肖像画 — 白黒 — 彼らはタバコを持っています

Use the uploaded photo as the ONLY identity reference. Automatically create an ultra-realistic, cinematic double-portrait composition using the uploaded person's face and identity.

FORMAT: Vertical 9:16 portrait, premium dark aesthetic.

MAIN PORTRAIT:
Create a large, close-up portrait of the uploaded person, occupying most of the frame. Give them thick, voluminous, naturally styled dark hair, stylish rectangular eyeglasses, subtle earrings, and a black oversized jacket. They hold a cigarette casually near their lips, with delicate smoke drifting through the air. Their expression is confident, calm, and slightly mysterious, with their gaze directed toward the camera.

Preserve the uploaded person's recognizable facial features, natural skin tone, face shape, and identity. Adapt the hairstyle and accessories naturally to match the reference aesthetic.

CENTER OVERLAY:
In the lower center, add a smaller, highly realistic black-and-white portrait of the SAME person, shown shirtless with a muscular, athletic upper body, looking upward with a serious expression. Give this smaller portrait a bright, soft white glowing outline that separates it from the background.

BACKGROUND:
Use a softly blurred, warm beige and gray indoor background with cinematic shadows and subtle atmospheric smoke. Keep the main portrait in natural warm colors and the smaller muscular portrait in monochrome.

COMPOSITION:
The large portrait must dominate the image. The smaller grayscale portrait overlaps the chest area in the bottom center, maintaining the same layered double-exposure aesthetic as the reference.

STYLE:
Photorealistic face, natural skin texture, realistic hands, detailed hair, cinematic editorial photography, soft dramatic lighting, high contrast, subtle film grain, premium moody atmosphere, sharp facial details.

IMPORTANT:
Use the uploaded person's identity consistently in both portraits. The smaller portrait must clearly look like the same person as the main portrait. Preserve the reference's overall framing, pose, lighting, and layered composition.

No text, no watermark, no logos, no extra people. Automatically generate the complete image using only the uploaded photo, without requiring additional instructions.

プロンプト全体をコピーし、必要に応じて独自の参考写真を使用してください。