
ChatGPT
PortraitChatGPT
Aesthetic wind motion blur portrait
Creator @YangOnchain
Prompt
Generate a vertical 3:4 aesthetic wind-style portrait photograph, with the theme of "[Theme Name]". The main subject is a female figure wearing a lightweight white long dress or off-white gauze dress, running, spinning, turning back, or being blown by the wind in a natural scene. The character's posture should have obvious dynamics, with long hair and the skirt pulled into flowing trajectories by the wind, creating the effect of slow shutter photography, motion blur, and trailing ghosts. The overall look is not an ordinary clear photo, but an aesthetic dynamic photograph with a sense of dreaminess, film, and oil painting. The image should look like "the wind is passing through the character's body," with the character, skirt, hair, and background all having soft ghosting and a sense of the flow of time. The scene can be [Scene], such as wilderness flower fields, lakeside meadows, misty forests, golden wheat fields, coastal mountain roads, twilight gardens, hillside grasslands, or riverside reed beds. Visual requirements: 1. The character's silhouette is soft but recognizable, do not deform. 2. The white dress should be light, fluttering, with a sense of gauze and layering. 3. Long hair is blown up by the wind, forming a natural scattered feeling. 4. The background has horizontal or rotational motion blur. 5. Low saturation colors, with vintage film grain. 6. Soft light, can be early morning, dusk, cloudy, backlighting, or misty light. 7. The image has a sense of poetry, solitude, freedom, and flow. Photography style: Slow shutter photography, long exposure trails, soft focus, film grain, low saturation colors, natural light, dreamy portraits, motion blur, wind-style photography, vintage art photos, poetic cinematic feel. Do not: No text, no watermarks, no logos, no overly clear commercial portraits, no posed feel, no cartoonish feel, no 3D feel, no cluttered backgrounds, no multiple people, no exaggerated expressions.
Copy the text, add your own details, and try it in your image model.