{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
{ "prompt_type": "descriptive_replication", "reference_adherence": "STRICT_VISUAL_FIDELITY", "aspect_ratio": "9:16", "identity_lock": { "priority": "ABSOLUTE", "instruction": "Use MY EXACT FEMALE MODEL Adèle. Facial identity must be perfectly preserved with zero drift. All details below must be followed exactly.", "cranial_structure": { "shape": "elongated oval skull with wide bizygomatic width", "forward_growth": "maxilla and mandible projected forward, strong skeletal framework", "facial_thirds": "upper : middle : lower = 1:1:0.9, slightly elongated lower third", "cheekbones": "high‑set, razor‑sharp, beginning directly under eyes, sweeping laterally, forming deep hollow cheeks with natural submalar shadows", "mandible": "sharply defined, gonial angle approx. 118°, clean V‑shaped jawline", "chin": "slightly pointed but firm, aligned with nasal septum" }, "eyes": { "shape": "almond‑shaped, deeply set", "canthal_tilt": "positive 5° (outer corners higher)", "medial_canthi": "acutely angled, extending downward", "scleral_show": "zero (lower lids tightly hug iris)", "eyelids": "hooded with minimal exposure", "iris": "light green with golden‑brown central heterochromia, distinct limbal ring, visible radial fibers", "eyebrows": "thick, straight, set low on supraorbital ridge, soft arch peaking slightly lateral to lateral limbus, individual hairs visible" }, "nose": { "dorsum": "straight, smooth", "bridge": "refined, narrow", "tip": "precisely sculpted, slightly rounded, delicate cartilage definition", "nostrils": "narrow, symmetrical, slight columellar show" }, "lips": { "fullness": "full, lower lip to upper lip ratio approx. 1.3:1", "cupids_bow": "pronounced with distinct peaks", "vermilion_borders": "distinct", "oral_commissures": "sharp, turning slightly upward", "philtrum": "short, well‑defined, approx. 11 mm" }, "skin": { "tone": "warm‑neutral with subtle olive undertone, faint golden undertones in cheeks", "texture": "hyperrealistic, photorealistic, ultra‑detailed", "pores": "barely visible micropores 0.1–0.2 mm in T‑zone", "micro_roughness": "present", "orange_peel": "ultra‑fine micro‑depressions visible under side lighting", "subsurface_scattering": "moderate on nose, cheeks, earlobes, lifelike translucency", "vellus_hairs": "delicate on jawline and cheekbones", "micro_veins": "faintly discernible under translucent skin", "specular_highlights": "healthy on forehead, nose, cheekbones; matte cheeks" }, "hair": { "color": "chestnut brown with natural California blonde balayage (bronde), multi‑tonal, soft caramel and honey streaks", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "body": { "height": "approx. 172 cm", "build": "lean athletic, low body fat (18–20%) with hourglass figure", "neck": "slender, elongated", "clavicles": "prominent, horizontal", "shoulders": "faint muscle definition with delicate venous network", "curves": "subtle feminine curves", "chest": "full bust with natural projection" } }, "version_1_soft_smile": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft gentle smile, lips slightly parted, only a small amount of upper teeth visible, pleasant and relaxed", "gaze": "directly into camera lens, intense and confident", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_2_lip_bite": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "subtle gentle lower lip bite, tasteful and non‑exaggerated, lips slightly engaged with teeth", "gaze": "directly into camera lens, confident and focused", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "version_3_romantic_look": { "subject": { "demographics": "Young adult female (mid‑20s), my exact model Adèle", "hair": { "color": "chestnut brown with natural blonde balayage (bronde), multi‑tonal", "style": "long wavy, cascading over shoulders, with side‑swept curtain bang grazing outer eye corner", "texture": "glossy, healthy, individual strands visible, anisotropic reflections, no frizz" }, "face": { "expression": "soft affectionate expression, warm emotional gaze, relaxed eyelids, lips gently closed or slightly parted", "gaze": "directly into camera lens, warm and emotional but still direct eye contact", "eyes": "confident focused gaze, slightly narrowed eyelids for a model‑like look, eyes clearly visible with strong presence, warm affectionate quality", "makeup": "none, natural look", "visibility": "full