Chapter 8: Sketching, Digital Refinement & Corrections

10-15% Rule Application: Sketching first can triple your keeper rate; Figma corrections can save 50% of re-generation costs

Content:

Pre-Generation Sketching:

• Digital vs. analog tools (pencils, stylus, paper)
• When to sketch before AI generation
• Using sketches as AI input (img2img, ControlNet, etc.)
• Tracing AI outputs for refinement
• Hybrid workflows combining hand and AI work
• Skill level requirements: You don't need to be an artist

NEW: Figma for Post-AI Corrections

• Understanding the Correction Workflow:
• When to correct vs. when to re-generate
• Cost-benefit analysis: Correction time vs. generation credits
• Types of errors best fixed in Figma vs. re-prompting
• Non-destructive correction approaches

• Using Figma Vector Tools for Cleanup:
• Vector shape tools for fixing AI artifacts
• Pen tool for refining edges and details
• Using frames and layers for organized corrections
• Color correction and adjustment layers
• Cloning and patching techniques in Figma
• Typography corrections for AI-generated text
• Symmetry tools for fixing asymmetrical errors

• Combining Multiple AI Outputs in Figma:
• Compositing different AI generations
• Taking the best elements from multiple outputs
• Blending techniques and masking
• Creating hybrid images from partial successes
• Managing layers for complex composites
• Maintaining style consistency across combined outputs

• Exporting Corrected Work for Re-rendering:
• Export settings that preserve correction quality
• Preparing corrected images as new AI inputs
• Using corrected outputs as reference images
• Creating img2img ready exports
• Documentation of corrections for prompt refinement
• Building a library of corrected assets

Advanced Figma Correction Techniques:

• Common AI Errors and Figma Solutions:
• Fixing hands and fingers (when salvageable)
• Correcting facial asymmetry
• Cleaning up background artifacts
• Fixing perspective issues
• Removing unwanted elements
• Adding missing details
• Correcting color inconsistencies

• Figma Plugins for AI Workflow:
• Useful plugins for image enhancement
• Cleanup and refinement plugins
• Color harmony tools
• Batch processing capabilities

• Decision Framework:
• Flowchart: Should I correct or re-generate?
• Time estimation for common corrections
• Quality threshold guidelines
• When Figma corrections are worth it
• When to just re-prompt instead

Practical Elements:

• Tool recommendations and comparison (cost, learning curve, AI compatibility)
• Workflow diagrams: Sketch → AI → Figma Corrections → Refine
• Exercise: Sketch, generate, trace process for one shot
• NEW: Figma Correction Exercises:
• Exercise 1: Fix common AI errors in Figma
• Exercise 2: Composite 3 AI outputs into one final image
• Exercise 3: Prepare corrected asset for re-rendering
• Best practices for sketch-based prompts
• Figma correction best practices guide

• ⚡ Quick Win: Stick figure compositions that improve AI output
• ⚡ Quick Win: 5-minute Figma fixes for common AI errors