, and stared at the camera up close. Negative prompt words Low definition, plastic skin, excessive dermabrasion, fake rain, cheap neon effects, exaggerated makeup, stiff movements, wrong fingers, extra arms, body deformation, inconsistent face, anime feel, excessive sharpening, messy background, text garbled, fonts blocking characters, watermarks, Logo, account information, puzzles.

Output details

Media type: Image prompt
Task type: Text To Image
Dimensions: 1024 × 1344
Aspect ratio: 16:21

Model: GPT Image

Continue exploring

All Prompts

Prompt details: Use the reference person as the only identity benchmark and maintain a high degree of consistenc

Reference preview

More like this