# GPT Image Hub prompt

## 찢어진 종이 이중 노출 패션 포트레이트
- URL: https://www.gptimagehub.com/ko/library/torn-paper-double-exposure-fashion-portrait-2
- Generator URL: https://www.gptimagehub.com/ko/generate?promptId=oss-ym-g2-25843
- Markdown URL: https://www.gptimagehub.com/api/agent/markdown?path=%2Fko%2Flibrary%2Ftorn-paper-double-exposure-fashion-portrait-2
- Category: 인물 및 사진
- Model: nanobananapro
- Aspect ratio: 9:16
- Style: photorealistic
- Tags: 시네마틱, 이중노출, 콜라주, 패션인물사진, 페이퍼티어, 남아시아
- Updated: 2026-10-27T01:51:43.098Z
- Created: 2026-10-27T01:51:43.098Z

비에 젖은 거리를 걷는 전신 모습과 몽환적인 대형 클로즈업을 찢어진 종이로 분할한 시네마틱 콜라주입니다. 인물, 전통 의상, 배경 도시를 커스텀해 보세요.

### Full prompt
```text
Vertical cinematic double-exposure paper-tear collage portrait of [a beautiful young South Asian woman].

Composition: One full-body figure walking forward in the lower center, with a giant ethereal close-up portrait of the same woman filling the upper two-thirds softly blended into the background. A torn white paper edge runs diagonally from lower left to mid-right, visually separating the giant portrait above from the street scene below, featuring realistic fibrous paper texture and drop shadow.

Subject & Attire: The walking figure wears [an elegant teal embroidered South Asian salwar kameez with gold embellishments, matching dupatta, wide trousers, gold sandals, necklace, and ornate gold jhumka earrings]. The upper close-up features silky dark hair parted to one side, warm medium-brown skin tone, delicate jewelry, and a serene fashion-editorial expression.

Setting: [An atmospheric wet urban boulevard with blurred streetlights and warm golden bokeh] at dusk, with reflective pavement, misty rain haze, amber highlights, and a soft overcast sky.

Style & Camera: Cinematic photorealism, fashion magazine editorial aesthetic, 85mm lens look, shallow depth of field, creamy bokeh, romantic moody lighting, intricate fabric embroidery, and subtle film grain. Dramatic scale contrast between the intimate close-up and the full-body subject.

Constraints: No text, logos, or watermarks. Exactly one full-body figure and one close-up portrait, with no additional people.
```