Rooftop Dusk — Retro Jacket Car Portrait

Use the uploaded [reference image] as the subject's exact face, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour. Subject: a man standing in the centre of a large institutional dormitory, facing the camera straight on from the chest up, arms at his sides, expression tense and resolute. Other figures in matching tracksuits sit and stand on bunk platforms behind him, all softly out of focus. Wardrobe: a plain teal-green zip-up tracksuit jacket with a single white stripe running down each sleeve, worn over a plain white crew-neck t-shirt, with matching teal-green tracksuit trousers. No numbers, no badges, no lettering anywhere on the clothing. Setting: a vast concrete hall lined with stacked steel bunk beds rising several levels up the wall behind him. Bare white walls, a single square ceiling light casting a hard pool of light, institutional and cold. Lighting: harsh overhead fluorescent light, cool and clinical, slightly greenish, with soft falloff into shadow at the edges of the frame. Camera: chest-up portrait at eye level, 50mm lens, subject tack sharp, the bunks and figures behind falling into shallow blur. Colour: desaturated cool palette, teal-green tracksuits against grey concrete and pale institutional walls, muted skin tones under fluorescent light. Mood: tense, uneasy, cinematic thriller still. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic cotton jersey and tracksuit fabric, believable fluorescent colour cast, fine film grain, cinematic colour grade, no distortion, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 320 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference face] --no distortion, extra limbs, warped hands, text artifacts, visible logos or brand names, numbers on clothing, badges, gibberish lettering, recognisable celebrity faces

Use the uploaded [reference image] as the subject's exact face, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour. Subject: a man standing alone on a wide landing, photographed from directly above and slightly behind, his head tilted back so he looks straight up into the lens. Arms hang loose at his sides, expression wary and searching. Wardrobe: a plain teal-green zip-up tracksuit with a white stripe over each shoulder, a white crew-neck t-shirt beneath, matching teal trousers and plain white slip-on canvas shoes. No numbers, no badges, no lettering anywhere on the clothing. Setting: an enclosed stairwell painted in flat saturated colour blocks — mint green, mustard yellow, coral pink and soft blue — with staircases running off at conflicting angles. The floor is dark grey studded rubber with a regular pattern of raised circles. Lighting: flat even artificial light with no strong shadows, so the colour blocks read as clean solid planes. Camera: full body shot from a steep overhead angle looking straight down, 35mm lens, slight wide-angle distortion at the frame edges, everything in focus. Colour: vivid saturated full colour, flat pastel-bright wall panels against the deep teal tracksuit and dark rubber floor, natural skin tones. Mood: surreal, disorienting, graphic, cinematic. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic tracksuit fabric and rubber floor texture, crisp flat colour planes with no banding, cinematic colour grade, no distortion of the face, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 350 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference face] --no distortion, extra limbs, warped hands, text artifacts, visible logos or brand names, numbers on clothing, badges, gibberish lettering, recognisable celebrity faces

Use the uploaded [reference image] as the subject's exact face, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour. Subject: a woman lying on her back on a floor covered in open newspaper pages, laughing with her head tilted and her eyes crinkled shut. Both hands are raised to her face, fingertips lightly touching the frames of her sunglasses. Wardrobe: a bright red ribbed turtleneck sweater with long sleeves pushed toward the wrists, and oversized heart-shaped sunglasses with translucent red-tinted lenses and thin gold frames. Small gold stud earrings. Setting: layered sheets of black-and-white newsprint spread flat beneath her, columns of type and small photographs visible. Lighting: soft, even, bright daylight from above, no harsh shadows, clean and cheerful. Camera: directly overhead flat-lay portrait looking straight down, 50mm lens, subject centred, whole frame in focus. Colour: vivid full colour, saturated scarlet knit and red lenses against cool grey-and-white newsprint, warm golden hair, natural glowing skin tones. Mood: joyful, playful, spontaneous, full of energy. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, individually rendered curls, realistic ribbed-knit weave and printed newsprint texture, true-to-life colour grade, no distortion, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 300 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference face] --no distortion, extra limbs, warped hands, text artifacts, cleavage, revealing necklines, readable headlines, gibberish lettering

