A good prompt is surprisingly similar to a photographic briefing.
In Zappa, every prompt is constructed according to a fixed sequence. Not because it sounds nice, but because image models respond better to a clear visual structure.
The basic structure of a fashion prompt consists of:
1. Image type and framing
Is it a close-up, half body, or full-body image?
2. Body orientation
In which direction is the body facing and where is the weight distributed?
3. Head position
Where is the model looking and at what angle is the head positioned?
4. Arms and hands
Hands often determine the tension of an image.
5. Legs and points of contact
Where does the body touch the environment?
6. Clothing and material
How does the fabric fall, where is there tension or movement?
7. Lighting
The direction of the light determines shape and depth.
8. Environment
Only visible elements—do not add anything that is not there.
By using this sequence, an image model understands the image almost as if it were a studio setup.
That makes the difference between an AI image and an editorial fashion image.