Complete Guide to AI Image Creation
π Quick Actions
β‘ AI Prompt Builder
π― Quick Reference
Ready to start your AI image journey!
Pre-Project Checklist
- β‘ Tool selected
- β‘ Goals defined
- β‘ Content prepared
Per-Image Process
- β‘ 3-5 concepts generated
- β‘ Specific prompts created
- β‘ 4-6 variations tested
- β‘ Results refined
- β‘ Post-processing planned
Emergency Troubleshooting
Weird text? β Text Issues solutions
Inconsistent style? β Consistency solutions
Inappropriate imagery? β Abstract Concepts help
Poor quality? β Quality Control tips
Not matching vision? β Layout Control solutions
π 4-Phase Workflow
Phase 1: Preparation
- Define Goals: Purpose, audience, usage context
- Select Tool: Based on budget and requirements
- Prepare Content: Clean text, identify themes
Phase 2: Ideation
- Input to Claude: Share content and context
- Request Concepts: Multiple approaches (literal, symbolic, abstract)
- Clarify Details: Style, mood, cultural considerations
Phase 3: Creation
- Generate Prompts: Tool-specific, optimized
- Test & Iterate: 4-6 variations per concept
- Create Variations: Adjust style and mood
Phase 4: Refinement
- Evaluate: Against original goals
- Post-Process: Add text, combine elements
- Build Library: Save successful prompts
π οΈ AI Image Tools: Choose Your Platform
Select the right tool for your project needs and available resources:
Midjourney
Strengths: Exceptional quality, active community
Access: Subscription required, Discord interface
Use when: Creating mood boards, artistic interpretations
Adobe Firefly
Strengths: Rights-cleared, Adobe integration
Access: Creative Cloud subscription
Use when: Professional marketing, branded content
Google ImageFX/Gemini
Strengths: Speed, Google Workspace integration
Access: Often included in Google subscriptions
Use when: Quick concepts, presentations
Canva AI
Strengths: User-friendly, built-in templates
Access: Freemium model
Use when: Social media graphics, simple layouts
β οΈ Common Problems & Solutions
Text Issues
- Generate base images, add text separately
- Use "no text" in prompts
- Focus on symbolic imagery
- Create designated text areas
Layout Control
- Break complex scenes into parts
- Use composition keywords: "centered," "rule of thirds"
- Generate multiple variations
- Describe spatial relationships clearly
People & Faces
- Use silhouettes or back views
- Focus on gestures over faces
- Try abstract representations
- Use crowd terms like "diverse congregation"
Consistency
- Develop style templates
- Use identical base prompts
- Generate series in same session
- Keep detailed prompt notes
Abstract Concepts
- Use symbolism: doves, light, water, mountains
- Focus on nature scenes
- Create atmospheric backgrounds
- Use metaphorical approaches
π‘ Best Practices
Prompt Writing Excellence
- Start simple: Add complexity gradually
- Use descriptive adjectives: serene, dramatic, ethereal
- Specify style: photorealistic, artistic, minimalist
- Include lighting: soft light, golden hour, backlighting
Quality Control Process
- Generate multiple options before choosing
- Test simple concepts first
- Keep a failed prompt log
- Save successful prompts
Efficiency Strategies
- Batch similar requests
- Create prompt templates
- Build keyword libraries
- Use Claude for prompt refinement
Cultural & Theological Sensitivity
- Consider audience expectations
- Avoid overly literal interpretations
- Focus on universal themes
- Review with others before use
Remember: AI image generation is iterative. Expect to refine and adjust as you learn what works best for your needs.
Vibe coded by Tim Pennington-Russell with claude.ai and Gemini for First Baptist Church, Washington DC.