
This is a luminous polished 2.5D semi-realistic anime fantasy game key visual with delicate character design, graceful and confident linework, smooth cel-inspired shading, and softly painted color gradients. The rendering preserves an elegant anime illustration foundation while incorporating more pronounced semi-realistic CG modeling, softly sculpted facial planes, refined material definition, and coherent dimensional depth. Use smooth three-dimensional color modeling to give the face, hair, fabric, ornaments, and water a clear but restrained sense of volume. Define the facial contour, nose bridge, lips, eye sockets, individual hair masses, layered fabric, and metallic ornaments through subtle tonal transitions and physically coherent highlights, while avoiding an overly photorealistic or fully rendered 3D appearance. Keep shadows open, transparent, and chromatic rather than dark gray or muddy blue. Use a gradual highlight roll-off and reserve near-white values for only a few small focal accents. Control the intensity and bloom of halos, magical symbols, reflective metal, water highlights, and backlighting so that their internal shapes remain clearly visible. Preserve readable detail in both illuminated and shaded areas, with smooth transitions between the bright character and the environment. Dolia_HZJR, an elegant bust-length fantasy poster inside a luminous underwater palace. She rests one hand against a vast shell-shaped window and looks quietly toward the ocean beyond, her profile illuminated by drifting schools of glowing fish. Her other hand holds a small flower-shaped pearl close to her chest. Flowing curtains, suspended bubbles and delicate coral silhouettes frame her without obscuring the character. Side profile, contemplative expression, asymmetrical composition, soft cyan window light, warm pearl highlights, atmospheric depth, refined fantasy game illustration, polished 2.5D anime rendering, clean controlled details, no grain
Parameters used to generate this content
The node graph used to generate this content — pan and zoom to explore, or download it to run in ComfyUI.
AI models used to generate this content