And overall character identity from the high-resolution reference image Prompts

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a ipencil in one hand, matching the original hand position, angle, and posture precisely. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, no distortion in fingers or fan shape, professional commercial illustration quality. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering.

Use 100% face as an reference image

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a pen, and the tip of the pen is clearly and firmly touching the floor. The pen tip must be in direct contact with the ground, not floating, not lifted, with a visible point of contact between the pen tip and the floor surface. Maintain the exact hand position, wrist angle, and arm posture from the pose reference image. Do not change the direction or length of the pen. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, accurate object contact, no distortion in fingers or pen shape. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering. Use a white background.

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a pen, and the tip of the pen is clearly and firmly touching the floor. The pen tip must be in direct contact with the ground, not floating, not lifted, with a visible point of contact between the pen tip and the floor surface. Maintain the exact hand position, wrist angle, and arm posture from the pose reference image. Do not change the direction or length of the pen. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, accurate object contact, no distortion in fingers or pen shape. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering. Use a solid blue background.

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a pen, and the tip of the pen is clearly and firmly touching the floor. The pen tip must be in direct contact with the ground, not floating, not lifted, with a visible point of contact between the pen tip and the floor surface. Maintain the exact hand position, wrist angle, and arm posture from the pose reference image. Do not change the direction or length of the pen. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, accurate object contact, no distortion in fingers or pen shape. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering. Use a solid blue background.

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a pen, and the tip of the pen is clearly and firmly touching the floor. The pen tip must be in direct contact with the ground, not floating, not lifted, with a visible point of contact between the pen tip and the floor surface. Maintain the exact hand position, wrist angle, and arm posture from the pose reference image. Do not change the direction or length of the pen. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, accurate object contact, no distortion in fingers or pen shape. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering. Use a solid blue background.

Using the attached low-resolution image as pose reference and the attached high-resolution image as character style reference, generate a high-quality character image with the exact same pose and gesture. The character is holding a pen, and the tip of the pen is clearly and firmly touching the floor. The pen tip must be in direct contact with the ground, not floating, not lifted, with a visible point of contact between the pen tip and the floor surface. Maintain the exact hand position, wrist angle, and arm posture from the pose reference image. Do not change the direction or length of the pen. Maintain the same facial proportions, body ratio, outfit design, and overall character identity from the high-resolution reference image. Ultra high resolution, clean silhouette, smooth surfaces, refined edges, natural hand anatomy, accurate object contact, no distortion in fingers or pen shape. Soft and bright lighting, neutral clean background, marketing-ready visual, sharp focus, detailed textures, 4K quality, studio-level rendering. Use a solid blue background.

