Back to Cinematic photo prompts

Mirror Selfie Couple AI Photo Prompt

Use the uploaded reference photographs as the sole and highest-priority identity references for both people. Recreate the exact same man and woman shown in the uploaded images throughout all three panels. IDENTITY CONSISTENCY IS THE ABSOLUTE HIGHEST PRIORITY. The man and woman must remain the same recognizable individuals in every panel, with no face changes between shots. FACE & IDENTITY — STRICT Preserve both subjects' exact facial identities, facial structure, proportions, natural asymmetry, face shape, forehead, eyebrows, eye shape and spacing, eyelids, nose, lips, cheeks, jawline, skin tone, age appearance, hairline, and distinctive facial characteristics from the uploaded references. Do not generate generic versions of them. Do not beautify, redesign, reshape, slim, enlarge, smooth, de-age, age, or reinterpret their faces. Maintain authentic skin pores, fine texture, natural complexion, realistic facial anatomy, and subtle imperfections. The woman must have the exact same face in Panels 1, 2, and 3. The man must have the exact same face in Panels 1, 2, and 3. Create a vertical three-panel photographic triptych, with three distinct photographs stacked vertically in a single 4:5 composition. OVERALL STYLE Ultra-realistic cinematic lifestyle photography of the same young South Asian couple enjoying a romantic date at a cozy, elegant restaurant. The restaurant has a dimly lit romantic atmosphere with a textured cream-colored stone wall, warm ambient lighting, subtle shadows, and sophisticated intimate décor. All three panels should feel as though they were photographed during the same evening at the same restaurant, with consistent clothing, hairstyles, accessories, skin tones, and overall visual identity. Use realistic DSLR photography, natural expressions, authentic skin texture, cinematic warm lighting, shallow depth of field, subtle photographic grain, and premium editorial lifestyle photography. --- TOP PANEL — MIRROR SELFIE The woman stands with the man beside her, taking a happy mirror selfie using her phone. The woman wears a beautiful red floral dress and has her long dark hair naturally styled. The man stands beside her wearing a black shirt and glasses, smiling naturally toward the mirror/camera. The woman holds the phone naturally at a believable selfie angle. Capture a genuine happy couple moment rather than a stiff posed photograph. Show realistic mirror reflections, accurate phone proportions, natural hand anatomy, realistic clothing folds, and subtle restaurant reflections. --- MIDDLE PANEL — ROMANTIC CLOSE-UP A candid intimate close-up photograph of the same exact couple. They are sitting or standing very close together, looking deeply into each other's eyes with warm, genuine smiles. Their expressions should communicate natural affection and comfort without appearing exaggerated or staged. Frame both faces prominently while preserving their exact facial identities. Use soft warm restaurant lighting falling naturally across their faces, with realistic highlights in the eyes and gentle shadows across the facial contours. Use shallow depth of field with the textured cream stone wall softly blurred behind them. --- BOTTOM PANEL — JOYFUL SELFIE The same exact couple takes another joyful selfie together. They are laughing naturally and leaning slightly toward each other. Their expressions should feel spontaneous and candid, capturing a genuine happy moment from the same date. Maintain the exact same facial identities, hairstyles, clothing, accessories, and overall appearance established in the first two panels. --- TABLE & PROPS Include an elegant restaurant table naturally visible within the composition. On the table: Elegant glasses containing red cocktails A beautiful bouquet of pink polka-dot flowers Subtle tableware and refined restaurant details Keep all props physically realistic and naturally positioned. LIGHTING & CAMERA Warm cinematic restaurant lighting with practical ambient lights and subtle highlights. Professional full-frame DSLR / mirrorless photography. Vertical 4:5 composition Three clearly separated photographic panels Realistic 50mm / 85mm portrait photography aesthetic Natural perspective Shallow depth of field Creamy realistic bokeh Authentic skin pores and fine facial texture Realistic hair strands Physically accurate reflections Natural shadows Subtle film grain Cinematic warm color grading High dynamic range Photorealistic low-light detail Premium editorial lifestyle photography Ultra-high-resolution realism STRICT NEGATIVE REQUIREMENTS No identity drift, no different faces between panels, no generic faces, no face replacement, no facial reshaping, no beautification, no plastic skin, no excessive skin smoothing, no artificial beauty filter, no altered nose, eyes, lips, jawline, or face shape, no de-aging, no aging, no distorted hands, no extra fingers, no duplicated people, no malformed anatomy, no unnatural reflections, no CGI, no 3D-rendered appearance, no illustration, no cartoon effect, no artificial-looking expressions. FINAL REQUIREMENT: The uploaded photographs are the definitive identity references. The final triptych must look like three genuine photographs of those exact same two people, photographed during the same romantic restaurant date, with perfect identity continuity across all three panels.

How to use this prompt

  1. Choose Copy prompt to copy the instructions.
  2. Open a fresh ChatGPT conversation.
  3. Upload the photo you want to transform.
  4. Paste the copied prompt and send it with your photo.
  5. If the result misses the mark, try another generation or adjust one phrase.

Tips for a stronger result

  • Leave room for portrait framing. For a 9:16 image, keep some space around the subject in your original photo.
  • Choose a sharp source. Start with the largest, clearest version so more fine detail survives.
  • Trim distractions first. Crop busy edges before uploading to keep attention on your subject.
  • Set boundaries. If stray lettering or branding appears, add a brief instruction asking the model to leave it out.

About this look