Галерея ИИ-промптов OpenZhenZhen · gpt-image-2.5 · image

Диалог сквозь века на улицах Токио

A Century-Spanning Time Dialogue on the Streets of Tokyo

«Диалог сквозь века на улицах Токио» — работа с использованием gpt-image-2.5 на OpenZhenZhen. Изучите исходный промпт и пример генерации для собственных проектов.

Диалог сквозь века на улицах Токио — AI image creation
Диалог сквозь века на улицах Токио

Промпт

Промпт на китайском

身份依赖：执行时提供identity_reference，并使用固定参考ref-10a-v1对应出现在两个时代中的成年女性；不声称来源远程封面图已附加或检查。创作一幅电影感横向宽景城市画面：Tokyo (东京)（默认：“Tokyo (东京)”）在左侧的1920s (20世纪20年代)（默认：“1920s (20世纪20年代)”）与右侧的present day (当代)（默认：“present day (当代)”）之间自然融合。依据来源提供的历史视觉细节，左侧呈现木质町屋店铺、瓦屋顶、狭窄巷道、人力车、老式有轨电车和温暖纸灯笼，成年行人穿着符合时代的服装，整体采用克制的棕褐与琥珀色光线。右侧展现同一城市的玻璃钢结构高楼、霓虹店面、高架铁路、LED广告牌、当代咖啡馆、自行车、出租车和紧凑型汽车，以冷青和洋红城市灯光照亮。潮湿路面反射两侧的色彩。在画面中央避免生硬分割：让门廊、轨道和光带从旧时代有机演变为现代形态。将identity_reference中的同一位成年女性表现为一次跨时代相遇：左侧她穿着得体的20世纪20年代服装，右侧的当代版本穿现代街头服饰；两人在过渡区域附近相遇，好奇地彼此注视，并交换一件简单日常物品，例如折叠报纸。保持两个版本的身份一致，同时清楚区分衣着与时代。采用平视电影式横向广角构图、层次丰富的城市纵深和轻微氛围雾，重点细节聚焦于人物互动。把来源列出的建筑与街景作为场景依据，不添加无依据地标，也不宣称是档案级复原。招牌可以是图形，但不得出现乱码或虚构可读文案。不增加其他前景人物、标志或水印，也不承诺特定画幅比例、相机型号或分辨率。

Промпт на английском

Identity dependency: provide identity_reference at execution and use fixed reference ref-10a-v1 for the clearly adult woman appearing in both eras; the remote source cover is not claimed attached or inspected. Create a cinematic wide horizontal scene of Tokyo (东京) (default: "Tokyo (东京)"), blending 1920s (20世纪20年代) (default: "1920s (20世纪20年代)") on the left into present day (当代) (default: "present day (当代)") on the right. Use the supplied historical concept details: the left side has wooden machiya shopfronts, tiled roofs, narrow lanes, rickshaws, an old streetcar and warm paper lanterns, with period-appropriate adult clothing and subdued sepia/amber light. The right side shows the same city with glass-and-steel towers, neon storefronts, an elevated rail line, LED billboards, a contemporary café, bicycles, taxis and compact cars, lit by cool cyan and magenta city glow. A wet street reflects both palettes. At the center, avoid a hard dividing seam: allow a doorway, rail line and bands of light to transform organically from old to new. Show the same clearly adult woman from identity_reference twice as a time-travel encounter: her left-side 1920s appearance wears modest period-appropriate clothing, while her right-side present-day appearance wears contemporary street clothes; they meet near the transition, look at one another with curiosity, and exchange a small everyday object such as a folded newspaper. Keep both versions recognizably the same identity while clearly distinguishing clothing and era. Use a cinematic eye-level wide composition, layered urban depth, atmospheric mist and selective detail focused around their interaction. Treat the supplied architectural and street details as the scene basis; do not add unsupported landmarks or claim archival reconstruction. Signs may be graphic but must not contain gibberish or invented readable copy. No extra foreground characters, logos, watermark, exact aspect-ratio, camera-model or resolution guarantee.

Открыть модель API

Больше промптов для изображений и видео · Создать изображение · Создать видео · Создать на ИИ-холсте