Prompt Detail
台北暴雨中的湿身浪漫自白 | AI Prompt Gallery
15 seconds, vertical screen 9:16, cloudy day diffused natural light + humid texture of rain wetting everything, raindrops occasionally splashing onto the phone lens causing local blur, under the arcade on a Taipei street, protagonist Taiwanese girl [Image 1], long hair wet from rain sticking to cheeks and neck, white thin shirt slightly sheer and wet. Throughout the process, the girl holds the phone with one hand for a selfie POV, the lens shakes slightly with the wrist. 0-5 seconds: Lens handheld intense shaking, the first few frames of the picture have rain splashing onto the lens causing blurred water droplets; the girl holds the phone with one hand facing herself, the other hand pulling someone running, the lens bumps and shakes with the running; after running into the arcade she leans against a pillar gasping for breath, wet hair sticking to her forehead, casually flicking hair strands which immediately slip back down; Line (panting and laughing): "Hahaha I'm dying from running! I told you it would rain and you didn't believe me! Look, look, it's all wet!" 5-9 seconds: She slowly slides down sitting against the pillar, the phone lowers with her, background is the curtain of rain formed by pouring rain outside the arcade and blurred car light spots, she looks down at the soaked shirt and wrinkles her nose, looks up at the lens showing a helpless yet amused expression; Line (helpless and acting cute): "I look super ugly like this, don't film! ...Eh but do you think this rain is actually quite romantic?" 9-12 seconds: She takes out a pack of crumpled tissues from her pocket, pulls one out to wipe her cheeks, then pulls another one and extends it towards the lens—gesturing with her eyes, corners of her mouth slightly raised; [Rain sound + Street ambient sound, no lines] 12-15 seconds: She slowly brings the phone closer, her face gradually fills the screen, nose tip almost touching the lens, the lens slightly loses focus and then pulls back to clear; Line (suddenly gentle): "Remember to bring an umbrella next time, dummy... but getting caught in the rain with you is actually okay." No text, subtitles, LOGO or watermarks allowed.
Content: The prompt is structured as a timeline with 4 segments (0-5, 5-9, 9-12, 12-15 sec), each containing camera movement, subject action, and lines. Overall: POV selfie perspective, Taipei arcade, rainy day. Style: realistic, romantic. Composition: vertical 9:16, handheld shake, close-up to extreme close-up. Lighting: cloudy diffused natural light + rain wetness. Color: cool tones. Materials: rain, wet hair, wet shirt, blurry droplets. Camera: handheld shake, defocus then refocus.
Pros: 1. Clear timeline segmentation, easy for AI to generate sequentially. 2. Rich details, specific emotions and actions, coherent narrative.
Cons: 1. Prompt length may exceed model context limits. 2. Relies on reference image for protagonist appearance; mismatch without it.
Reference image: Obvious dependency on reference image ([Image 1])
Source materials are collected from publicly accessible web pages. Contact us if a rights issue needs review. Copyright & Privacy