Visual Description for AI Image Generation

The Prompt

Create a detailed image generation prompt for: [CONCEPT/SCENE]

Structure the prompt with these layers:

1. **Subject:** What is the main focus? (be specific about pose, expression, action)
2. **Setting:** Where? (environment details, time of day, weather)
3. **Composition:** How is it framed? (close-up, wide shot, bird's eye, rule of thirds)
4. **Lighting:** What mood does the light create? (golden hour, dramatic rim light, soft diffused, neon)
5. **Style:** What does it look like? (photorealistic, oil painting, anime, film noir, editorial fashion)
6. **Color palette:** Dominant colors and temperature (warm/cool)
7. **Mood/Atmosphere:** The feeling it evokes (one word)
8. **Technical details:** Camera/lens simulation (85mm f/1.4, macro, tilt-shift)

Provide:
- Full combined prompt (one paragraph, comma-separated)
- Negative prompt (what to exclude)
- 3 variations (same subject, different styles)
- Recommended generator (Midjourney/DALL-E/Stable Diffusion) and why

Keep prompts under 75 words for best results. Front-load the most important elements.

How To Use It

Describe what you want to see conceptually. The AI will translate your vision into the specific descriptive language that image generators respond to best. Different generators have different strengths — the recommendation helps you pick the right tool.

Why It Works

AI image generators respond to specific technical language (lighting terms, composition rules, style references) much better than vague descriptions. This prompt ensures every important dimension is addressed, producing results that match your intent rather than leaving too much to chance.

Variations

Style consistency:

I want to create a series of images in a consistent style. Define a style guide: color palette, lighting approach, composition rules, and recurring visual motifs. Then generate 5 prompts that look cohesive.

Reference-based:

Describe this image I'm imagining using only visual terms an AI generator would understand. It looks like [REFERENCE A] meets [REFERENCE B] in the style of [ARTIST/ERA].

Iterative refinement:

This prompt produced [DESCRIBE RESULT]. Diagnose what went wrong and rewrite the prompt to fix: [SPECIFIC ISSUE - wrong lighting, wrong composition, wrong style].