ちぎり絵コラージュポスター

Turn the uploaded photo into a single high-end torn-paper editorial art poster — not a collage of several images. Use a vertical 3:4 diptych composition: the top holds the real original photograph, clearly visible; the bottom rebuilds the same scene as a "torn-paper scene reconstruction". The two bands are close to equal in height, separated by a clean horizontal edge — no haze, no bleeding between them. Output the face of the poster only: no picture frame, no wall, no hand-held mockup.

The top band occupies roughly 47–50% of the frame and preserves the original photograph faithfully. Keep the viewing angle, the direction of the vanishing point, the slope of the road, the buildings on either side, the eaves, the doors and windows, the signage, the vehicles, the plants, the overhead wires, and the number and posture of any people, along with the original natural light and colour. Apply only light travel-magazine grading and the thinnest film grain; it must still read clearly as real photography. The sky, the road surface or quiet parts of the surroundings may be extended naturally to fit the frame, but do not repaint buildings, move people, alter vehicles, stretch, mirror or duplicate the subject.

In the lower band, do not simply put a paper filter over the same framing, and do not assemble rectangular photo fragments. First choose the single dominant vanishing line of the space, the main facades on either side, and one to three objects that best identify the place — a pedestrian, a vehicle, a lantern, an umbrella, overhead wires, a plant, a sign or a roof. Actively delete repeated window panes, complicated crowds, irrelevant small objects and photographic noise, then reorganise what remains into a smaller, clearly outlined, physical torn-paper scene.

The reconstructed scene fills about 60–82% of the lower band, centred or slightly offset, with roughly 20–35% of continuous ivory paper left around it. Re-judge the scale and visual hierarchy against the original so that the most recognisable building, object or figure becomes the anchor on the sheet; do not transcribe the photograph's full composition. A small torn circular window may be added in the white space, enlarging one real detail taken from the original, connected back to the main scene by a single graphite arrow.

Build the facades, road surface, eaves, plants, sky and figures piece by piece from hand-torn matte fibre paper, thin rice paper, corrugated card, kraft paper and a little old printed stock. Every major shape must show clearly visible fibrous torn edges, slight overlaps, real paper thickness, lifted corners and short, natural cast shadows. Torn edges may be irregular, but the outline of the scene, the direction of the space, the boundaries of the buildings and the identity of the subject must stay legible. No smooth vector slices, no laser-cut edges, no flat digital shadows.

Take the colour entirely from the photograph above, reducing its complexity to five to seven greyed paper tones. Build the skeleton from the true dominant colours of the walls, the road, the roofs, the planting and the people, and keep exactly one characteristic accent from the original — a rust red, a cobalt blue, an olive green or whatever the photograph actually offers. Do not apply a fixed Japanese-travel palette to every location.

Text — generated
Arrange a restrained group of handwritten editorial type in the ivory white space to the right or below: one all-capitals headline of two or three words distilled from the place, the subject or the mood; a fine rule beneath it; a small numbered note under the rule; and one descriptive line of no more than twelve words. The type sits quietly on the blank paper and never covers the torn-paper scene.