translating these pictures into the versions of shadow puppetry
AI image made with GPT Image 2.5 Flare by @daaihq
Prompt
Please create each photo I upload into a separate high-end design poster, without multi-image collages—output each photo individually. The overall layout adopts a 3:4 vertical composition, with the upper and lower sections strictly 1:1 in height, each occupying 50% of the frame. The upper half retains the original photo, preserving the main subject's identity, structure, pose, realistic texture, natural lighting and shadows, and original color atmosphere. Perform only subtle high-end color grading to give it the texture of artistic photography and Eastern exhibition imagery. To fit the frame, the environmental background may be naturally extended, but the main subject must not be stretched, distorted, or altered. The lower half extracts the **most identifiable subject, outline, pose, and narrative relationships** from the photo, reconstructing them into a **traditional Chinese shadow puppet**, performing on a translucent rice-paper screen illuminated by warm light. Regardless of whether the original image features a person, animal, architecture, plant, object, or landscape, it is interpreted as a "manipulable character": the subject is broken down into a few jointed components connected by rivets, with the pose drawn from the most expressive moment in the original photo, accompanied by one or two slender bamboo control rods extending from the bottom of the frame. Do not depict realistically or cartoonishly; instead, re-summarize using the formal grammar of shadow puppets: lateralization, planarization, decorative cutouts, cloud patterns and geometric carved motifs, partial semi-transparent color dyeing. Retain the most recognizable features of the original object, allowing viewers to identify it at a glance, while sensing it has been reinterpreted and hand-carved by a master artisan. Light and shadow are the soul of this work: a warm-colored lamp behind the screen makes warm light spots shine through the cutouts, with the dyed sections presenting semi-transparent vermilion red, indigo blue, olive green, and amber yellow; the edges of the shadow puppet feature soft projected halos and subtle glows, as if slowly moving. There may be extremely subtle double shadows or dynamic blurs, capturing the instant of the puppet in performance. The composition centers on **one scene, one puppet + ample negative space**. The screen is a warm field of beige or light amber light, naturally darkening at the edges; the main subject occupies only about 25%–35% of the frame, which may be off-center or positioned at the golden ratio, with the remaining space kept serene, like a pause before the play begins. Do not include audiences, full-stage views, or complex props. The color palette extracts 2–3 colors with the highest emotional value from the upper photo, translating them into semi-transparent dyes for the shadow puppetry, unified with the warm light field; shadows use deep brown and warm sepia tones, rather than cold black. Text intervenes sparingly, perhaps as vertical small text writing a "play title" or short phrase, like a handwritten program from a troupe, quietly placed in one corner of the screen—no restrictions on language. The overall presentation weaves **shadow puppet cutouts, semi-transparent color dyeing, warm-lit screen, soft projections, and Eastern negative space** into a cohesive light-shadow visual. The core principle is: invite the photo's protagonist onto the screen, letting it perform an entire play in a single pose. Avoid full realism, cluttered scene-filling, cheap "guochao" aesthetics, plastic textures, 3D rendering, cartoonish feel, and templated poster designs.
Source: @daaihq on x.com