Dark-Fantasy Guangzhou City Poster
海報與插畫

Dark-Fantasy Guangzhou City Poster

平面插畫,東方幻想風格高階城市海報設計,豎版9:16構圖,整體採用對角線+S型流動構圖,從左下向右上延展,畫面以深邃黑色為背景,自上而下漸變至濃烈暗紅色,形成強烈冷暖對比與空間縱深,背景帶微弱星塵與顆粒質感。畫面中央一條金色流動能量線條如火焰般蜿蜒貫穿,自底部向上延伸,具有流體質感、粒子光效與漸變高光,區域性帶細微能量碎屑與體積光。 金色流光中逐層浮現廣州城市地標建築群:廣州塔為視覺核心,比例突出,周圍融合珠江新城高樓群、獵德大橋及現…

提示詞

平面插畫,東方幻想風格高階城市海報設計,豎版9:16構圖,整體採用對角線+S型流動構圖,從左下向右上延展,畫面以深邃黑色為背景,自上而下漸變至濃烈暗紅色,形成強烈冷暖對比與空間縱深,背景帶微弱星塵與顆粒質感。畫面中央一條金色流動能量線條如火焰般蜿蜒貫穿,自底部向上延伸,具有流體質感、粒子光效與漸變高光,區域性帶細微能量碎屑與體積光。 金色流光中逐層浮現廣州城市地標建築群:廣州塔為視覺核心,比例突出,周圍融合珠江新城高樓群、獵德大橋及現代與嶺南建築元素,建築採用“精細線描 + 金色發光體塊”表現,輪廓清晰、細節豐富,在金色光暈映襯下彷彿懸浮於虛空,形成超現實空間層次,遠景輕微霧化增強縱深感。 畫面底部為一位東方白髮女性形象,長髮飄逸,如煙似霧,與金色流光自然銜接並逐漸融合,髮絲半透明帶漸變光感,姿態柔美,雙目微閉,神情寧靜,懷抱一束多彩鮮花,花間點綴微光粒子與星點效果,象徵人與城市能量的精神連線,人物細節適度簡化以突出整體設計感。 光影集中於金色流線、建築與人物輪廓,形成強烈明暗對比與視覺聚焦,整體氛圍宏大、神秘、具有東方神話意境且略帶治癒感。色彩以黑與暗紅為基底,高亮鎏金為主視覺強調,金色具備豐富明暗層次,輔以小面積高飽和花束色彩點綴,整體高階剋制。 頁面文字與畫面融合排版:頂部居中宋體大字“廣州·中國”,下方小字“2026/04/20”,再下方小字“LIYUE”,文字採用淡金色或柔和暖白色,與整體光影統一。高品質細節,電影級光影表現,體積光與粒子細節豐富,畫面乾淨無噪點,超高畫質8K解析度,商業級海報質感。

How to use this prompt

1
Review the full prompt

Read the complete Dark-Fantasy Guangzhou City Poster prompt and identify the subject, style, camera, lighting, and composition requirements before generating.

2
Customize variables

Replace bracketed or argument-style placeholders with your product, character, brand, scene, color palette, or aspect ratio requirements.

3
Generate and iterate

Open https://www.gptimagehub.com/zh-TW/generate?promptId=cmogwtuu5003hzpdybo2xe7ua, generate the image, then refine the prompt with more specific subject, text, layout, or negative constraints if needed.

Prompt FAQ

What is the Dark-Fantasy Guangzhou City Poster prompt best used for?

This prompt is best used for 海報與插畫 images where you want a reusable structure, detailed visual direction, and consistent output quality.

Can I edit the prompt before generating?

Yes. The full prompt text is visible on this page so you can change subjects, product names, colors, composition, camera terms, aspect ratio, and style notes before generation.

Which model should I use with this prompt?

Use the model shown in the prompt metadata as the default starting point. If another image model supports the same aspect ratio and instruction style, you can adapt the prompt and compare results.

其他模板