Color and Computation

Color in digital space is not paint. It is light. It is numbers. RGB values that our screens translate into photons, wavelengths that our eyes interpret as sensation.

Working with color computationally means understanding this translation. A palette is not a set of pigments but a set of coordinates in a multidimensional space. HSL, LAB, OKLCH—each color model offers a different geometry for navigating that space.

Perceptual Uniformity

The challenge of digital color is perceptual uniformity. In RGB space, equal numerical steps do not produce equal perceptual steps. A gradient from dark to light may appear to accelerate or decelerate depending on the hue.

OKLCH solves this by mapping color to human perception rather than hardware. A ten-unit step in lightness looks the same regardless of hue. This makes it invaluable for generative work where mathematical relationships must translate to visual harmony.

Harmony Through Constraint

I often begin with severe constraints. Three colors. A single hue with varying lightness. Complementary pairs locked to specific saturations. These constraints force creativity, preventing the overwhelm of infinite possibility.

The best palettes emerge from understanding relationships. Analogous colors create calm. Complements create tension. The algorithm can explore these relationships systematically, but the artist must recognize which explorations succeed.

Light as Material

Unlike paint, which absorbs and reflects, screen color emits. This changes everything. Dark backgrounds conserve energy. Bright colors can overwhelm. The additive nature of light means white is presence, not absence.

Working with light means respecting its intensity. Subtlety often succeeds where boldness fails. The most powerful digital colors are often the quietest ones.