🖼️ Image & Visual (AI Art)

Realistic Mirror-Selfie Scene Creation

📁 Image & Visual (AI Art) 👤 Contributed by @dorukkurtoglu@gmail.com 🗓️ Updated
The prompt
“Mirror-selfie scene in a modern apartment interior. A young woman with long, naturally wavy light-brown hair wears a dark baseball cap (strap visible at the back), a black short-sleeve football jersey with pink lettering and the number ‘10’, and black high-waist leggings. She is turned back to camera, facing the mirror; her face is mostly not visible (rear/three-quarter from behind). She holds a black smartphone in her right hand at shoulder height, elbow bent ~90°, wrist straight, the rear camera facing the mirror so the phone screen in the mirror shows the same scene (phone-within-phone recursion). Left arm relaxed down by her side. Camera & composition: shot as a mirror selfie from behind; lens ~26–28mm equivalent, eye-level camera height; portrait orientation; framed from mid-thigh to above the head, centered on the jersey text and number. Subtle soft daylight from the room; no harsh shadows. Environment: white door on the left with metal lever handle, narrow wall section, open doorway leading to a minimal kitchen (tall fridge/oven column visible), pale neutral walls, light flooring. Background slightly soft but readable. Wardrobe details to preserve: jersey text “MESSI” in rounded pink letters above a large pink “10”; pink line style consistent; jersey fit slightly loose; cap color dark; leggings matte black; hair length reaches mid-back with a few loose curls. Pose & micro-angles: head slightly tilted right, cap brim angled slightly downward; shoulders relaxed; pelvis square to mirror; spine neutral. Right hand grip firm with fingertips visible at the phone’s far edge; phone edges parallel to mirror frame; phone centered roughly over the spine line. Lighting: soft natural indoor light, neutral white balance, no flash, no color cast. Reflections are physically correct with no duplication or misalignment. Aesthetic: clean, realistic, unfiltered, true-to-life skin and fabric textures, sharp focus on jersey lettering and phone; minimal grain. Crop & aspect: primary 3:4 (portrait); safe alternatives 9:16 and 4:5 maintaining full jersey text and number in frame.” Consistency Locks (do not alter): Same woman: hair length, color, wave pattern; dark baseball cap with back strap; black jersey with pink “MESSI” + “10”; black leggings; right-hand phone hold at shoulder height; left arm relaxed; mirror-selfie from behind; apartment/kitchen layout; door with metal lever handle; overall camera height ≈ eye level; phone recursion visible on screen. Keep proportions and body type; no slimming/lengthening. Keep colors and typography of jersey lettering exactly; no restyling. Maintain environment geometry and object placement.

How to use this prompt

Copy the prompt above or click an "Open in" button to launch it directly in your preferred AI. You can then customize the wording to match your exact use case — for example replacing placeholders like [your topic] with real context.

Which AI model works best

These prompts are written for image-generation models (Stable Diffusion, Midjourney, DALL-E 3, Flux) — not chat LLMs. Copy them into your image tool. Midjourney v7 excels at photorealistic portraits; Stable Diffusion 3.5 is the best for fine-tuning and custom checkpoints; DALL-E 3 integrates seamlessly with ChatGPT.

How to customize this prompt

Keep the style descriptors and lighting keywords — these are what make the output consistent. Change the subject, background, and pose freely. Add or remove quality modifiers like "hyper-detailed", "cinematic lighting", "35mm film". For Stable Diffusion, use weight syntax: (keyword:1.3) to emphasize.

Common use cases

  • Generating consistent social-media visuals at scale
  • Creating hero images for blog posts or landing pages
  • Producing concept art and mood boards for clients
  • Generating product photography without a studio
  • Crafting personal avatars and profile pictures

Variations

Adapt the tone (more casual, more technical), change the output format (bullet points vs. paragraphs), or add constraints (word limits, target audience).

Related prompts