原始文档
GPT Image 2 · en
film photography, candid street snapshot aesthetic, razor-sharp focus on subject, shallow depth of field with soft blurred urban background, bright daylight, slightly overexposed highlights, vivid contrast, subtle analog film texture, heavy film grain, nostalgic cinematic atmosphere, Tokyo backstreet neighborhood scene near a quiet train station, narrow pedestrian street lined with Japanese convenience stores, vintage vending machines, small ramen shops, hanging signs with faded typography, bicycles parked along tiled sidewalks, utility poles and overhead wires stretching across the sky, scattered fallen leaves on the ground, distant pedestrians and passing taxis softly blurred in the background, warm afternoon sunlight with deep blue sky, realistic street fashion photography style, effortless cool street vibe, vibrant colors with slightly muted faded film tones, beautiful 19-year-old Chinese female influencer, fair porcelain skin with cold pale undertones, exquisite natural makeup, glossy soft lips, defined brows, delicate lashes, soft messy long black hair with natural flowing curves, wearing an off-shoulder white fluffy faux fur jacket, playful yet subtly seductive expression, lazy dreamy vintage filter, ultra high-quality details, intentionally mundane phone-camera snapshot feeling, casual accidental composition, imperfect framing, realistic iPhone photography texture, spontaneous candid energy, highly attractive girl casually posing in the middle of the sidewalk, body facing away from the camera while turning her head back toward the lens with direct eye contact, relaxed posture, soft wind moving her hair, emotional youthful atmosphere, modern Asian street fashion editorial, soft haze, layered composition, masterpiece, best quality, ultra detailed, slight motion blur from slow shutter, authentic everyday realism, “BubbleBrain” small handwritten signature text on bottom corner --ar 9:16使用方法
- 复制完整原始提示词，并保留其中的参数。
- 按需求替换主体、文字和构图要求。
- 在兼容的外部工具中使用，核对当前使用条件并检查结果。
开始前需了解
- 先确认是否需要参考图。标注兼容模型，不代表该示例由这个模型生成。
- 核对名称、数字与画面文字；不同模型版本的结果可能不同。