Prompt 正文
默认展示英文原文。点击“用它生成”后，首页会默认载入此草稿。
Using REFERENCE_0 as the ancient-art composition source, transform it into a believable, collectible fine-art photograph as if the scene had been captured by a top photographer in the same historical era. Do not simply render the painting style as a photo; instead reverse-engineer the painting as a compressed memory of a real landscape and reconstruct the real scene behind it.
Keep the core subject logic: exactly {argument name="rider count" default="3"} ancient women riding exactly {argument name="horse count" default="3"} pale horses, moving together across the frame, with pink-to-coral garments and traditional updo hairstyles. Preserve the feeling of lightness, spring colors, and spaciousness from the reference, but convert the abstract watercolor blobs into a real environment.
New photographic scene: place the riders in a wide misty spring meadow with soft green grass, shallow reflective puddles, distant fog, blurred flowering trees in pale pink and fresh yellow-green, and foreground shrubs/branches echoing the referenceâs pink and green color fields. The landscape should feel humid, early morning, and atmospheric, with large negative space above the riders and a distant tree line fading into haze.
Photographic treatment: long-lens documentary/fine-art photography, natural dawn light, gentle backlit mist, realistic horses and fabric motion, shallow atmospheric depth, muted pastel palette, soft contrast, high dynamic range, no ink outlines, no watercolor texture, no illustration, no mural or religious icon style.
Composition constraints: keep the riders small-to-medium in the lower middle of the frame, traveling horizontally; maintain a vertical 4:3 composition with broad sky/fog above and textured meadow below. Avoid modern objects, roads, buildings, visible text, logos, or fantasy elements.
Optional customization: set the historical period to {argument name="historical period" default="Tang dynasty"}, the season to {argument name="season" default="early spring"}, and the mood to {argument name="mood" default="misty serene"}.
