Criação de Personagem Realista com Divisão em Três Painéis

0,0 (0 votos)
Grátis Influenciadora de IA
Voltar
OBJETIVO
Gerar um quadro visual detalhado e consistente do seu personagem a partir de uma imagem.
ONDE USAR
Nano Banana 2, GPT Image 2 ou outras ferramentas de imagem
PROMPT PRINCIPAL
Based on the attached reference image, generate a three-panel character reference sheet. The composition is a horizontal layout, equally divided into three side-by-side vertical panels, with clean, thin dividing lines between panels. Strictly maintain the character's facial features, hairstyle and hair color, body proportions, clothing style, and accessories exactly as in the reference image; all three panels feature the same character in the same outfit.

Left panel — Front full-body, headless. The character's body faces the camera directly, from the shoulders down to the feet, with both arms naturally hanging at the sides, hands relaxed and open, weight evenly distributed on both feet. No head, no hair—nothing protrudes above the shoulder line, no hair draping over the chest or shoulders. The neck ends at the base of the throat with a clean, flat, sharply edged horizontal cut-off, like a headless mannequin, with a crisp sculptural edge—no blurring, no softening, no fading, no smoke, no afterimage, no transparency, no blood, no anatomical details; above the cut-off is only blank background. This panel retains full headroom space, with ample blank area above the shoulders to keep the character's proportions and positioning in the frame consistent with a normal full-body portrait.

Center panel — Back full-body, with head. The same character is photographed straight from behind, standing upright, with hair naturally falling from the back, back of the clothing construction, skirt hem/pant shape, and footwear clearly visible, both arms naturally hanging at the sides, hands relaxed, weight evenly distributed on both feet, fully framed from the top of the head to the shoes.

Right panel — Bust close-up, identity lock. The same character in a tight composition from slightly above the head down to the collarbone, showing only the very top edge of the clothing, with the face occupying most of the frame—a true close-up shot. The body faces the camera directly, head level, eyes staring straight at the lens, lips closed and relaxed, with a neutral, restrained expression. Eyebrows, eyes, eyelashes, lip texture, and key identifying features are clearly discernible at close range.

Unified background for all three panels: Flat, even 18% middle gray seamless background, single color value, no gradients or falloff (flat 18% gray seamless). Shadowless frontal illumination with no direction and no shadows, even and consistent fill light around all sides, no key light side, no shadow side, no rim light, no edge light (shadowless frontal illumination); no projections on the background, no contact shadow under the feet (zero cast shadow); lighting completely consistent across all three panels.

Photorealistic texture (photorealistic), skin showing fine, even pores and subsurface scattering, individual strands of hair distinctly visible, fabric weave clear (fabric weave), subtle film grain. Real human, real camera capture—absolutely no plastic feel, no rendering, no CGI, no AI skin smoothing, no excessive flattening.

Guia de Implementação

Para criar um quadro visual realista do seu personagem, siga os passos abaixo:

  • Escolha uma ferramenta de criação de imagem: Use plataformas como Nano Banana 2, GPT Image 2 ou qualquer outra de sua preferência.
  • Anexe a imagem do seu personagem: Carregue a imagem de referência que deseja usar.
  • Configure as dimensões da imagem: Defina o layout para 16:9 para garantir a melhor proporção e qualidade.
  • Inicie a geração da imagem: Cole o prompt fornecido na ferramenta escolhida. Certifique-se de seguir os requisitos descritos para cada painel da imagem.
  • Finalize e baixa a imagem: Assim que a imagem for gerada, faça o download na melhor resolução possível para garantir qualidade visual e consistência.

Não perca esses prompts em ação

Eu aplico esses prompts na prática em vídeos rápidos, mostrando o resultado real e ajustes que não cabem no texto.

Ver no YouTube