Stable Diffusion Prompt Text

Stable Diffusion is an AI system that generates images from text prompts. Crafting effective prompts is key to producing high-quality AI art with Stable Diffusion. This article provides prompt engineering tips and real-world examples to help you get started.

When creating a prompt, first decide on the subject matter and style you want the AI to render. The more descriptive your prompt, the better the results. Use adjectives and style cues to set a specific mood and direct the AI’s creative process.

Basic Prompt Structure

Prompts generally follow this basic structure:

"A [style] [subject] [modifiers], [style cues], [composition]"

Let’s break this down:

Style: The art genre or medium, like “oil painting” or “digital illustration.”
Subject: The main focus, like “wizard” or “dystopian cityscape.”
Modifiers: Descriptors to tweak the subject, like “elderly wizard” or “gritty dystopian city.”
Style cues: Artistic directions, like “intricate detail” or “Rembrandt lighting.”
Composition: Guidance on framing, perspective, etc.

Prompt Engineering Tips

Here are some key tips for engineering effective prompts:

Be specific. The more details you provide, the less room there is for unwanted outputs.
Limit scope. Stick to 1-2 main subjects; complex scenes can confuse the AI.
Use references. Name a specific artist, art style, or existing work to guide the AI.
Iterate and refine. Tweak prompts over multiple generations to home in on your vision.

Useful Keywords

These keywords can help provide direction:

Style: oil painting, watercolor, pen and ink, vector art, low poly, pixel art
Quality: intricate detail, depth of field, cinematic lighting, rule of thirds
Mood: somber, lighthearted, mystical, nostalgic
Composition: portrait framing, golden ratio, leading lines, negative space

Stable Diffusion Prompt Examples

Here are some real-world examples of effective Stable Diffusion prompts:

Fantasy Character Portrait

An intricate fantasy oil painting portrait of an elderly wizard with a long white beard, purple robes, and wizard's hat, dramatic Rembrandt lighting, highly detailed

Futuristic Cityscape

A retrofuturistic 80s digital illustration of a gritty dystopian megacity skyline, with flying cars, neon lights, and art deco skyscrapers, cyberpunk concept art trending on Artstation

Still Life Painting

A photorealistic Baroque still life oil painting of fruit in a bowl with dramatic chiaroscuro lighting, by Caravaggio

Advanced Prompting Techniques

As you gain experience with Stable Diffusion, try these advanced techniques:

Stylistic References

A fantasy landscape in the style of John Howe's Lord of the Rings art, mystical ruins overgrown with vines

Directing Focus

A close-up portrait of an astronaut's helmet visor reflecting a vibrant nebula and stars, intricate NASA space suit in sharp focus

Compositional Guidance

A full body portrait of a knight in armor, standing contrapposto holding a sword pointing down, detailed gothic architecture in the background to the left and right, depth of field

Iterative Refining

Start broad, then add details. For example:

wizard --> elderly wizard with beard --> intricate fantasy painting of elderly wizard --> Caravaggio lighting...

Conclusion

Crafting effective prompts is key to creating high-quality AI art with Stable Diffusion. Follow the tips and examples in this guide as a starting point, then refine through experimentation. Remember to be specific, limit scope, and provide ample style and compositional direction. With practice, you’ll be able to consistently produce AI art that matches your creative vision.