Image Nano Banana Pro
Fractured Reflections: Cinematic Portrait
A high-fashion close-up with shattered glass framing the subject. Realistic refractions echo the same identity, highlighting dramatic lighting, black editorial clothing, and a deep red background.
Prompt
Transform the reference photo into a high-end cinematic portrait with a shattered glass foreground while preserving the exact original facial features and hairstyle. The subject must remain the same person as the reference image, maintaining identical facial structure, skin tone, and identity without modification or invented elements. The subject wears only black clothing styled in a bold, high-fashion editorial manner, maintaining a dark and elegant aesthetic. The composition is a tight close-up shot captured from a top-down camera angle, focusing closely on the subject's face and upper body. The background is clean, minimal, and a deep solid red color that contrasts with the black clothing and skin tones. Realistic shattered glass fragments are suspended in mid-air around the border of the composition. Each glass shard reflects and refracts fragments of the original reference photo itself, including portions of the subject's face, eyes, skin, clothing, and lighting, creating multiple distorted reflections of the same subject across the broken glass. The reflections must be optically accurate with realistic refraction, mirrored angles, light bending, depth variation, and sharp glass edges. The subject behind the glass remains sharp, powerful, and emotionally intense, while reflections in the shards echo the same identity from different angles. Lighting is cinematic and dramatic with controlled highlights on the glass, soft shadows on the subject, and film-grade color grading. Style: cinematic movie poster, psychological drama, premium photography, hyperrealism, ultra-detailed, professional lighting, sharp focus, 4K resolution, no distortion. Aspect Ratio: 3:4 portrait. Image quality: ultra-high resolution, photorealistic editorial finish.
Published: April 19, 2026 by Jarvis AI