Three part prompt:
1) show it the reference image and paste


Analyze the provided reference image and generate a modular, hyper-detailed, objective description of its content. Structure the output into four distinct, labeled modules: POSE & GESTURE, WARDROBE & PROPS, ENVIRONMENT & LIGHTING, and OPTICAL & COMPOSITION."

"For the content of these modules, focus exclusively on technical, photorealistic terms, avoiding all stylistic or interpretive language. Ensure every element meets these mandatory detail requirements:"

Mandatory Detail Requirements: Subjects & Precise Pose: Use directional vectors and approximate angles. Detail number of people, gender, and physical build. Specify the degree of head and torso rotation relative to the camera (e.g., “head rotated 5–10° right”). Describe precise limb placement, including joint angles (e.g., “right elbow bent 90°”), weight distribution (e.g., “weight centered, left knee slightly elevated”), and gaze direction relative to the lens. Clothing/Props: Describe material textures using PBR-friendly terms (e.g., “matte,” “glossy,” “light-scattering,” “porous”). Specify color, fit, and how props are held or positioned relative to the body, including the texture and placement of any fibers or smoke. Describe all colors as seen in the reference image, avoiding real-world interpretation defaults. Environment/Setting: Specify location, time of day, and directional lighting (e.g., “diffused overhead,” “rim light from left”). Note texture contrast between elements (e.g., “cracked earth vs. porous bread crumb”). Framing & Composition (Optical Mandates): Specify camera angle (e.g., “eye-level, horizontal”), suggested lens focal length (e.g., “85 mm prime”), and depth of field (e.g., “shallow cinematic focus, background blur/bokeh”). Define subject placement mathematically (e.g., “centered vertically, filling 70 % of frame”).


2) paste this in the same chat following the description


Generate the final structured image prompt for an ultra-realistic photograph. Use the descriptions from the specified modules, maintaining the modular structure with clear headings and using punctuation (like | and :) to separate constraints.

The prompt will end with this full mandate: 'This final image must be an absolutely hyper-realistic, professional photograph. Captured with a high-end full-frame DSLR camera and prime lens, using natural and studio lighting. Every detail must exhibit true physical-based rendering (PBR), cinematic depth of field, and authentic light falloff. Focus on intricate textures, micro-details, and realistic surface scattering for all materials (skin, fabric, metal, organic elements). Absolutely no digital illustration, painting, cartoon, 3D render artifacts, synthetic, or stylized effects whatsoever. The final output must be indistinguishable from an actual photograph shot on location by a master photographer.'


3) then copy the text output from that prompt, and paste in a new chat (new chat optional but it stops any unintended interference from the image)