Diwali Diya Portrait in Warm Light

Use the uploaded [reference image] as the subjects' exact faces, likeness and skin tone. Create a hyper-realistic photograph styled as an authentic late-1980s colour print. Subjects: a couple standing close together on a grassy hillside, posed for a holiday photograph. He stands slightly behind and to her left with one hand in his jeans pocket, chin raised, looking coolly toward the camera. She stands in front of his shoulder with her hands loosely clasped in front of her, head tilted, smiling gently at the lens. Hair: his is dark, thick and worn high with strong volume swept up and back off the forehead in a full period blow-dry. Hers is a big voluminous mane of long chestnut curls with heavy body, falling well past the shoulders and lifting away from the face. Wardrobe: both wear matching oversized nylon shell jackets in bold colour-block panels of teal, magenta and cream, worn open. He has a plain white t-shirt beneath, acid-wash high-waisted jeans and a black digital wristwatch, with dark aviator sunglasses. She has a hot pink top, a wide black leather belt with a large buckle, acid-wash jeans and oversized cream and black hoop earrings. Color Grade (Authentic 1980s C-41 Print): Aged physical photograph aesthetics. The cyan dye layer has faded faster than magenta and yellow, creating a warm amber-pink cast strongest in the corners. Blacks are lifted and milky; shadows sit warm grey-brown with a slight magenta tint. Overall low contrast with soft highlight roll-off and gentle halation glow around bright light sources. Visible silver-halide grain through the midtones and mild edge vignetting. Print Treatment: Authentic aged print look with a soft diagonal crease across one corner, fine white surface scratches, and lightly worn, frayed edges. The color is faded unevenly toward the borders for a nostalgic, family-album feel. Technical: Ultra-detailed, razor-sharp on faces, 50mm lens, shallow depth of field, 9:16 aspect ratio, cinematic film grain, natural skin texture with visible pores.

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 at the centre of a press scrum, head and shoulders filling the middle of the frame, chin level and expression composed and unreadable as cameras fire around him. Wardrobe: a black tailored blazer over a black roll-neck jumper, and black rectangular sunglasses with a thin metal browline. Setting: a crowded press line. A dense ring of handheld microphones on black foam windshields pushes in from the bottom of the frame toward his face, their flags plain matte black and completely unbranded. Behind and around him, out-of-focus photographers hold professional cameras with large lenses, several flashes firing at once. Lighting: hard direct flash from multiple angles, bright specular highlights on the sunglasses and hair, deep falloff into shadow behind the crowd. Camera: chest-up portrait at eye level, 85mm lens, subject tack sharp with the microphones nearest the lens softly blurred in the foreground. Colour: rich full colour, warm amber tungsten light behind the crowd against cool white camera flashes, deep true black wardrobe rather than grey, warm natural skin tones. Mood: intense, cinematic, the moment before someone speaks. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, individually rendered strands, realistic foam windshield texture and lens glass, believable flash falloff, 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 320 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference face] --no black and white, monochrome, greyscale, desaturated, distortion, extra limbs, warped hands, text artifacts, readable network logos, brand names on microphone flags, 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. Subjects: a couple in the front seats of a car at night, shot from the back seat. She sits in the passenger seat turned toward him, cheek resting on her hand with her elbow on the console, looking at him softly. He is at the wheel in a three-quarter view, one hand on the steering wheel, glancing across at her. Wardrobe: she wears a full oversized faux fur coat in warm caramel and honey tones over a dark top. He wears a black wool overcoat over a black shirt. Setting: the interior of a modern car with pale cream leather seats and dark trim. Through the windscreen, a wide spread of out-of-focus city lights fills the background in warm amber and cool white bokeh. Lighting: low ambient night light, warm dashboard glow on their faces from below, cool blue-white spill from the city through the glass, deep shadows throughout. Camera: waist-up two-shot from between the rear seats, 35mm lens, shallow depth of field with the windscreen lights thrown fully out of focus. Colour: rich full colour, warm caramel fur and cream leather against black wool, amber and gold city bokeh with cool blue highlights on the glass, warm natural skin tones under the dashboard glow. Mood: intimate, quiet, cinematic, late-night calm. Rendering: ultra-detailed and razor-sharp on both faces, natural skin texture with visible pores and fine detail, individually rendered fur fibres, realistic leather grain and windscreen reflections, believable circular bokeh, warm cinematic colour grade, fine 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 320 --quality 2 --seed [insert seed] --iw [insert image weight 1-3 for the reference faces] --no black and white, monochrome, greyscale, desaturated, distortion, extra limbs, warped hands, text artifacts, visible logos or brand names, warped steering wheel

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 sitting on the ground of a rooftop car park with his back resting against the rear bumper of a parked car, one knee bent and one leg stretched out, looking up and away from the camera with a calm, unhurried expression. Wardrobe: a vintage colour-block windbreaker with bold chevron panels in black, red and mustard yellow, worn open over a plain white t-shirt. Light-wash straight-leg jeans and black low-top canvas sneakers with white soles. Setting: the open top deck of a concrete multi-storey car park at dusk. A boxy 1980s saloon car fills the frame behind him, its red tail lights lit and glowing. A city skyline and a few distant lights sit far behind, softly out of focus. Lighting: fading blue-hour daylight with a warm glow from the car's tail lights spilling onto the tarmac around him, low contrast, soft and even on the face. Camera: full body shot from a low seated angle, 35mm lens, moderate depth of field so the car stays sharp and the skyline falls away. Colour: rich full colour, deep teal-blue dusk sky against warm red tail lights and the mustard and red of the jacket, cool grey concrete, natural skin tones. Mood: nostalgic, still, quietly cinematic, late-eighties film aesthetic. Rendering: ultra-detailed and razor-sharp, natural skin texture with visible pores and fine detail, realistic nylon windbreaker sheen and denim weave, believable tail-light glow and reflections on wet-looking tarmac, fine film grain, 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, visible logos or brand names, readable number plate, gibberish lettering