Use 100% face as an reference image

Gen photo of extremely ordinary and unremarkable iPhone selfie, with no clear subject or sense of composition—just a quick accidental snapshot. The photo has slight motion blur and uneven lighting from streetlights or indoor lamps, causing mild overexposure in some areas. The angle is awkward and the framing is messy, giving the picture a deliberately mediocre feel, as if it was taken absentmindedly while pulling the phone from a pocket. The main character is [uploaded person's face], and [NAME] stands next to him, both caught in a casual, imperfect moment. The background shows a lively los angeles street at night, with neon lights, traffic, and blurry figures passing by. The overall look is intentionally plain and random, capturing the authentic vibe of a poorly composed, spontaneous iPhone selfie.{ "title": "MASTER-ULTRA REALISTIC IDENTITY-LOCK PORTRAIT PROMPT (NIHAL FACE EDITION)", "description": "Production-grade, ultra-photoreal 8K portrait prompt for generating realistic, model-like images of the same person as the reference selfie, with 100% identity preservation and raw dermatological skin realism. Designed for SDXL / RealVis / Juggernaut XL / Flux-style models with InstantID or IP-Adapter FaceID. Absolutely no face change, no age change, no plastic skin, no AI/CGI look.", "contextual_parameters": { "engine_recommendation": { "primary_models": [ "SDXL Juggernaut XL v9", "RealVis XL v4", "SDXL base photoreal fine-tunes" ], "alternative_models": [ "Flux.1 dev (only if you already know its behavior and keep strong texture prompts)" ], "reasoning": "SDXL-based photoreal models currently provide the best balance of beautiful aesthetics with raw dermatological skin texture while keeping identity strict through InstantID / IP-Adapter FaceID.[file:1]" }, "identity_control": { "method": "InstantID + IP-Adapter FaceID / FaceID-PlusV2-SDXL", "reference_face_requirement": "Sharp, unfiltered selfie with visible natural skin texture, no beauty filter, even lighting, straight or slight 3/4 angle.[file:1]", "strict_identity_settings": { "ip_adapter_weight": 0.90, "instantid_weight": 0.65, "subject_identity_lock": 0.95, "facial_landmarks_lock": 0.90, "skin_texture_lock": 0.92, "reject_if_eye_distance_change_percent_over": 6.0 }, "constraints": [ "No change to bone structure, jawline, nose shape, eye distance, lip shape, or natural facial asymmetry.", "No age change, no gender change, no beauty-filter slimming or stylization of the face." ] }, "camera_and_resolution": { "base_generation_resolution": "1024x1536 or 1024x1024 (portrait aspect preferred for face).[file:1]", "upscale_strategy": "Generate at SDXL-native, then upscale 2x–4x (Hires fix or external upscaler) to reach 4K–8K. Avoid direct 8K generation in one pass to prevent artifacts.[file:1]", "virtual_camera": { "primary": "Phase One IQ4 150MP medium-format digital back", "lens": "Schneider Kreuznach 80mm portrait lens at around f/5.6", "alternate_emulation": "Hasselblad X2D / Canon R5 / Leica M10-P with 50–85mm portrait lenses for natural perspective and realistic lens behavior.[file:2]" } }, "sampler_and_quality": { "sampler": "DPM 2M or DPM 3M SDE Karras", "steps": 30, "cfg_scale": 4.5, "hires_fix_denoise": 0.20, "adetailer_denoise": 0.15, "notes": "Too high CFG or denoise can cause plastic skin or overbaked contrast; staying in these bands preserves pores and vellus hair.[file:2]" }, "safety_and_style_constraints": { "no_ai_look": true, "no_cartoon_or_anime": true, "no_makeup": true, "no_body_horror": true, "no_child_appearance": true, "goal_style": "High-end fashion/editorial + National Geographic documentary realism with filmic grading, but absolutely photographic and natural.[file:1]" } }, "master_prompt": { "positive_prompt": "Ultra-photorealistic, masterpiece, 8K UHD, raw photo, professional studio-grade portrait of the exact same male identity as the reference selfie, no face change, no age change, no gender change, preserve original bone structure, natural asymmetry, same nose shape and bridge width, same jawline curvature, chin shape, eye distance, eyelid structure, eyebrow thickness and direction, lip contour and proportions, authentic facial hair pattern, micro scars, moles and unique facial landmarks. Model-like handsome masculine look with a lean, athletic, swimmer-type physique, around 6 feet tall appearance, defined neck and shoulder muscles visible in frame, confident but relaxed energy, calm intense gaze straight into the camera, subtle natural smirk, natural expression with no overacting. Shot as if on a Phase One IQ4 150MP medium-format camera with a Schneider Kreuznach 80mm portrait lens at approximately f/5.6, diffraction-limited edge-to-edge sharpness, natural perspective, realistic depth compression, and medium depth of field. Face and eyes are razor sharp, with extremely detailed irises, clear limbal ring, subtle veins in sclera, natural moisture and reflections, single soft studio catchlight in each eye, realistic eyelash density and direction, perfectly preserved eyebrow shape with individually resolved hairs, authentic beard and moustache edges, natural beard density, visible stubble micro-texture and individual facial hairs without over-sharpened outlines. Ultra-detailed skin physics and dermatological realism: raw, unretouched skin texture with macro-visible pores at 20–40x magnification clarity on nose, cheeks, forehead and chin, epidermal irregularities, slight redness, faint freckles or pigmentation if present in the real face, natural uneven skin tone, asymmetrical pigmentation, micro-bumps, and realistic micro-shadows in pores. Directional vellus hair (peach fuzz) across cheeks, jawline and sides of the face, visible in rim light with individual micro-hairs catching the light, subtle natural skin oil with a gentle sebaceous sheen (not oily, not plastic), multi-layer subsurface scattering with soft reddish translucency in ears, nose tip and thin skin areas, natural micro-wrinkles and expression lines around eyes, forehead and nasolabial