GPT Image 2pixelminimal
Turn each uploaded photograph into its own independent, sophisticated design poster. Do not combine multiple photographs; output each separately. Use a 3:4 portrait composition, with upper and lower regions in a strict 1:1 height ratio, each occupying 50% of the canvas.
Keep the original photograph in the upper half, preserving the subject's identity, structure, pose, realistic texture, natural lighting, and original color atmosphere. Apply only subtle, refined color grading for an art-magazine and exhibition-image quality. The environmental background may be naturally extended to fit the format, but never stretch, distort, or alter the subject.
In the lower half, extract only the photograph's most recognizable subject, silhouette, structure, pose, and narrative relationships, and reconstruct them as low-resolution pixel-art illustration / retro 16-bit game art. Do not reproduce the entire photograph or redraw every object. Remove most irrelevant background and detail, retaining only the structures, directional flow, and visual memory cues that best represent the original. Reinterpret these through intentional pixel clusters, hard-edged color blocks, and geometric contours so their correspondence with the photograph above is immediately recognizable.
Use genuine pixel-art language: a limited palette, hard-edged pixels, no anti-aliasing, and clear stepped contours. Reorganize people, architecture, animals, plants, objects, or landscapes into a few large pixel clusters, without pursuing realistic materials or minute detail. Convey light, shade, and space through limited tonal steps, layered blocks, and restrained dithering, not gradients, blur, or a mosaic filter.
Make fresh art-direction decisions and redirect the lower-half composition. Even if the original subject is small, the background cluttered, or the composition ordinary, boldly remove, move, regroup, crop, enlarge, or reduce elements. Keep only genuinely valuable visual relationships and redesign the photograph as an independently successful work rather than faithfully copying its layout.
Maintain a small, stamp-like subject against an exceptionally large area of negative space. Place it freely according to its direction, proportions, and visual center of gravity: off-center, at an edge, suspended, or partially cropped. Negative space is not leftover background but a graphic element as important as the pixel subject. Use positive and negative shapes, density and sparseness, grouping and dispersal, distance, and asymmetrical balance to create pronounced breathing room, spatial presence, and pauses. Prefer drawing less to filling the frame.
Extract 2–4 of the most distinctive and lively colors from the upper photograph and reorganize them into a limited pixel palette. Moderately brighten, purify, and remove grayness as appropriate, keeping the result vivid, clean, light, and thoughtfully designed. Avoid muddy gray, gloom, visual clutter, and a cheap game-screenshot look.
Use text only as a very small editorial intervention, with no language restriction. Freely generate a few words or short phrases from the subject, place, action, emotion, or metaphor. Set them quietly within the negative space in small, restrained bitmap type, forming a refined relationship between typography and pixel imagery.
The overall sophisticated visual effect combines selective distillation, geometric pixels, bold reconstruction, extraordinarily generous artistic negative space, and editorial typography. Avoid object-by-object copying, excessive background retention, a filled frame, ordinary mosaic effects, smooth anti-aliasing, complex realistic detail, and formulaic pixel posters.