face visible, head perfectly straight, chin level, no tilt, no rotation" }, "body": { "pose": "standing perfectly straight, shoulders square to camera, torso facing directly forward, head aligned to camera sensor plane, no body rotation, no perspective distortion", "posture": "upright, symmetrical", "anatomy": { "curves": "subtle feminine curves", "chest": "full bust with natural projection", "details": "visible collarbones, slender elongated neck" }, "skin_texture": "visible fine pores, micro‑roughness, delicate vellus hairs, natural sheen from sebum film, no airbrushing, subsurface scattering on nose and cheeks" }, "clothing": { "description": "simple, elegant top (e.g., fitted black or neutral top), allowing focus on expression" } }, "environment": { "setting": "professional photo studio with neutral gray seamless background", "background": { "description": "smooth gray studio backdrop, evenly lit, no gradients, no objects, no shadows on wall", "lighting": "soft frontal studio lighting, symmetrical illumination on both sides of face, clear catchlights in both eyes, no dramatic shadows, no side lighting, no rim lighting" } }, "camera_and_technical": { "perspective": "straight‑on, eye level, perfectly centered", "camera_position": "professional studio camera with 85mm equivalent portrait lens", "framing": "vertical 9:16, from mid‑chest (upper bust) to slightly above head, subject centered, shoulders fully visible", "focus": "sharp on entire face, hyperrealistic detail", "visual_fidelity": "hyperrealistic, photorealistic skin texture, visible pores, natural skin shading, no beauty filters, no airbrushing, high resolution studio photography, professional portrait quality" } }, "global_negative_prompt": [ "head tilt", "angled head", "head rotation", "body rotation", "looking away", "looking sideways", "looking off-camera", "3/4 view", "asymmetrical shoulders", "perspective distortion", "different face", "beauty filters", "airbrushed skin", "anime", "cartoon", "over-sharpening", "clean digital look", "perfect exposure", "smooth gradients", "visible ears", "ears showing", "messy hair", "flat hair", "greasy hair", "oily face", "greasy skin", "overexposed", "shiny T-zone", "glossy skin", "one-length haircut", "blunt cut", "excessive makeup", "CGI", "3d render", "plastic texture", "smooth", "airbrushed", "digital art", "painting", "deformed face", "asymmetrical eyes", "extra facial features", "blurry", "low detail", "unrealistic proportions", "bad anatomy", "acne", "skin imperfections", "blemishes", "redness", "pimples", "scars", "moles", "shadows on wall", "gradients in background", "objects in background", "side lighting", "rim lighting", "dramatic shadows" ] }
Use the provided reference image as the primary identity source. IDENTITY LOCK: Replicate the exact same person — facial structure, proportions, skin tone, and natural asymmetries must match precisely. Do not beautify or alter identity. SCENE & SETTING (INDOOR — MODERN YOGA STUDIO): Ultra-realistic photo captured inside a modern, high-end yoga studio with a dim, warm ambiance. Minimalist interior with polished wooden floors, soft indirect lighting, and subtle architectural elements. Background slightly out of focus but readable, with natural imperfections and depth. FRAMING & CAMERA (CRITICAL): Smartphone perspective (NOT DSLR), slight wide-angle distortion typical of phone cameras. Camera positioned very low, near floor level, slightly behind the subject, angled upward (low-angle shot). Close proximity to the subject creates subtle perspective exaggeration (near-body elements appear slightly larger). Framing is slightly tilted, imperfect, and casual — not centered, not symmetrical. Mid-body to upper-thigh composition. POSE & BODY LANGUAGE: The woman is on a pink yoga mat in a four-point position (hands and knees on the ground). Knees slightly apart, feet relaxed. Back naturally arched (not exaggerated), hips elevated as the highest point closest to the camera. Torso angled forward while the head turns over the shoulder toward the camera. Pose feels candid and mid-movement, not staged — confident, controlled, subtly playful. EXPRESSION: Direct eye contact with the camera. Calm, slightly playful expression with relaxed lips. Natural facial asymmetry and micro-tension — no perfection or stiffness. BODY (SEXY BUT REALISTIC): Naturally athletic physique