Illuminating the limits of the international unit of light, the candela

The candela, the international unit of light used for nearly a century, underpins photometry. However, a new study reveals that measurements based on it inaccurately gauge the brightness of colored light sources and significantly diverge from human perception. Led by perception researcher Karl Gegenfurtner at Justus Liebig University Giessen, an international team discovered a straightforward rule for predicting perceived brightness: it can be approximated by considering only the highest of the three weighted color components—red, green, and blue.

This rule explains over 95% of brightness judgments and surpasses all existing photometric models. As the candela relies on human perception, the discrepancies between its derived quantities and actual perception have implications for lamp labeling, display calibration, and lighting planning in various settings. The study, published in Proceedings of the National Academy of Sciences, shows that saturated colors seem much brighter, and blue light plays a more substantial role in perceived brightness than photometric measurements suggest.

The discrepancy stems from the 1920s assumption that a color mixture's brightness equals the sum of its parts. Traditional brightness matching methods are also slow, challenging, and inconsistent among observers. To overcome this, the Giessen team asked participants to sort colored patches from bright to dark, yielding remarkably consistent results over time, even when repeated at home or online.

By aggregating these data, the team tested various models, including luminance, radiance, and established color appearance models like CAM16, finding that none captured the observed pattern. The breakthrough came when they found that weighting the red, green, and blue components and keeping the largest value perfectly predicted brightness judgments.

This simple rule aligns with brain responses to flickering lights under both fast and slow modulation. In real-world lighting situations, participants consistently chose illuminations rated as brighter by the new model, indicating its relevance for everyday seeing. As modern LEDs allow for flexible spectral control without changing luminance, the findings offer a reliable method for measuring colored light brightness and a computational basis for perceiving light in everyday environments.

This could greatly impact building lighting design, display calibration, and the development of more energy-efficient lighting.