OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

书籍主角文字云书法排版

Book Protagonists Word Cloud Calligraphy Typography

16:9的2x2书籍主角文字云书法作品：为四个非版权角色分别从微型文字密度、间距、换行、字号和负空间构成远近皆可辨的代表性轮廓。

书籍主角文字云书法排版 — AI image creation
书籍主角文字云书法排版

提示词 / Prompt

中文提示词

创建一幅16:9作品，严格排列为2x2网格，服务于四个非版权书籍主角：一名虚构的公版侦探、一名虚构的公版探险家、一名虚构的公版博物学家和一名虚构的公版漫游者。每个主角对应一个排版书法面板。根据提供的名称推断每个主角的视觉DNA，只选择与其关联的最具代表性的轮廓、物体、场景、生物、建筑、姿态或符号；除非已提供，不得硬编码电影、引用、角色细节、物体、场景、文字或符号。每个画面主要由密集微型文字而非普通绘画构建，并使用与主体相关的主题片段和元数据作为纹理来源。若没有确切文本来源，只能为该主角创造主题片段、元数据、章节式短语、标签、日期、关键词和氛围词。让主体从排版密度、间距、换行、文字大小和负空间中显现，远看保持强烈且可读的轮廓，近看能发现成千上万个微小单词、短语、说明和文字碎片。使用单色墨水、陈旧纸张、档案颗粒和复古印刷纹理；四格风格统一，但每个主体的轮廓和文字节奏可以不同。保持网格平衡，让微型文字看起来像文字而非随机噪点。避免普通绘画插画、硬编码的受版权媒体引用、未提供的引文、无关泛化符号、浑浊轮廓、难读的大字、彩色渐变、现代光亮设计、杂乱、水印和标志。

English prompt

Create a 16:9 artwork arranged as a precisely aligned 2x2 grid for four non-copyrighted book protagonists: a fictional public-domain detective, a fictional public-domain explorer, a fictional public-domain naturalist and a fictional public-domain wanderer. Generate one typographic calligram panel per subject. For each panel, infer the subject's visual DNA from the supplied name and choose only the most iconic silhouette, object, scene, creature, architecture, gesture or symbol associated with that subject; do not hard-code an unprovided film, quotation, character detail, object, scene, text or symbol. Build each image primarily from dense microtext rather than conventional drawing, using subject-specific thematic fragments and metadata as the texture source. If an exact text source is absent, invent only thematic fragments, metadata, chapter-like phrases, labels, dates, keywords and atmospheric wording appropriate to that subject. Make each subject emerge from typographic density, spacing, line breaks, text scale and negative space, with a strong readable silhouette from a distance. Close inspection should reveal thousands of tiny words, phrases, captions and fragments. Use monochrome ink, aged paper, archival grain and vintage printed texture; keep all four panels stylistically coherent while allowing each subject's silhouette and text rhythm to differ. Ensure the grid is balanced and the microtext remains text-like rather than random noise. Avoid normal painted illustration, hard-coded copyrighted media references, supplied-absent quotes, generic unrelated symbols, muddy silhouettes, illegible large text, colorful gradients, modern glossy design, clutter, watermark and logo.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作