with soft curves. Subtle muscle tone (legs, glutes, arms, core) — visible but not extreme. Realistic anatomical proportions and joints. Natural asymmetry and posture tension — no exaggerated or artificial features. SKIN (CRITICAL): True-to-life skin rendering: visible pores, subtle peach fuzz, slight uneven skin tone, natural highlights and soft shadows, minor imperfections. NO plastic smoothing, NO airbrushed or CGI look. HAIR (NATURAL BUT CONTROLLED): Long, softly wavy hair falling naturally with gravity. Mostly cohesive, clean direction, minimal flyaways. Lightly groomed appearance with subtle natural texture. OUTFIT (FITNESS — REALISTIC FABRIC): Vibrant turquoise athletic shorts (short length), textured sweat-wicking fabric. Fitted tank top with thin straps, slightly loose — one strap subtly slipping off the shoulder. Fabric behaves physically correct: natural compression, slight wrinkles, tension at seams, realistic stretch. No painted-on effect. LIGHTING (PHYSICALLY ACCURATE): Dim indoor lighting with warm tones. Soft directional light creating gentle shadows and natural highlights on the body. Slight uneven illumination for realism. Image remains clear due to realistic smartphone HDR behavior. CAMERA & RENDERING (SMARTPHONE REALISM): Slight noise in shadow areas. Balanced dynamic range (minor highlight clipping, shadows not fully recovered). Natural skin tones with slight warmth. Subtle sharpening only. No cinematic grading, no HDR exaggeration, no DSLR compression. COMPOSITION: Casual lifestyle composition, slightly off-center. Natural spacing, imperfect crop, slight tilt. Foreground emphasis due to low angle, background softly receding. FINAL RESULT: A believable, ultra-realistic smartphone photo captured from a low angle inside a yoga studio. Natural, immersive, slightly intimate, with realistic imperfections in skin, lighting, and environment. Indistinguishable from a real candid fitness photo — NO CGI look, NO artificial perfection, NO exaggerated anatomy.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.
Create ONE single cinematic frame or a short 3-6 second video sequence of the EXACT SAME real human being, with permanently anchored identity, where Reference A and Reference B ALWAYS represent the SAME PERSON, reconstructed physically and perceptually inside the world of Reference C. Master dominance order: identity geometry and ratios first, dermis and skin tone second, beard and under-eye third, body coherence fourth, camera and lighting fifth, scene and style last. Lower systems MUST always adapt to higher systems and may never override them. Source ownership: facial geometry, stable identity ratios, asymmetry, and expression baseline come from A; dermal truth, pore structure, skin tone, undertone, melanin reading, redness pattern, beard density, beard irregularity, beard color variation, beard-to-skin transition, and under-eye truth come from A+B; body coherence, neck, shoulders, posture, and body scale come from B; environment, props, scene architecture, mood, and non-identity world come from C. Visible identity must ALWAYS read as A+B. Visible non-identity world must follow C exactly only when safe for identity. Absolute facial protected regions: eyes, eyelids, brow structure, brow asymmetry, canthal tilt, under-eye darkness, under-eye folds, under-eye depth, nose bridge, nostrils, philtrum, mouth corners, jawline, chin projection, beard boundary, cheek volume, and hairline are immutable and may not be altered by pose, lens, lighting, grading, environment, temporal interpolation, or cinematic style. Structural immutability lock: preserve exact facial geometry, exact feature placement, exact eye spacing, exact eye offset, exact nose width and projection, exact nostril asymmetry, exact mouth width, exact mouth-corner imbalance, exact mouth-to-chin distance, exact jaw width, exact chin projection, exact beard boundary, exact left-right asymmetry, and exact expression baseline. Forbidden: reshaping, symmetry correction, alignment correction, centering correction, optimization, beautification, model-face effect, influencer-face effect, younger-looking transformation, flattering reconstruction, glamourization, facial cleanup, cosmetic correction, facial redesign, or replacement by a generic attractive face. If the face looks better, cleaner, prettier, younger, smoother, sharper in an artificial