又是一个有趣的城市美食提示词。GPT Image 2 处理这类任务非常出色。

GPT-Image-2 · @Gdgtify · Sat May 16

在画廊中打开

分享链接 · https://openimages.ajiang.me/p/2055656013060010043/

又是一个有趣的城市美食提示词。GPT Image 2 处理这类任务非常出色。 <instructions> 为意大利执行此操作 Input = Country 扮演世界级的美食造…
camera_lens · macrocamera_lens · tilt-shiftcolor_palette · warm-tonescomposition · multi-panellighting · high-keyquality_flags · embedded-text

提示词(原文,逐字保留)

Another fun food prompt for cities. GPT Image 2 does a very good job with these. <instructions> do this for Italy Input = Country Act as a world-class food stylist and geography expert. 1. For the Input Country, identify 4 cultural hubs or less-famous cities (avoiding the obvious capital if possible). 2. For each city, identify a specific Iconic Landmark or Street Scene . 3. Identify the Signature Ingredients or Famous Dish specific to that city's region. Function Draw ($ City, $ Landmark, $ Ingredients) Anchor: [Miniature $ Landmark] :: [$Ingredients]::4 Morphology: Edible architectural diorama of [$Landmark], entire landscape and structures constructed via the assemblage of [$Ingredients], masonry and terrain formed from the constituent parts of the local cuisine (e.g., stacked items for walls, liquids for rivers, powders for ground), isometric food styling on a clean serving tray::3 Material Physics: Physical properties of [$Ingredients] applied to architecture, organic food textures replacing concrete and steel (e.g., porous bread, viscous sauce, fibrous vegetables), appetizing gloss and freshness::3 Illumination: High-key commercial food lighting, warm backlighting to show steam or translucency, soft shadows, mouth-watering aesthetic::2 Render Stack: Macro food photography, tilt-shift lens, f/2.8, 8k, isometric view, Michelin star plating aesthetic, Unreal Engine 5 render::1 Negative: [Inedible materials, plastic, clay, raw meat (unless appropriate), human hands, cutlery, text, labels, blurry, dirty tray]:: -1 Output: 2x2 grid. 16:9 For each panel, execute the Draw Function for one of the cities. Name of the city and the dish underneath. </instructions>

中文译文

又是一个有趣的城市美食提示词。GPT Image 2 处理这类任务非常出色。 <instructions> 为意大利执行此操作 Input = Country 扮演世界级的美食造型师和地理专家。1. 针对输入国家,确定 4 个 文化中心或不太出名的城市 (如果可能的话,避免明显的首都)。2. 针对每个城市,确定一个具体的 地标建筑 或 街景。3. 确定该城市地区特有的 招牌食材 或 名菜。Function Draw ($ City, $ Landmark, $ Ingredients) Anchor: [Miniature $ Landmark] :: [$Ingredients]::4 Morphology: [$Landmark] 的可食用建筑立体模型,整个景观和结构通过 [$Ingredients] 的组合构建,砌体和地形由当地美食的组成部分构成(例如,堆叠的物品作为墙壁,液体作为河流,粉末作为地面),等距美食造型,置于干净的上菜托盘上::3 Material Physics: [$Ingredients] 的物理特性应用于建筑,有机食物纹理取代混凝土和钢铁(例如,多孔面包、粘稠酱汁、纤维蔬菜),诱人的光泽和新鲜感::3 Illumination: 高调商业美食照明,暖色背光以显示蒸汽或半透明感,柔和阴影,令人垂涎的美学::2 Render Stack: 微距美食摄影,移轴镜头,f/2.8,8k,等距视图,米其林星级摆盘美学,Unreal Engine 5 渲染::1 Negative: [不可食用材料、塑料、黏土、生肉(除非合适)、人手、餐具、文字、标签、模糊、肮脏的托盘]:: -1 Output: 2x2 网格。16:9 每个面板对其中一个城市执行 Draw Function。城市名称和菜肴名称标注在下方。</instructions>

来源与署名

原文由 @Gdgtify 发布在 X:查看原推文。 本页逐字保留原文并提供机器翻译的中文解读;版权归原作者所有。

在画廊中浏览

同工具的更多提示词