Create images

Remaining credits 0

1/5

Prompt3987 / 5000

Style Direction: Ultra-photorealistic cinematic lifestyle photography, blending Japanese fresh minimalism with Korean tonal aesthetics. Emphasize a sense of relaxation and emotional softness, with a clean, gentle visual style that combines fashion editorial storytelling with everyday warmth.

Theme and Atmosphere: A quiet, lazy afternoon by the window, conveying calmness, softness, and intimacy. Use light and body language to express a “healing daily life” mood.

Reference Baseline: Strictly use the uploaded subject as the sole identity reference. Accurately reproduce facial structure, proportions, skin tone, hairstyle, and all recognizable features. No stylization or alteration of the real face.

Aspect Ratio and Framing: 4:5 vertical composition. Full-body or half-body framing depending on the reference, prioritizing upper body and leg interaction. Centered alignment with balanced visual weight.

Subject Placement: The subject leans by the window in a relaxed posture, occupying the right two-thirds of the frame. The left side is left open to show a softly blurred outdoor street scene, creating an emotional contrast between interior and exterior.

Interaction Design: Relaxed posture, hands naturally overlapped and resting between the legs. Back gently leaning against the wall, legs slightly bent, conveying a casual, lived-in interaction.

Real Subject (Photorealistic Rendering)
Facial Fidelity: 100% preservation of original facial structure, proportions, skin tone, and hair color. Skin texture must be realistic, with visible pores and natural light transitions, avoiding any artificial smoothing. Eyes appear soft and slightly unfocused, with a faint smile conveying relaxed ease.
Expression: Calm, slightly distant, with a soft, dreamy gaze and naturally relaxed lips, expressing a gentle and healing presence.
Pose: Based on the reference sitting pose — back lightly against the wall, legs slightly bent and overlapped, hands resting naturally on the legs, shoulders fully relaxed with no stiffness.

Wardrobe Reconstruction:
Inner Layer: White spaghetti-strap tank top with ribbed texture, closely fitting the skin with visible material detail.
Outer Layer: Blue-and-white striped shirt, loose fit, sleeves slightly rolled, natural fabric folds, evenly spaced stripes.
Bottom: Likely shorts or a short skirt (frame limited to legs), maintaining a casual homewear aesthetic.
Material Rendering: Skin-contact softness of the tank top, cotton texture of the shirt, and gravity-driven folds must all be clearly visible and realistic.

Lighting and Environment
Lighting Setup: Side-back natural light simulating sunlight from the window on the left. Light grazes the face, shoulders, and clothing, creating soft highlights, while shadows transition naturally (e.g., neck, inner arms), enhancing depth and atmosphere.

Background:
Interior: Light gray matte wall paired with an off-white cushion (soft, voluminous, with natural folds under light).
Exterior: Blurred street scene outside the window, with faint silhouettes of buildings and pedestrians in cooler tones, contrasting with the warm interior light.
Environmental Details: Dark window frame (metal or wood texture), clean windowsill, realistic cushion volume and folds, minimal clutter to maintain a clean, minimalist home aesthetic.

Output Style
Image Quality: 8K resolution with RAW-level detail. Hair strands, skin pores, fabric textures, and light gradients must be rendered with extreme realism.
Color Grading: Low-saturation soft tones dominated by off-white, light gray, and pale blue, with warm yellow highlights simulating afternoon sunlight. Overall palette should feel fresh, calm, and healing.
Post-processing Look: Cinematic grading with preserved shadow detail, soft highlights, and moderate contrast to reinforce the relaxed emotional tone.
Lens Language: 35mm prime lens, f/2.8 aperture, with slight background blur and natural depth transition, keeping the subject as the visual focus.

Source Prompt

Lazy Afternoon Window Calm

The prompt text, model, aspect ratio, and reference images were prefilled from GPT Image Prompts.