Use the uploaded [reference image] as the subjects' exact faces, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour, styled as a 1950s street photograph. Subjects: a couple walking toward the camera along a wet city pavement. She is a step ahead, head thrown back mid-laugh, one hand lifting the side of her skirt as she strides. He walks just behind at her shoulder, grinning at her, a vintage twin-lens reflex camera held in one hand. Wardrobe: she wears a sapphire-blue satin tea-length dress with a fitted waist, a modest wrapped portrait neckline that sits on the shoulders, and a full pleated skirt; with a wide-brimmed navy hat, a diamond collar necklace, short white gloves and navy heels. He wears a brown herringbone tweed three-piece suit, a white shirt and a dark tie. Setting: a 1950s London street after rain, the pavement glossy with reflections. A red telephone box at the left edge, a black classic saloon car parked at the kerb, cast-iron street lamps, and Big Ben rising through grey mist in the background. Lighting: soft, flat, overcast daylight with a faint warm glow from the street lamps, damp atmosphere, gentle haze. Camera: full body shot at chest height, 50mm lens, subjects sharp, background softly falling off. Colour: rich full colour with a period grade — saturated sapphire blue and pillar-box red against muted greys, browns and wet-stone tones. Mood: joyful, cinematic, nostalgic, mid-century romance. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic satin sheen and tweed weave, believable wet-pavement reflections, fine period film grain, no distortion, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 350 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference faces] --no distortion, extra limbs, warped hands, text artifacts, cleavage, revealing necklines, modern cars, modern clothing, gibberish signage

Use the uploaded [reference image] as the subjects' exact faces, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour. Subjects: a couple mid-waltz on a polished parquet floor, caught in the middle of a turn. His right hand rests at her back, their other hands clasped and raised; they look into each other's eyes, both smiling softly. Her skirt sweeps outward with the movement. Wardrobe: she wears a floor-length scarlet satin ballgown with a fitted bodice, a modest straight neckline, narrow shoulder straps and an enormous full circle skirt. Drop earrings, hair in a low chignon with soft tendrils at the nape. He wears a classic black tuxedo with satin lapels, a white dress shirt and a black bow tie. Setting: a grand historic ballroom with gilded panelling, tall arched doorways and warm gold walls receding into soft darkness. An enormous crystal chandelier hangs directly above them, dozens of candle-shaped lamps glowing. Lighting: warm amber candlelight falling from the chandelier above, a soft pool of light around the couple, the far end of the room dropping into shadow. Camera: full body shot from slightly below eye level, 35mm lens, chandelier included in the top of the frame, moderate depth of field. Colour: rich full colour, deep scarlet satin against warm gold and amber architecture, black tuxedo, glinting crystal, warm candlelit skin tones. Mood: cinematic, sweeping, romantic, timeless grandeur. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic satin sheen and fabric drape in motion, individually rendered crystal droplets, warm cinematic colour grade, no distortion, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 350 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference faces] --no distortion, extra limbs, warped hands, text artifacts, cleavage, revealing necklines, blurry faces, duplicated chandeliers

Use the uploaded [reference image] as the subject's exact face, likeness and skin tone. Create a hyper-realistic, photorealistic image in full colour. Subject: a woman standing at the polished chrome stern railing of a moving yacht, body turned to the side, one hand resting on the rail behind her, looking out toward the camera with a relaxed, self-possessed expression. Her hair lifts in the wind. Wardrobe: a crisp white linen shirt with the collar open and sleeves rolled to the forearm, worn loose over matching wide-leg white linen trousers. Slim black rectangular sunglasses, a fine gold chain necklace and a thin gold bracelet. Hair: long, dark, almost black, blown back and to one side by the sea breeze. Setting: deep blue open sea with a white churning wake trailing behind the boat, hazy blue-grey mountains along the horizon, clear bright sky. Lighting: strong natural midday sunlight, crisp and clean, bright reflections scattered across the water. Camera: three-quarter body shot from just below eye level, 35mm lens, deep depth of field so the wake and distant mountains stay legible. Colour: vivid full colour, deep sapphire and turquoise sea, brilliant white linen, warm sunlit skin tones, pale blue sky. Mood: effortless, free, luxury summer travel. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic linen weave and creasing, believable water spray and foam, true-to-life daylight colour grade, no distortion, no extra fingers, correct anatomy. Add a small, subtle, semi-transparent "promptybox.com" watermark in the bottom-right corner. Aspect ratio 9:16. Optimised for Gemini and ChatGPT / DALL·E (also works in Midjourney). Parameters: --ar 9:16 --style raw --v 6 --stylize 300 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference face] --no distortion, extra limbs, warped hands, text artifacts, cleavage, revealing necklines, visible logos or brand names