Transferencia de estilo piedra fragmentada
Transforma cualquier sujeto en una escultura 3D hiperrealista de fragmentos de pizarra y arenisca sobre un fondo minimalista de estudio.
Los modelos económicos no renderizan tipografía de forma fiable; este caso necesita un nivel superior.
Imagen generada por nosotros con gpt-image-2.
Prompt
Perform an image-to-image style transfer using [INPUT_IMAGE] as the source. Recreate the main subject from that image as a hyper-real 3D studio composition built entirely from fragmented stone. The subject must be assembled from separate, clearly detached rock pieces with small visible gaps between shards, with no pieces merging together. Material look: fragmented slate and sandstone shards with chiseled edges, crisp fractures, visible stone grain, micro-scratches, and realistic surface roughness. Color palette: predominantly dark slate tones with subtle warm-ochre sandstone accents. Lighting: soft studio key light from the top-left, gentle fill light, subtle contact shadows beneath each shard, realistic ambient occlusion in the crevices, and minimal, clean reflections. Background: a minimal off-white seamless backdrop with no texture. Framing: centered, clean, straight-on composition with the subject fully readable. Add a few tiny debris chips floating or resting near the base for added depth. Preserve the subject's identity, proportions, and recognizable silhouette from the input image while transforming all surfaces into stone fragments. Render in hyper-real style with high detail, sharp focus, and 8k quality. Avoid: text, typography, logos, watermarks, signatures, extra props, busy backgrounds, fog, heavy bloom, cartoon or illustration style, low-poly geometry, plastic or metal materials, glossy paint, melted shapes, merged fragments, an unreadable subject, blur, noise, low resolution, oversharpening halos, distorted face or body, extra limbs, or deformed geometry. Output should use the input image's aspect ratio, an off-white background, a straight-on camera angle, centered framing, and medium camera distance. Technical parameters: style strength 0.75, identity preservation 0.9, very high detail level, soft top-left studio lighting preset, subtle shadow intensity, clearly visible gaps between shards, a few tiny debris chips, and no text in the final image.
Queda 1 hueco sin rellenar. Si copias así, el generador dibujará los corchetes tal cual dentro de la imagen.
Variables editables
Cambia cualquier valor y el prompt se reescribe arriba. Copia el resultado ya listo.
La imagen de origen proporcionada por el usuario cuyo sujeto principal se transformará en una escultura de piedra fragmentada.
Genera la imagen aquí
Con tu propia clave de OpenAI. Gratis para nosotros, lo que gastes es tuyo.
Se guarda solo en este navegador y va directa a la API de OpenAI: no pasa por ningún servidor nuestro, porque no hay ninguno en medio. Puedes comprobarlo en la pestaña de red del inspector.
Rellena las variables que quedan antes de generar: si no, el modelo dibujará los corchetes.
