Put the cover under the lamp. Each of the five things that made it work is a layer you can switch on: the words, the grid it sits on, the colours it uses, the subject, the light.
GPT-image2 / Midjourney5. Two model names appear side by side, directly setting up a comparison arena that makes people want to know which one generates something more like a real photo.
Posted by KimWang on Xiaohongshu, 23 Apr 2026. Shown with credit and a link back.
Keep all five layers and swap what’s under them: your shop, your budget, your city. You get four covers and the post text.
Verbatim. Copy it, or let Queenie swap in your own details.
Referencing the case study "GPT-image2 is truly 🐮🍻 the models are iterating so fast" — its old photo restoration approach, period texture, visual storytelling and posting tone — generate a new set of retro old-photo image-and-text posts. New topic: {fill in the photo topic you want to restore, rebuild or recreate} Reference platform: Xiaohongshu Content direction: retro old-photo image-and-text Creation method: keep the original case's nostalgic emotion, film/photo-paper texture, character relationships, period feel of the scene and restoration constraints; if it is photo restoration, you must stay faithful to the original subject's identity, facial features, composition and historical context, avoiding excessive beautification, repainting, false historical information and a modern digital plastic look. Please output: 1. Cover title and subtitle; 2. A 6-9 page image-and-text script, each page containing a visual description, main copy, period/texture keywords and a save prompt; 3. Visual prompts that can be used directly for image generation or photo restoration, explaining the requirements for characters, scene, lighting, film grain, damage repair and colour restoration; 4. Posting suggestions: title length, tags, image ratio and platform tone. Constraints: do not replicate the original author, original people or brand marks; do not fabricate historical facts; when real people's photos are involved, prioritise identity consistency, and do not make them younger, beautify, slim the face or alter facial features.
