GPT Image 2ai-art

Fingers Untying Shadows in a Summer Alley

Fingers Untying Shadows in a Summer Alley

An AI-generated image capturing a quiet moment of a young woman adjusting her hair in an old alley where shadows and summer light intersect. Set against a Showa-era wooden shopping street, strong sunlight from the upper right illuminates the figure. A realistic Japanese fashion portrait with low-saturation browns and ivory tones.

Prompt

Even the fingertips touching my hair have been caught by the summer light. Only the shadows of the old alley are a little cool. ☀️🍃📷 #AIart #AI写真 #夏コーデ [GPT Image2 Prompt] Theme: Fingers untying shadows Subject: Vertical 2:3 photo. In a Japanese alley lined with old wooden shops, a young woman stands slightly right of center. The shot captures from the top of her head to near her thighs, with the figure occupying about 55% of the frame. The dark wall in the background and the strong sunlight slanting in from the upper right are prominently shown. Character/Facial Expression: Bright ivory-beige skin with natural blush and delicate texture. Light rose blush, champagne beige and pink brown eye makeup, thin long eyelashes, natural tear troughs, soft bright eyebrows, sheer pink-beige lips. Expression is calm and slightly pensive, face turned slightly left from front, gaze looking toward the left back. Small catchlight in the eyes. Clothing/Pose: Thin white sleeveless blouse with fine pintucks and modest lace, soft cotton material, cool summer wear that naturally shows shoulders and collarbone. Bottom is a high-waisted light flared skirt with pale blue floral pattern. Her right hand is raised near the temple, fingertips gently adjusting hair. Right elbow bent outward, left arm naturally down by the side. Weight slightly shifted to one leg, shoulders relaxed. Background/Lighting: Old wooden shopping street with Showa-era remnants. Faded wooden boards, peeling paper, old signs, blurred unreadable Japanese-style posters. Deep shadows in the foreground and left side of the frame. The face, shoulders, chest, and right arm are hit by hard warm light from the upper right. The light creates a large diagonal triangle, casting dark shadows behind the figure and in the lower left. Subtle brown reflected light from the wooden wall bounces onto the skin. Composition/Camera: Almost eye-level from slightly below chest height. Equivalent to 50mm, distance about 2.5 meters from the subject. The figure is placed slightly right of center, leaving wide dark areas at the top and left. The bright wall on the upper right and the black shadow on the left are divided diagonally. Focus is precisely on the face and right hand, with the background slightly blurred but still distinguishable in wood and paper texture. Texture/Style: Realistic Japanese fashion portrait. Natural skin, individually visible hairs, fine highlights in milk tea ash brown, thin see-through bangs, layers around the face, medium-long hair to below shoulders. Ends loosely curled inward, a few stray hairs catching the light. Low saturation browns, ivory, and pale blue as base tones, with soft film grain and deep blacks. Negative: Unnatural face, unnatural gaze, extra fingers, missing fingers, broken wrists or elbows, fusion of hair and fingers, overly transparent clothing, floating, unnatural gravity, incorrect perspective, shadows contradicting light source, readable text, garbled text, logos, watermarks, excessive skin smoothing, plastic-like skin.

Kynvio Prompt Workflow

How to use this prompt

1

Open it in Image Studio

Send the prompt to Kynvio AI without starting a generation.

2

Adapt the details

Replace the subject, setting, colors, text, or composition for your goal.

3

Review and generate

Choose your final settings, then click Generate when you are ready.

Related GPT Image 2 prompts

About this prompt

Will opening this prompt spend credits?

No. The button only prefills Image Studio. Credits are used only after you start a generation.

Can I edit the prompt first?

Yes. The transferred text is fully editable before you generate an image.