{ "_meta": "Charity Right brand photography style — forensic analysis from 8 reference images across Ramadan homepage, One Eats One Waits, Make It Right, Fidya/Kaffarah pages", "anti_ai_rules": { "NEVER_SAY": [ "golden hour", "warm amber glow", "beautiful bokeh", "cinematic", "hopeful eyes", "gentle smile", "content expression", "grateful", "Sony A7", "Leica", "85mm f/1.4", "Portra 400", "shallow depth of field", "editorial photography", "professional", "dignified" ], "WHY": "These are the top AI photography prompt clichés. Every AI image uses them. They produce smooth, over-warm, emotionally directed stock photos that scream synthetic." }, "light": { "description": "Single hard directional source. NOT even soft glow everywhere.", "observed_in_refs": [ "cr-hunger6-zakat: harsh midday Middle Eastern sun, hard shadows under chin, blue sky", "cr-final: late afternoon side-light, long shadows on cracked earth, NOT golden", "fidya-hero: single window, hard beam falling across table and face, rest is dark", "fidya-quran-elder: single small window right side, chiaroscuro, deep shadows", "fidya-child-meal: classroom window right side, directional, hard shadows on desk", "oneeats UK: tungsten overhead + window, mixed color temps, NOT uniform warm" ], "prompt_pattern": "Lit by a single [window/afternoon sun/overhead bulb]. Hard directional light from the [left/right]. Deep shadows on the opposite side. Unlit areas stay dark." }, "color": { "description": "MUTED and slightly DESATURATED. NOT oversaturated warm tones.", "palette": { "dominant": ["dust beige", "concrete grey", "faded ochre", "muted teal sky"], "accents": ["terracotta fabric", "olive green", "faded blue", "worn brass"], "avoid": ["rich amber", "deep orange", "vivid warm tones", "saturated anything"] }, "grade": "Slightly pulled-back saturation. Faded, dusty quality. Blacks are lifted slightly (not crushed). Highlights have a faint yellow-green cast. Shadows lean cool/neutral, not warm." }, "skin_texture": { "description": "REAL skin with visible imperfections. This is the #1 AI tell.", "details": [ "Visible pores on nose and cheeks", "Uneven skin tone — darker patches around eyes, lighter on forehead", "Dust or slight dirt on face/hands (not clean-scrubbed)", "Chapped or dry lips", "For elders: deep wrinkles, age spots, visible veins on hands", "For children: slightly ashy knees/elbows, scabbed mosquito bites, runny nose", "Single catchlight in each eye from the single light source — NOT ring lights" ] }, "hair": { "description": "Messy, uncombed, natural. NOT styled or neat.", "details": [ "Fly-away strands catching backlight", "Dusty, not freshly washed", "Irregular — some parts flat, some sticking up", "For headscarves: fraying edges, loose threads, real textile weave visible" ] }, "clothing": { "description": "Simple, worn, faded — but CLEAN (not filthy). Real fabric behavior.", "details": [ "Fabric is thin, slightly translucent where light hits", "Visible stitching, slightly uneven hems", "Faded color — a shirt that was once bright blue is now pale", "Creased and wrinkled from being worn all day", "Collar slightly askew, buttons not perfectly aligned", "SPECIFIC garments: shalwar kameez, dupatta, cotton kurta — not generic 'traditional dress'" ] }, "composition": { "description": "IMPERFECT documentary framing. Not centered, not perfectly composed.", "details": [ "Partial figures intruding at frame edges — someone's arm, shoulder, back", "Subject slightly OFF-CENTER — not rule-of-thirds perfect", "Foreground OBSTRUCTION — shot through a doorway, past a wall edge, between objects", "Background has DEPTH with real objects at multiple planes", "NOT: single subject centered on clean bokeh background (that's the AI default)" ], "observed": [ "cr-hunger6-zakat: someone's arm/torso enters from left edge, buildings and wires behind", "fidya-feeding: volunteer's arm enters from left, other people mid-ground", "kaffarah-community (the good one from other pages): multiple depth layers, cluttered" ] }, "environment": { "description": "SPECIFIC, CLUTTERED, REAL locations. Not clean minimalist settings.", "details": [ "Walls: crumbling plaster patches, exposed brick, repair marks, damp stains", "Hanging objects: cloths on nails, pots on hooks, photos in cheap frames", "Floor: cracked tiles, bare concrete, swept-dirt ground, woven mats with fraying edges", "Infrastructure: power lines overhead, water tanks on roofs, satellite dishes", "Furniture: wooden charpai with sagging rope, worn benches, plastic chairs", "Regional markers: Urdu/Arabic signage visible in distance, specific cookware, clay water pots" ] }, "grain_and_noise": { "description": "Visible film-like grain, especially in shadows and mid-tones.", "details": [ "Fine luminance noise throughout", "Color noise in deep shadow areas", "Slight chromatic aberration at high-contrast edges", "NOT: clinical digital sharpness with zero noise" ] }, "emotion_direction": { "description": "NEVER direct the emotion. Describe the PHYSICAL ACTION instead.", "wrong": "A child with hopeful grateful eyes smiling gently", "right": "A child mid-chew, mouth slightly open, looking sideways at something outside the frame. One hand holds a piece of flatbread, the other rests on the metal plate.", "principle": "Real documentary photos capture people DOING things, not FEELING things for the camera." }, "aspect_ratio": "2:1 wide landscape (matching site card slots at ~1400x700)", "kaffarah_page_slots": { "kaffarah-community.jpg": { "context": "Kaffarah card — feeding 60 people as expiation. Shows communal child feeding.", "current_problem": "Gritty, dirty children in an aid-distribution queue with workers in grey vests. Feels like poverty-tourism NGO stock.", "replacement_brief": "Children eating a communal meal. NOT a handout queue. Kids sitting together, eating, mid-action. Real location with clutter. Muted light." } } }