folds, subtle under-eye texture with real tear ducts and waterline detail. Dermatological authenticity: no beauty-filter smoothing, no pore loss, no over-processed plastic surfaces, all real-world imperfections preserved in a flattering, cinematic way. Skin material and optics described with physical accuracy: realistic micro-surface roughness, non-uniform BRDF distribution across T-zone and cheeks, micro-occlusion shadowing in pores and fine creases, natural specular highlights shaped by true skin oil patterns instead of uniform gloss, proper dermal thickness feeling, translucent quality in backlit areas, realistic thermal signature implied by slightly warmer ears and nose, documented micro blemishes and scars if present, preserved micro-markers like wrinkle topology, pore clusters and fine hairs. The face must look exactly like an ultra-high resolution photograph from a professional medium-format system, not AI-generated. Lighting and atmosphere: natural diffused key light around 5400–5600K (daylight-balanced softbox or large window) placed at about 45 degrees to the face in classic Rembrandt or loop-lighting configuration, soft but directional, sculpting cheekbones and jawline. Add a hard raking side-light component from the key direction to carve out pores and vellus hair, creating micro-shadows across the skin surface, emphasizing texture without harsh clipping. Gentle negative fill on the shadow side to slim and sculpt the face, avoiding flatness, with cinematic chiaroscuro contrast and full tonal range from rich shadows to soft highlights. Subtle cool rim light on jawline and hair to separate the subject from the background, revealing hair edges and peach fuzz, tiny dust particles and light haze in the beam for very light volumetric atmosphere. Shadows are realistic and soft-edged, with smooth gradations, no blown highlights, no crushed blacks, no HDR halos. Color science and grading: neutral white balance with accurate, lifelike skin color, ProPhotoRGB / Hasselblad or Phase One natural color science look, soft filmic S-curve contrast, 24-stop-like dynamic range behavior, gentle highlight roll-off, rich but natural midtones, no strong color cast, no orange skin, no magenta skin. Subtle Kodak Portra 400-style organic film grain, very fine and integrated into the image (grain inside the texture, not painted on top), cinematic yet realistic grading with slightly warm skin tones and slightly cooler, desaturated background shadows. No over-saturation, no excessive clarity, no unnatural teal-orange extremes; everything feels like a professional, high-end editorial portrait. Wardrobe, background and composition: fitted, stylish outfit such as a charcoal blazer, dark smart shirt or henley that enhances the masculine model look, with visible fabric weave, stitching details and natural folds creating micro-shadows; no loud logos, no distracting graphics. Background is a premium but simple environment: either a modern studio with a neutral gradient backdrop or an upscale rustic pub / luxury interior with dark wood, glass reflections and warm tungsten bokeh, softly blurred. Composition follows rule-of-thirds with a head-and-shoulders or slightly above-chest crop, straight-on or very subtle 3/4 angle, camera at eye level for strong presence. Slight lens compression for flattering facial proportions, natural DOF with sharp face and gently blurred background, beautiful optical bokeh with rounded highlights and realistic depth layering. Overall mood: cinematic, masculine, confident, authentic, like a real high-end professional photoshoot for a fashion magazine cover or premium advertising campaign. Image must feel like a genuine high-resolution camera capture with zero AI or CGI artifacts, absolutely no change to the original face identity, and fully preserved original skin texture and individuality. (ultra photoreal, masterpiece, best quality, raw photo:1.4), (8K UHD, hyper-detailed:1.3), (natural skin pores, epidermal irregularity, vellus hair, subsurface scattering:1.4), (hard raking light micro-shadows in pores:1.3), (documentary realism, National Geographic + high-end fashion editorial aesthetic:1.3), (unretouched authentic photograph:1.3)", "negative_prompt": "plastic skin, wax texture, thick skin, porcelain skin, mannequin, doll, CGI, 3D render, game render, toon, cartoon, anime, comic style, illustration, painting, drawing, cel shading, stylized face, AI texture patterns, pore tiling, uniform skin texture, fake specular highlights, plastic shine, over-smoothed skin, airbrushed, beauty filter, FaceApp look, Snapchat filter, Instagram filter, retouched beauty guru makeup, heavy foundation, contouring, glitter, fantasy makeup, skin-smoothing apps. face change, identity change, identity drift, different person, swapped face, altered face, morphed face, changed bone structure, changed jawline, jawline slimming, jawline reshaping, nose reshaping, nose reduction, nose enlargement, changed eye shape, enlarged eyes, narrowed eyes, misaligned eyes, crossed eyes, lazy eye, warped eyelids, deformed face, asymmetry beyond natural, extra eyes, extra teeth, double pupils, split face, duplicated facial features, distorted mouth, broken lips, melted skin, dislocated features, wrong face proportions. wrong body proportions, huge head small body, short height look, distorted neck, broken shoulders, twisted torso, bad anatomy, missing fingers, extra fingers, fused fingers, deformed hands, warped ears, incorrect ear position, unrealistic skull shape, bizarre limb lengths, unnatural posture, broken spine. low resolution, low quality, pixelated, muddy details, heavy compression, JPEG artifacts, noise blobs, banding, posterization, ringing, halos, chromatic aberration glitches, lens dirt artifacts on face, motion blur on face, out-of-focus eyes, blur on pores, depth of field focused on wrong plane, fake bokeh overlay on skin. overexposed highlights, blown-out skin, crushed blacks, extreme contrast, neon saturation, orange-teal overgrading, strong color cast (green, magenta, cyan), flat