角色觉醒姿态全景构图预设

UserBigcatgj
原作者: てんねん
原文链接
2026/8/15
Nano Banana
故事分镜角色设定二次元动漫

效果展示

图片案例

示例图片 1

提示词内容

【覚醒ポーズプロンプト】
subject: >
  Use the same character as reference A.
  Preserve all facial features, hairstyle, clothing, body proportions, colors,
  line weight, shading, and art style exactly as in the reference.
composition: >
  IMPORTANT: Change the framing from the reference.
  The camera must be positioned **much farther away** so the character’s entire body
  is fully visible **with extra space above, below, and around** the character.
  The character must appear **significantly smaller** than in reference A,
  occupying only about **20–30% of the frame height**.
  Place the character exactly at the center of the frame with wide open space
  surrounding them for later energy effects.
  The field of view should be noticeably wider than the reference.
  The final image must be rendered in a **16:9 aspect ratio**.
action: >
  Set the character in the awakening stance shown in reference B:
    - feet shoulder-width apart
    - both hands tightly clenched into fists
    - arms slightly bent and tense
    - chest slightly forward, posture firm and braced
    - intense, focused expression
location: >
  Reuse the same background environment as reference A.
  Expand the visible portion of the background **significantly**
  to match the widened, zoomed-out framing.
  Maintain the same textures, perspective, colors, and drawing style.
style: >
  Match the exact style of reference A:
  identical linework, shading, coloring, and background rendering.
camera_lighting: >
  Use the same lighting conditions as in reference A.
edit_instructions: >
  Recreate the character and background from reference A exactly.
  BUT modify the camera distance so the entire character fits in the frame
  with **substantial space** around them.
  The camera should be pulled back enough that the character appears small
  within a much wider environment.
  Change only:
    - the camera distance (strong zoom out)
    - the char

想要不同效果?只改几处即可

AI 自动找出可替换的部分(主体/风格/场景),填入想要的值,其余保持不变

提示词广场登录