Stable Diffusion is a powerful AI image generation model that can create images based on text prompts. With the right prompts, Stable Diffusion is capable of mimicking a wide variety of art styles. Crafting effective prompts is key to getting good results. This article provides an overview of different art styles that can be prompted in Stable Diffusion, along with prompt examples and tips.
Impressionist Style
Impressionist art is characterized by visible brush strokes, an emphasis on light and color over detail, and realistic depictions of scenes from everyday life. To prompt Stable Diffusion to generate art in an Impressionist style, keywords to include are:
- Impressionist
- Visible brush strokes
- Outdoor scene
- Soft lighting
Prompt example:
An impressionist oil painting of a sunny park with people relaxing on the grass, visible brush strokes
Surrealist Style
Surrealist art aims to unlock the unconscious mind through unexpected juxtapositions of images. Prompts for Surrealist art should include:
- Surrealist
- Dreamlike
- Unexpected combinations
- Vivid colors
Prompt example:
A surrealist digital painting of a giant snail crawling up a staircase in an orange sky, very detailed
Pop Art Style
Pop art uses recognizable imagery from popular culture rendered in an eye-catching, graphic style. Good keywords for Pop Art prompts are:
- Pop art
- Graphic
- Vibrant colors
- Popular culture reference
Prompt example:
A pop art graphic painting of Marilyn Monroe wearing Campbell's soup can earrings
Tips for Prompting Art Styles
- Reference specific artists to better capture a style – “in the style of Van Gogh”
- Use multiple style keywords – “impressionist, visible brush strokes”
- Specify a medium – “oil painting”, “pen and ink drawing”
- Give compositional details – “a close up portrait”
- Describe lighting – “soft lighting”, “dramatic shadows”
- Limit variation by locking down details as much as possible
Useful Websites for Stable Diffusion Prompts
Here are some helpful sites for finding and sharing Stable Diffusion prompts:
- Lexica – Search engine for prompts and images
- Prompt Mania – Prompt sharing and inspiration
- Stable Diffusion Upscaler – Upscale and expand images
- Automatic1111 Webui – Local Stable Diffusion interface
- Stable Diffusion Subreddit – Tips, tricks, and prompts
Prompt Engineering Fundamentals
Crafting effective Stable Diffusion prompts is part art, part science. Here are some key principles of prompt engineering to understand:
Be Specific
Narrow down exactly what you want to generate. Overly broad or abstract prompts produce unpredictable results.
Mind the Details
Small descriptive details make a big difference. Include colors, lighting, angles, medium, etc.
Apply Constraints
Limit what the AI generates by locking down as many details as possible. This reduces unwanted outputs.
Try Variations
Experiment with adding, removing, or changing prompt components to refine the output.
Advanced Prompting Techniques
As you become more experienced with Stable Diffusion, you can start utilizing more advanced prompting techniques:
Style Transfer
Blend multiple styles in one prompt – “A cubist interpretation of The Starry Night”
Image to Image
Seed an existing image to expand, edit, or change its style
Masking
Focus generation on certain regions of the image
Prompt Weighting
Control the influence different prompt components have on the output
Responsible AI Use
As with any powerful technology, it’s important use Stable Diffusion responsibly by:
- Avoiding biased, dangerous, or unethical prompts
- Giving proper artist credit and permissions
- Considering how generated art impacts real artists
Used conscientiously, Stable Diffusion offers exciting new creative potential. With the prompting fundamentals covered here, you’re ready to start exploring that potential yourself.