lighting, on-camera flash look, harsh specular hotspots, wrong shadow direction, inconsistent lighting, HDR fake glow, bloom, glows around edges, hard vignettes, fisheye distortion, extreme wide-angle selfie distortion, warped perspective, tilted horizon (unless artistically intentional in full-body shots). old age (if not requested), exaggerated wrinkles inconsistent with real age, zombie skin, horror elements, gore, fresh injuries, fantasy implants, cyberpunk chips, sci-fi tattoos, face piercings not present in reference, excessive jewelry, distracting accessories, unrealistic clothing logos. any sign of synthetic, uncanny-valley, AI-generated appearance: over-sanitized skin, perfect doll-like features, absence of pores and vellus hair, plastic or waxy reflections, painterly edges, oversharpening halos, unnatural edge contrast, smear artifacts, or anything that does not look like an authentic high-resolution photograph from a real camera." }, "identity_lock_positive_prompt": "Ultra-photorealistic 8K UHD raw portrait photo of the exact same male identity as the reference selfie. Preserve 100% of the original facial identity with no face change, no age change, no gender change: keep identical nose shape and width, bridge and tip form, same jawline and chin geometry, same cheekbone structure, same eye distance, eye shape, eyelid fold pattern, same eyebrow shape and thickness, same lip shape and fullness, same natural facial asymmetry, same beard and moustache pattern, same hairline and forehead proportions. Realistic, model-like version of this real person, with slightly enhanced grooming and lighting but absolutely unchanged facial identity. Face framed in a professional portrait style: direct camera gaze with confident but relaxed expression, calm intensity in the eyes, subtle natural smirk or neutral relaxed mouth position, micro-expression lines maintained for authenticity. Detailed skin with clearly visible pores and vellus hair at macro scale, real-world imperfections, no beauty-filter smoothing, subsurface scattering in ears and nose, neutral white balance, cinematic lighting, shot on Phase One IQ4 150MP with an 80mm Schneider portrait lens, in-camera look, unretouched authentic photograph aesthetic.", "identity_block_negative_prompt": "face change, identity change, different person, altered face, morphed face, swapped face, changed nose, changed jawline, changed chin, changed lips, changed eye shape, eye displacement, ear misalignment, skull reshaping, exaggerated symmetry, idealized generic model face, identity drift, age change, deformed face, cartoon face, anime face, stylized face, low resolution deformities, thick plastic skin, waxy skin, airbrushed skin, beauty filter artifacts, CGI look, 3D render, doll-like appearance, overly smooth poreless skin, blur on face, bad anatomy, wrong face proportions.", "step_by_step_instructions": [ { "step": 1, "title": "Prepare the reference selfie", "details": "Use a clean, unfiltered photo of your face with natural or soft window lighting, no beauty filters, and visible skin texture. Crop to a portrait framing where your full face is clear, eyes sharp, and no strong motion blur.[file:2]" }, { "step": 2, "title": "Load model and identity modules", "details": "Load an SDXL-based photoreal model such as Juggernaut XL v9 or RealVis XL v4. Enable InstantID or IP-Adapter FaceID (FaceID-PlusV2-SDXL) as ControlNet / adapter, and load your reference selfie into the identity input slot.[file:1]" }, { "step": 3, "title": "Set identity-lock weights", "details": "Set IP-Adapter weight around 0.88–0.92 and InstantID weight around 0.60–0.70. Lock facial landmarks and skin texture with high weights (around 0.9) so that your face structure and pores are preserved, but lighting and pose can still adapt naturally.[file:2]" }, { "step": 4, "title": "Paste positive and negative prompts", "details": "Copy the full master positive_prompt into your Positive Prompt field and the full negative_prompt into your Negative Prompt field. For identity-critical runs, optionally also include the identity_lock_positive_prompt and identity_block_negative_prompt in the same respective fields for extra strictness.[file:1]" }, { "step": 5, "title": "Configure sampler and resolution", "details": "Use sampler DPM 2M or DPM 3M SDE Karras, around 30–40 sampling steps, CFG scale 4–5, and base resolution around 1024x1536 or similar portrait aspect. Enable Hires fix or an external upscale pass (2x–4x) to reach 4K–8K while preserving skin texture.[file:2]" }, { "step": 6, "title": "Generate and inspect skin texture", "details": "Generate the portrait, then zoom into 100–300% on nose, cheeks and forehead to verify that pores, vellus hair and micro-wrinkles are clearly visible, and that skin does not look plastic or airbrushed. If it looks too smooth, slightly lower CFG and denoise, or strengthen vellus hair / pores terms.[file:2]" }, { "step": 7, "title": "Optional ADetailer / face pass", "details": "If the initial generation loses some micro texture, run a low-denoise face detailer pass (around 0.12–0.18) focused on the face with a short prompt emphasizing macro pores, vellus hair and raw skin detail, without changing identity.[file:1]" } ], "output_format_guidelines": { "file_formats": [ "16-bit TIFF (for maximum editing latitude)", "High-quality JPEG or PNG (for web and sharing)" ], "resolution_targets": [ "Minimum: 4096 px on the long side", "Preferred: 6144–7680 px on the long side for 8K-style detail" ], "post_processing_notes": [ "Avoid aggressive skin retouching, frequency separation or blur that removes pores.", "If adding grain, keep it subtle and film-like so it integrates with existing texture rather than covering it.[file:2]", "Do not use automated face-restoration tools that create plastic or waxy skin." ], "usage_hint": "In most UIs, paste only the master positive_prompt into the main prompt box and the negative_prompt into the negative box. Use the JSON here only as a reference structure for your workflow, not as text inside the model prompt field." } }

