Best Practices
Refine ur skillzzz
Last updated
Refine ur skillzzz
Last updated
[put the image here - collage of example images showing good prompt techniques]
Creating amazing AI-generated images is both an art and a science. This guide provides expert techniques, tips, and best practices to help you get the most out of Nova-47's image generation capabilities.
Whether you're a beginner or experienced with AI art, these strategies will help you create more impressive, accurate, and beautiful images with any of our commands.
The prompt is the most important factor in AI image generation success. Here are strategies for crafting powerful prompts:
Subject
The main focus of your image
"a medieval knight"
Action/Pose
What the subject is doing
"riding a horse into battle"
Environment
Where the scene takes place
"on a misty battlefield"
Lighting
How the scene is lit
"illuminated by dramatic sunset light"
Style
Artistic style reference
"in the style of a renaissance painting"
Details
Important specific elements
"with a red cape flowing in the wind"
Composition
How the image is framed
"close-up portrait shot"
This prompt contains:
Clear subject (white wolf)
Action/pose (standing on cliff)
Environment (rocky cliff, winter forest)
Lighting (northern lights, night sky)
Style (fantasy art)
Details (highly detailed fur)
Composition (wide-angle)
Be specific and detailed
Be vague or too general
Include artistic style references
Assume the AI knows what you want
Specify lighting and mood
Use confusing or contradictory descriptions
Focus on one main subject/scene
Request too many subjects or complex scenes
Use descriptive adjectives
Use subjective terms like "beautiful" or "good"
Include composition details
Forget about perspective or framing
Use common, clear language
Use obscure references or slang terms
Mentioning specific art styles significantly impacts results:
"Realistic"
Highly photorealistic rendering with accurate lighting and lifelike textures
"Anime"
Japanese animation style with vibrant colors and expressive features
"Watercolor"
Soft, flowing style with bleeding colors and gentle gradients
"Studio Ghibli"
Whimsical animation with intricate backgrounds and magical realism
"Cyberpunk"
Futuristic dystopian aesthetic with neon cityscapes and advanced technology
"Fantasy"
Magical realms with mythical creatures and epic landscapes
"Pixel Art"
Retro gaming style with limited color palette and clear pixels
"Concept Art"
Professional illustration style with detailed environments and cinematic composition
"Comic Book"
Bold outlines with dynamic action and comic panel aesthetics
"Gothic"
Dark medieval architecture with ornate details and mysterious atmosphere
"Film Noir"
High contrast black and white with dramatic shadows and moody lighting
"Vaporwave"
1990s internet aesthetic with glitch art and pastel colors
Different commands work best with different approaches:
Use a balance of detail and clarity
Works well with direct style references
Effective with most subject matter
Benefits from specific mood/atmosphere descriptions
Good with both realistic and stylized approaches
Focus on a single clear subject
Use simple backgrounds or no background references
Works best for character designs, objects, or symbols
Avoid overly complex scenes or multiple subjects
Excellent for cartoon, anime, and icon-like styles
Provide a concise description of the desired background
Specify the relationship between subject and background
Include lighting that complements the subject
Mention atmosphere and mood for cohesive results
Reference how the subject should integrate with the background
Leverage all the detail you can include
Experiment with different engine options for the same prompt
Be specific about materials, textures, and fine details
Great for complex scenes with multiple elements
Take advantage of negative prompts for precision
Use natural language descriptions
Excellent with landscape and environment descriptions
Take advantage of the lighting and color parameters
Great for photorealistic scenes
Works well with emotional or atmospheric descriptions
Keep prompts relatively concise
Excellent for stylized art and illustrations
Good balance of speed and quality for iterations
Works well with clear artistic style references
Great for character designs and simpler compositions
Negative prompts tell the AI what to avoid in your image:
"a fantasy castle on a mountain"
"a fantasy castle on a mountain"
negative prompt: dark, gloomy, ruins, damaged"
The seed value determines the specific randomization used to generate an image:
Save seeds of images you particularly like
Reuse seeds with slightly modified prompts for consistent subjects
Share seeds with friends so they can recreate similar images
Use different seeds with the same prompt to explore variations
Add these terms to enhance detail in specific areas:
Textures
"intricate details," "fine textures," "high resolution textures"
Lighting
"volumetric lighting," "cinematic lighting," "golden hour lighting"
Quality
"high resolution," "8k," "highly detailed," "crisp details"
Camera
"close-up," "macro photography," "portrait framing," "symmetrical composition"
Mood
"dramatic," "vibrant," "atmospheric," "moody," "ethereal"
Combine different official styles for unique results:
You can also combine style descriptions in your prompt:
Social media posts
Square (1:1)
All commands
Desktop wallpapers
Widescreen (16:9)
All commands
Mobile wallpapers
Portrait (9:16)
All commands
Discord banners
Wide (16:9)
Premium commands
Character portraits
Portrait (3:4)
All commands
Landscape scenes
Widescreen (16:9) or Panoramic (2:1)
Premium commands
For maximum quality, follow this workflow:
Generate multiple variations using regenerate button
Select the best composition based on overall structure
Upscale to enhance details and quality
Remove background if needed for your use case
Save the final result in appropriate format for your needs
Try these patterns for interesting results:
Style Fusion
"Portrait of a warrior, mix of cyberpunk and feudal Japanese style"
Temporal Blend
"A city half in daylight and half at night, seamless transition"
Concept Mashup
"A library that's also an aquarium, books floating among fish"
Material Transformation
"A forest where the trees are made of crystal, sunset lighting"
Instead of describing exactly what you want to see, describe the concept:
Too much clutter
Simplify prompt, focus on fewer elements
Unrealistic proportions
Add "anatomically correct" or "realistic proportions"
Wrong style
Be more specific about artistic style references
Incoherent scene
Focus on a central concept, reduce complexity
Poor composition
Specify camera angle and framing
Low detail
Add detail-boosting terms, use upscale
Unwanted elements
Use negative prompts to exclude specific items
Start simple and gradually add complexity as you learn
Keep a prompt journal of your most successful prompts
Study other AI art to learn effective techniques
Iterate and refine rather than starting from scratch
Combine commands for the best results (generate → upscale → remove bg)
Share your creations with the community to get feedback
Experiment constantly - AI image generation is still evolving!
Remember that AI image generation is a creative partnership between you and the AI. The more you practice and refine your approach, the better your results will be. Happy creating! 🎨✨