Prompt page
Generate Rainy Listening Bar Portraits
Prompt
Reference contract: Use the uploaded image only as the identity anchor for the subject. Preserve the subject’s recognizable likeness, face shape, eye spacing, nose shape, lip shape, skin tone and undertone, hairline, age cues, and body proportions. Change only the wardrobe, setting, and photographic treatment. Render the subject as a naturally present person in the alley, not a pasted reference face. Create a photorealistic front-facing, eye-level, waist-up portrait of the subject outside a hidden listening bar in a rain-wet neon alley, with a narrow doorway, dark wood facade, unreadable vinyl posters, warm amber interior light, and blue-magenta city reflections on the pavement behind. The subject faces the camera directly, relaxed and grounded, with the face fully unobstructed. Wardrobe: a matte black technical jacket layered over a textured knit top, with realistic rain-darkened fabric edges, seam tension, zipper teeth, and small water droplets on the shoulders. Composition: eye-level waist-up framing, subject sharp in the foreground, background receding through layered doorway, reflections, and alley depth. Leave subtle negative space on one side for a clean social crop. Do not force the hands; keep them naturally outside frame unless the model includes relaxed partial hands at the lower edge. Lighting: warm practical light from the bar doorway mixes with cool neon spill from the alley. The warm key light must remain dominant on the face so the subject’s skin tone stays natural, while cool reflections affect only the jacket edges, background, and wet surfaces. Realistic catchlights, soft jaw shadows, collar shadows, and consistent depth of field must connect the face, neck, shoulders, and clothing. Shot on Sony A7R V, 35mm f/1.4 Sigma Art, f/2.0, 1/500s shutter speed, ISO 200, naturalistic editorial street portrait, not movie-poster stylized. Details include wet asphalt reflections, wood grain by the doorway, worn sticker residue with no readable text, fabric weave, water beads, natural pores, fine hair strands catching light, and subtle high-frequency sensor grain. No in-image text, no logos, no watermark, no pasted-on face, no cutout edges, no mismatched face and neck lighting, no warped anatomy, no plastic skin, no heavy rain covering the face, no exaggerated cinematic grading. 4:5 vertical aspect ratio.
Model
ChatGPT




