Many artists improve their technical skills through structured drawing practice - carefully analyzing and recreating reference images. However, beginners and intermediate digital artists often face two key challenges:

• Difficulty knowing how to systematically study reference images
• Lack of real-time feedback on their work-in-progress

Our solution organizes the reference study process into three fundamental stages:

• Composition - Guidance for spatial arrangement
• Value - Tools for tonal accuracy
• Color - Support for hue/saturation matching

Key features include:

• Adaptive visual guides that respond to user input
• Automated analysis of value and color accuracy
• Non-prescriptive feedback that encourages artistic reflection

User studies with intermediate artists demonstrated:

• Seamless integration with existing workflows
• Effective promotion of self-assessment during creation
• Flexibility to support diverse artistic approaches

This work explores how computational tools can support artist-driven learning while preserving creative autonomy.

(This project was published to ACM UIST 2025)