M2 research perceptual graphic-emphasis analysis beyond contrast

Problem

Research whether Dataface should develop a chart-analysis tool that goes beyond standard accessibility contrast checks to model graphic emphasis, including mark size or pixel area, stroke thickness, color dimensions beyond lightness, and human-perception color models. Define when this lens is useful, where it could mislead, and what a practical first version should measure.

Context

• Standard contrast checks do not capture how visual emphasis actually behaves in charts.
• Graph-library quality work would benefit from a better model, but this can easily become pseudo-precision if overbuilt.
• The near-term need is to decide whether a practical first tool exists, not to solve perception science fully.
• Dataface now has local Leonardo-style palette tooling, including pairwise perceptual checks, which provides a practical precedent for bounded analysis instead of purely theoretical discussion.
• This research should treat the Leonardo-style workflow as a useful starting point for local analysis patterns, thresholds, and reporting shape, while recognizing that chart-emphasis analysis will need to account for geometry and density in addition to color separation.
• When color-space analysis is required, prefer OKLCH-based reasoning when possible because it better matches perceptual intent for lightness, chroma, and hue decisions; use other spaces only where simulation or comparison tooling still requires them.

Possible Solutions

• A - Rely only on traditional accessibility contrast metrics and manual design review: lowest complexity, but incomplete.
• B - Attempt a full perceptual model immediately across all chart types: ambitious, but too large and uncertain.
• C - Recommended: do bounded research to define when emphasis analysis is useful, where it misleads, and what a first measurable prototype should include.

Plan

1. Review existing chart-design problems where contrast checks are insufficient.
2. Survey plausible measurable factors such as area, stroke, color space, and density, including where OKLCH is the right working space for analysis.
3. Review how the new local Leonardo-style tooling could be reused, extended, or adapted for emphasis analysis beyond pairwise palette checks.
4. Define a narrow first-pass analysis model with explicit limits and non-goals.
5. Recommend whether to prototype, defer, or fold the work into design review only.

Implementation Progress

Review Feedback

• [ ] Review cleared