Puzzle

Use the uploaded photo as the only face and character reference when generating the image. The model must copy 100% of the visual characteristics from the photo at render time, including: same gender, same age impression, same body type and overall proportions, same facial features and bone structure (jawline, cheekbones, chin, nose, eyes, eyelids, eyebrows, lips, ears), same skin tone and skin texture, and any unique marks such as moles, freckles, scars, wrinkles, pores. Preserve the exact hairstyle, hair volume, hairline, direction and length of the hair, including any messy strands or texture details. Keep the same facial expression, gaze direction, mouth position, and subtle emotion as in the photo. Do not change gender, do not make the person younger or older, and do not beautify or stylize the face in a way that alters identity. Treat the reference photo as a hard constraint: the generated character must look like the exact same person from the original image in every detail. COMPOSITION & SCENE A vertical cinematic portrait, medium shot showing the character’s head and upper torso. The character appears as a human jigsaw puzzle gradually assembling, with puzzle pieces forming and dissolving around the head and shoulders. The main puzzle silhouette defines the contour of the head, with some missing pieces revealing a dark background beneath. One realistic human hand in the foreground holds a single puzzle piece that clearly belongs to the character’s face, about to fit it into place. Composition centered on the face, with puzzle edges framing the head and scattered pieces around the borders, clean and uncluttered. STYLE & RENDERING Hyper‑realistic photographic style with high‑end commercial retouching, extremely detailed skin and puzzle textures. The puzzle pieces have slightly glossy edges, fine cut lines, and soft, realistic shadows between each piece. The skin appears as if printed perfectly onto puzzle cardboard, with the cut lines following the natural contours of the face. High resolution, ultra‑sharp focus on the face and nearby puzzle area, with subtle softness in the background pieces. No painterly brushstrokes or cartoon aesthetics; everything should look like a real studio photograph. LIGHTING & MOOD Dramatic studio lighting with a soft key light from a 3/4 frontal direction, creating gentle contrast and depth on the face and puzzle surfaces. A subtle rim light outlines parts of the head and puzzle silhouette, separating the character from the dark background. Shadows are soft but defined, enhancing the three‑dimensional feel of the puzzle pieces. Overall mood is introspective and mysterious, suggesting themes of identity, memory, and self‑assembly. COLOR PALETTE Dark, matte grays and blacks dominate the background and outer puzzle pieces. Skin tones exactly match the reference photo, with natural warmth and realistic color variation. The hand and facial puzzle pieces have slightly warmer highlights to draw attention to the assembly process. No neon or overly saturated colors; keep a grounded, cinematic palette with controlled contrast. CAMERA & PERSPECTIVE Portrait lens look, equivalent to 50–85mm, with natural perspective that preserves the proportions of the reference face. Slightly low or eye‑level angle to create intimacy and presence with the character. Shallow depth of field: the face and central puzzle area are in crisp focus, while background pieces and far edges fall into a gentle blur. Subtle lens bokeh in the background, no fisheye or wide‑angle distortion. HAND & PUZZLE INTERACTION The hand is hyper‑realistic, correctly proportioned, with natural skin texture, subtle veins, and soft specular highlights. Fingers gently pinch a single puzzle piece whose printed skin texture aligns perfectly with a missing section of the face. The piece should be clearly readable, with visible cut contours and a slight shadow on the puzzle below to reinforce depth. The gesture conveys care and precision, as if carefully reconstructing the person’s identity. NEGATIVE PROMPT / CONSTRAINTS No changes to gender, age impression, or core facial structure of the character. No anime, cartoon, comic, vector, or painterly illustration style. No extra objects, text, UI elements, logos, or watermarks in the frame. Avoid extra limbs, deformed hands, misplaced puzzle pieces, or duplicated facial features. Exclude heavy film grain, motion blur, strong chromatic aberration, glitch effects, or overlays that obscure details of the face and puzzle. You can paste this as a single prompt in Nano Banana (Image to Image) with your photo uploaded as reference; ajuste só detalhes de cena/ângulo conforme o uso.