Reconstruct and upscale the attached portrait as a higher-resolution version of the same pre-rendered 3D game asset.
This is an image-restoration and detail-reconstruction task, not a painting or stylistic reinterpretation. Preserve the visual language of late-2000s fantasy RPG CGI: clearly modeled 3D geometry, smooth rasterized surfaces, baked textures, controlled material shading, ambient occlusion, directional game lighting, subtle specular highlights, rendered depth, and clean anti-aliased edges.
Keep the character’s exact identity, species, facial expression, pose, silhouette, anatomy, proportions, costume, armor, weapons, accessories, colors, lighting, camera angle, framing, and background composition. Infer missing detail conservatively from the source. Improve facial definition, fabric texture, metal surfaces, hair, fur, scales, and geometric clarity while making everything look native to the original 3D render.
Do not produce an oil painting, digital painting, concept-art illustration, drawing, or hand-painted interpretation. No visible brushstrokes, canvas texture, impasto, watercolor effects, painterly edges, sketch lines, cel shading, photorealism, cinematic redesign, excessive sharpening, smooth AI airbrushing, halos, text, borders, or invented decorative details.
Do not change the pose, expression, costume, proportions, camera angle, composition, or background. Do not expand, crop, rotate, letterbox, or reframe the portrait.
The final image should look as if the original 3D scene, character model, textures, materials, and lighting were rendered again at a much higher resolution using the same engine and art direction—not repainted by an illustrator.
Preserve the input image’s exact 40:51 aspect ratio (equivalent to the original 120:153 proportions). Generate at the highest available native resolution. Return only one finished portrait.
Reconstruct and upscale the attached portrait as a higher-resolution version of the same pre-rendered 3D game asset.
This is an image-restoration and detail-reconstruction task, not a painting or stylistic reinterpretation. Preserve the visual language of late-2000s fantasy RPG CGI: clearly modeled 3D geometry, smooth rasterized surfaces, baked textures, controlled material shading, ambient occlusion, directional game lighting, subtle specular highlights, rendered depth, and clean anti-aliased edges.
Keep the character’s exact identity, species, facial expression, pose, silhouette, anatomy, proportions, costume, armor, weapons, accessories, colors, lighting, camera angle, framing, and background composition. Infer missing detail conservatively from the source. Improve facial definition, fabric texture, metal surfaces, hair, fur, scales, and geometric clarity while making everything look native to the original 3D render.
Do not produce an oil painting, digital painting, concept-art illustration, drawing, or hand-painted interpretation. No visible brushstrokes, canvas texture, impasto, watercolor effects, painterly edges, sketch lines, cel shading, photorealism, cinematic redesign, excessive sharpening, smooth AI airbrushing, halos, text, borders, or invented decorative details.
Do not change the pose, expression, costume, proportions, camera angle, composition, or background. Do not expand, crop, rotate, letterbox, or reframe the portrait.
The final image should look as if the original 3D scene, character model, textures, materials, and lighting were rendered again at a much higher resolution using the same engine and art direction—not repainted by an illustrator.
Preserve the input image’s exact 40:51 aspect ratio (equivalent to the original 120:153 proportions). Generate at the highest available native resolution. Return only one finished portrait.