A developer has published a color space and procedural generation algorithm designed to make it easier for digital tools—such as character creators and art software—to represent diverse human skin tones.
According to the project, the work identifies what the creator calls the “broadest inclusive range of colors in the RGB color space that correspond to plausible, but simplified skin tones.” The goal is to find simple equations defining that range, making it usable across different contexts.
The creator acknowledges that representing human skin tones digitally remains a complex problem. Current approaches vary widely: emoji standards use roughly 5 shades; makeup brands might offer around 50 options; and digital character creators sometimes present all 16.7 million RGB colors without guidance. The work attempts to find a middle ground between these extremes.
The methodology involved manually labeling a large number of colors as plausible or implausible skin tones using a custom interface. The creator describes the resulting data shape as resembling “a banana that swoops between black and white,” skewed toward red tones and away from blue. To make this shape mathematically tractable, the developer applied principal component analysis, a statistical technique for simplifying complex data.
The final output includes a JavaScript-based color picker and a Python procedural generation algorithm (with JavaScript equivalents) that samples from the defined color space. Users can integrate this into their own tools.
The creator emphasizes significant limitations. Skin tone varies across different body areas and is affected by melanin concentration, blood flow, light scattering through skin layers, and conditions like vitiligo and freckles. Health conditions can produce skin colors far outside typical ranges. Additionally, the work reflects the subjective choices and perceptions of a single individual, not peer-reviewed research. Colors also display differently across screens and lighting environments.
The project is framed as “good enough” for simplified use cases like character creators and digital art, rather than as authoritative or comprehensive. The creator notes that addressing these limitations would require broader investigation and research collaboration.
Key facts
- The algorithm uses principal component analysis to define a mathematical color space for skin tones
- The approach aims to balance between overly limited preset palettes and overwhelming all-16-million-color options
- The methodology involved manually labeling colors as plausible or implausible based on the creator’s subjective assessment
- The creator acknowledges the work has significant limitations and should not be treated as authoritative or scientifically rigorous