AIToolDesk

AI image generators can produce a striking picture from a single sentence, but getting the image you actually had in mind takes a bit of method. This guide gives you a reusable formula and a troubleshooting list.

The prompt formula

Describe the image in this order:

1. Subject: who or what, doing what.
2. Setting: where and when.
3. Style: photo, watercolor, flat illustration, 3D render and so on.
4. Composition: close-up or wide shot, angle, where the subject sits in the frame.
5. Light and color: soft morning light, high contrast, muted palette.
6. Format: aspect ratio, for example square for social posts or wide for banners.

Before and after

A coffee shop.

A small neighborhood coffee shop on a rainy morning, seen from outside through a large window. A barista pours latte art at the counter. Warm interior lights, cool blue street tones, photographic style, shallow depth of field, wide 16:9 frame.

The second prompt fixes the viewpoint, mood, lighting and format, so far less is left to chance.

Details that change results the most

• Lighting words such as golden hour, overcast, studio lighting or neon.
• Camera language such as close-up, aerial view, low angle or macro.
• Medium such as oil painting, pencil sketch, paper cut-out or isometric illustration.
• Mood such as calm, dramatic, playful or minimal.

Fixing common problems

• Too busy: remove adjectives and describe one main subject.
• Wrong style: move the style description to the start of the prompt.
• Garbled text in the image: keep any words very short, or add the text afterwards in a design tool.
• Odd hands or faces: regenerate, choose a wider shot, or use the tool’s edit feature to fix one area.
• Inconsistent results: keep the prompt identical and change only one detail at a time.

Use images responsibly

Check each tool’s terms for commercial use. Avoid prompts that copy a living artist’s distinctive style or recreate trademarked characters and logos, and don’t create realistic images of real people without their permission. If an image could be mistaken for a real photo, consider labeling it as AI-generated.

Key takeaways

• Subject, setting, style, composition, light and format: in that order.
• Lighting and camera words change results more than extra adjectives.
• Change one thing at a time when refining.