way, more symmetric, or more attractive than A+B, the output is invalid. Skin tone and dermis lock: preserve exact skin tone, exact undertone, exact melanin distribution, exact redness pattern, exact tonal variation, exact natural chromatic depth, exact beard-to-skin transition, visible pores with non-uniform distribution, realistic tonal irregularity, realistic micro imperfections, natural oil reflectivity variation, realistic subsurface depth, forehead slightly more reflective, cheeks medium, under-eye darker and thinner, and chin with denser pore visibility. Forbidden: whitening, darkening, recoloring, ambient tint contamination, warm shift, cool shift, cinematic skin grading, orange-teal contamination, skin glow, skin polish, beauty retouch skin, pore reduction, wrinkle reduction, smoothing, plastic skin, wax skin, porcelain skin, AI skin, fake hyper-detail, invented pores, or artificial gritty texture. Under-eye truth lock: preserve darkness, folds, depth, fatigue, and hollow if present; forbid brightening, smoothing, cleanup, concealer effect, or youth correction. Beard integrity lock: preserve irregular density, natural edge, uneven growth, patchiness if present, exact color variation, and exact natural reading from A+B; forbid lineup cleanup, barber sharpening, density normalization, beard filling, edge stylization, or cosmetic beard design. Body coherence lock: body, neck transition, shoulder relation, posture, and scale must support the same-person reading without changing facial identity or making the head look pasted, disconnected, too large, too small, or mismatched. Camera safety lock: use identity-safe lens behavior and perspective; preserve natural craniofacial reading; forbid wide-angle distortion, selfie distortion, nose enlargement, face stretch, compression artifacts, warped perspective, focal distortion, or any camera effect that changes identity reading. Lighting integrity lock: lighting must adapt to the preserved face and never redefine the face. Environment lighting from C may shape the world, but may NOT recolor, flatten, beautify, contaminate, or cosmetically modify the face. Facial tone, undertone, dermis, contrast, and shadow logic must remain anchored to A+B even inside C. Forbidden: color cast on skin, washout, flattening, mixed-color contamination on face, unnatural relighting, cinematic grading that changes facial reading, or environment tint contamination on skin. Temporal hard lock for video: identical identity, asymmetry, dermis, skin tone, beard, under-eye, geometry, and facial proportions in every frame; zero drift, zero flicker, zero morphing, zero smoothing over time, zero beard shimmer, zero detail pumping, zero geometry breathing, zero identity evolution. If drift appears, revert to base identity truth from A+B. ZERO TEXT ERADICATION SYSTEM: the final output must contain ZERO readable text of any kind anywhere in the frame. If Reference C contains text, letters, words, numbers, captions, subtitles, signs, logos, labels, brands, packaging text, poster text, interface text, watermark, signature, stamp, UI marks, or readable characters, all of them must be completely removed and replaced with clean, natural, scene-consistent reconstruction. Leave no traces, no ghosting, no partial letters, no blur patches, no typographic residue, no censored bars, and no artifact edges. Clean negative filter: forbid pasted-face look, split face, duplicate features, warped ears, warped beard, warped jaw, facial blur, CGI look, over-sharpening, over-clarity, over-contrast skin, fake realism, extra fingers near face, or hand interference over the face. Failure conditions: block output if there is identity drift, symmetry increase, beautification, undertone shift, skin-tone shift, dermal loss, under-eye alteration, beard inconsistency, facial contamination from scene lighting, camera distortion, perspective distortion, temporal instability, any readable text, any watermark, any logo, any subtitle, or any signature anywhere. Final target: real human, zero deformation, zero beautification, exact identity locked, exact skin tone locked, exact dermis locked, exact beard locked, exact under-eye truth locked, identity-safe camera, identity-safe lighting, scene integrated from C without facial contamination, and zero readable text anywhere, optimized for Nano Banano Pro first and Kling 3.0 second.