{ "title": "01_circle_on_couch_horror_icons", "type": "image", "identity_reference": { "use_uploaded_photo": true, "preserve_face": true, "preserve_body": true, "match_outfit_and_accessories": true }, "scene": { "setting": "vintage living room arranged in a loose circle", "props": ["bean bag", "cigar"], "seating": "all characters seated casually on a couch set around the subject" }, "subjects": { "main": "the person from the uploaded photo sitting on a bean bag, wearing the same sweater, flared pants, brown platform boots, dark shades, jewelry", "companions_inspired": [ "silent brute in worn work clothes with a cracked hockey-style mask", "tall figure with blank white mask and dark coveralls", "small doll-like trickster with freckles and overalls, plastic grin", "pale clown-faced figure with fiery hair and exaggerated smile", "hooded figure with elongated ghost mask" ] }, "camera": { "framing": "medium-wide, eye level", "lens": "35mm", "distance": "3 meters", "composition": "subject centered, companions forming a semicircle" }, "lighting": { "key": "warm key on the subject", "fill": "low fill", "rim": "soft rim on villains", "mood": "moody and dramatic", "effects": ["subtle smoke", "cigar ember glow"] }, "style": { "look": ["realistic", "high-resolution", "staged photoshoot", "horror-comedy vibe", "That 70s Show seating reference"], "grain": "35mm film grain light" }, "quality": { "resolution": "4k", "sharpness": "high" }, "negatives": [ "do not alter face or identity", "no text or watermark", "no gore or blood", "no trademarks or direct character names", "no extra fingers", "no deformed hands" ], }

character --sref 3299013505 --v 7

same image ultra details

various characters both male and female using the attached image as reference. Give them lower facial detail than the reference image.

use the second input image to make the character from input image 1 be structured like that.

same image ultra details

high resolution photo --chaos 75 --ar 1:2 --sref 2977083251 --profile ooro6tr --stylize 750

character --sref 4015922619 --v 7

generates exactly the same image but with better quality

create a new character combining all the images

generates exactly the same image but with better quality

generates exactly the same image but with better quality

regenerate image with only 3 characters

Upscale the image to 4k by using your ai intelligence. Enable strict facial consistency mode. Prioritize the facial features from the provided reference image for all subsequent generations. Maintain the subject's identity accurately while only adapting the pose, lighting, and background. Do not alter the core facial structure.

generates exactly the same image but with better quality