Capture a candid photograph of two attractive adults in their late 20s at the Location. Style Person 1 with the Person 1 Hair, Person 1 Style and Person 1 Outfit, and style Person 2 with the Person 2 Hair, Person 2 Style and Person 2 Outfit. Show them naturally engaged in the Activity, with relaxed expressions, believable body language and an unplanned moment rather than a posed portrait. Use the Color Palette throughout the clothing and gallery environment, with the backless dress remaining the visual focal point. Light the scene according to the Lighting, preserving realistic exposure, natural skin texture, individual hair flyaways and authentic reflections in the wine glasses. Compose it as a flattering waist-up to three-quarter vertical frame with artwork, guests and gallery architecture softly visible in the background. Use the Camera Style and Format. Make it look like a real photo taken at the event, not a studio-perfect or AI-smooth render. No text, logos, watermarks, warped hands or artificial background blur.
Use these details:
Person 1 Hair: "Long auburn hair in loose, softly tousled waves"
Person 1 Style: "Glamorous, confident evening look with fresh youthful makeup"
Person 1 Outfit: "Backless floor-length evening dress with delicate strappy details and elegant jewelry"
Person 2 Hair: "Short, naturally textured dark hair with a relaxed side part"
Person 2 Style: "Elegant, approachable evening style with understated grooming"
Person 2 Outfit: "Tailored black dinner jacket, open-collar shirt and fitted trousers"
Location: "Contemporary art gallery during an evening exhibition opening"
Activity: "Standing close together while holding wine glasses and reacting to something nearby"
Lighting: "Soft gallery lighting mixed with warm highlights from the event space"
Camera Style: "Candid handheld smartphone photo with natural perspective and subtle low-light grain"
Color Palette: "Emerald green, auburn, black, warm white and muted gallery neutrals"
Format: "4:5 vertical portrait"
Using your own logo, product or face? Attach it in the chat and say which of the details it replaces.
Your result will come out close to the example rather than identical to it. Change the details to make it yours.
