Core technique
The prompt combines a tightly controlled three-quarter side-profile composition with a shallow-depth-of-field 85mm-style portrait treatment. It establishes a clear visual hierarchy: the face remains dominant, the hat is fully visible, the veil is readable, and the collar anchors the lower frame. Lighting from the front-right side reinforces the facial structure while allowing the veil to produce patterned shadows.
Recommended base models
Midjourney — well suited to the overall couture styling, composition, dramatic hat, and vintage editorial atmosphere, though fine veil transparency may need iteration. GPT-Image-1 — strong choice when preserving detailed relationships between the face, veil, hat, collar, and jewelry is important. FLUX — useful for photorealistic fabric, skin, hair, and jewelry detail; the intricate veil instructions may benefit from careful prompting. SDXL-based tools — flexible for controlled photorealism and negative prompting, particularly when using models or workflows tuned for faces and fashion photography.
Suggested negative prompt
cartoon, anime, illustration, painting, CGI, 3D render, doll face, plastic skin, excessive airbrushing, over-smoothed skin, beauty filter, distorted face, asymmetrical eyes, deformed eyes, extra eyes, malformed nose, distorted lips, bad teeth, exaggerated eyelashes, malformed ears, duplicate earrings, extra jewelry, deformed anatomy, unrealistic neck, warped hat, cropped hat, broken veil, opaque veil, incorrect veil mesh, oversized mesh, blurry face, low resolution, excessive noise, oversharpening, harsh flash, extreme shadows, neon colors, busy background, furniture, scenery, modern objects, text, logo, watermark, frame, distorted clothing, synthetic fabric, unnatural hair, flyaway hair, messy composition