
Original document
GPT Image 2 · en
Scene:
Use the reference board as the source atlas, but do not repeat the grid. Create one coherent square illustration of a quiet canal alley at dusk, seen at walking height, with warm lantern light and soft hand-painted atmosphere.
Subject:
Fuse the reference cues into one character scene: the round-glasses protagonist wears the teal wool coat and embroidered scarf, walks beside an old brick canal bridge, pushes the blue bicycle, and carries the brass compass. A ceramic cat sits near the wooden window frame; a small food stall glows under paper lanterns. The image should feel like all sixteen cues became one believable world.
Important details:
Preserve identity, palette, props, and illustrated texture from the reference board. Use the watercolor texture, ink wash clouds, mossy stair, warm film grain, and hand-painted sky as environmental details. The fusion should be seamless, not a collage; the viewer should sense many references were combined into one place.
Use case:
Second image in a reference-dependent sixteen-reference fusion series, testing whether many source cues can become a coherent walking scene.
Constraints:
Use the reference image as source material, but no visible 4 by 4 grid. No readable text, no logos, no watermark, no duplicated protagonist, no random extra props, no photorealism, preserve exact 1:1 aspect ratio.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.





