Candid Nighttime Park Portrait with Direct Flash

Create a spontaneous, motion-blurred nighttime portrait based on a reference photo, with direct phone flash and the feel of an imperfect candid snapshot

Лина Republished on Nov 20 0 Bookmarks 0 Copies
Bookmark 0

Example outputs

AI-generated image of a woman turns her head beside white stone park railings, with windblown hair and a fountain behind her in a dark autumn park, captured in a grainy, slightly underexposed direct-flash snapshot.
AI-generated image of a woman turns her head beside white stone park railings, with windblown hair and a fountain behind her in a dark autumn park, captured in a grainy, slightly underexposed direct-flash snapshot.

Prompt template

Format
Text
Inputs
Reference_photo (image) · Required
Outputs
Nighttime_portrait (image)
FACE AND APPEARANCE:
Preserve 100% likeness to the woman in the reference image. Do not alter her facial features, eye shape, eyebrows, nose, lips, face shape, jawline, skin color or texture, hair, or distinctive features. Do not change her age or natural body proportions.

A spontaneous nighttime snapshot in a Hype style, as if accidentally captured while she is moving. Vertical 9:16 format, shot on an iPhone with its built-in direct flash. The image should look like a genuine candid snapshot, not a professional fashion shoot.

COMPOSITION:
An eye-level portrait. The woman leans casually against white stone railings in a park. The framing is slightly off-center and imperfect, without precise symmetry. The white railings are prominent across the lower part of the image, while a fountain and dark park form the surrounding setting.

POSE:
The woman tilts her head slightly and casually shifts her weight onto the railings. One hand rests naturally on the railing, and the other hangs relaxed beside her body. At the moment of capture, she turns her head sharply to the side, making the photo feel like a candid moment caught in motion. Her gaze is directed to the side, not straight at the camera.

SHARPNESS 50:
The sharp head movement creates natural, medium-strength horizontal motion blur on the face, especially around the cheek contours, eyelashes, and individual strands of hair. The face remains recognizable and fully consistent with the reference. The hair has more noticeable motion blur from movement and wind. The body, hands, railings, and surrounding background remain much more stable. Do not blur the entire frame.

HAIR:
Long, voluminous hair flows freely in the wind. Preserve its original length, color, thickness, texture, and natural type. Individual strands move around the face and blur softly with the turn of her head. Do not change the hair length or create a different hairstyle.

CLOTHING:
A loose-fitting white oversized T-shirt, visibly wrinkled and slightly messy. Matte cotton, dropped shoulders, and a relaxed silhouette, without large lettering or prints. The T-shirt should look like real everyday clothing, not a styled fashion outfit.

SETTING:
An urban park in autumn, deep at night. A fountain is behind the woman, with dark trees and park paths around it. The background is legible, but most of it falls into darkness. The white stone railings provide a bright contrast against the surrounding night.

NIGHT:
Late night, with near-total darkness around the subject. The sky and distant background are deep and dark, with only a few park details faintly visible. Do not use daylight or natural evening light.

FLASH:
Hard, direct phone flash is the main light source. It sharply picks the woman and white railings out of the darkness. Her skin is lit brightly and naturally, while her hair and clothing have deep shadows and the background quickly falls into darkness. Local bright flash highlights on the white T-shirt and skin are acceptable. No studio lighting pattern.

FINISH:
A photorealistic nighttime photograph with the characteristic look of direct phone flash. Slightly underexposed, with noticeable analog grain, a light film texture, somewhat muted colors, natural skin texture, casual handheld imperfections, and slight optical softness. The image should feel raw, lively, and accidentally captured.

NEGATIVE:
smiling, looking directly at the camera, red dress, black coat, daylight, sunset, studio lighting, perfectly posed stance, perfect symmetry, strong bokeh, blurred background, HDR, beauty filter, plastic skin, heavy retouching, excessive sharpness, full-frame blur, digital motion blur across the entire image, glossy fashion look, CGI, 3D, AI look.

--ar 9:16 --raw --style motion 50

What you'll create

This prompt creates a spontaneous, photorealistic nighttime portrait of a woman leaning on white stone railings in a dark autumn park. Direct phone flash, windblown hair, and motion blur give the vertical image the feel of a grainy, imperfect candid snapshot

Generation settings

Aspect ratio: 9:16
Advanced settings
Negative prompt
smiling, looking directly at the camera, red dress, black coat, daylight, sunset, studio lighting, perfectly posed stance, perfect symmetry, strong bokeh, blurred background, HDR, beauty filter, plastic skin, heavy retouching, excessive sharpness, full-frame blur, digital motion blur across the entire image, glossy fashion look, CGI, 3D, AI look.

Usage instructions

Provide a reference photo of the subject, since the prompt relies on it to preserve her appearance. The motion should read primarily in the head and hair rather than across the whole frame; adjust the stated blur strength if you want a subtler or more pronounced effect

Frequently asked questions

What should the reference photo control?

Use it to preserve the subject’s facial features, appearance, hair, and natural body proportions

How much motion blur should appear?

Keep medium horizontal blur mainly on the face and hair; the body, hands, railings, and background should remain much more stable

What creates the nighttime look?

Direct phone flash lights the woman and white railings, while the park and distant background remain very dark

Comments

Tried this prompt? Share what you made, compare results, or ask a question

Up to 4 results. Images up to 2 MB; MP4, WebM, or MOV videos up to 4 MB each