Warm-cool Based Color Description for White Lighting

Abstract

We use lighting every day and modern lighting is more than just illuminating the darkness. Thanks to different uses of white lighting, different atmospheres as desired can be created in the same space. To achieve effective lighting design, it is necessary to accurately express the color of white lightings. Correlated color temperature (CCT) has been most widely used in the lighting market, which gives a single number to communicate color of white lights. However, it has been questionable about its visual resemblance of iso-CCT points, i.e., the same CCT values are not perceived as the same CCT. The current CCT is defined as a calculation method on the (u’, 2/3v’) chromaticity space, which is the previous (u, v) space. Recent studies have generally suggested using (u’, v’) chromaticity space instead of (u, v), but it is not conclusive due to lack of visual experimental data. This study approached the CCT perception issue with warm-cool feelings. It is widely accepted that CCT change is closely correlated with warm-cool feelings, i.e., lower CCT is perceived warmer and higher CCT is perceived cooler. In this regard, the same CCT would evoke the same level of warm-cool feelings. To achieve an accurate prediction of warm-cool feelings, it is necessary to estimate hue first, which has a close relationship with warm-cool feelings. In this study, a series of psychophysical experiments was performed to investigate hue and warmcool feelings of near-white lights using either a LED lighting booth or a LCD display. The experiments were designed so that participants could be adapted to the Reference condition before evaluating test colors. Two experiments were conducted. First, the hue experiment was carried out at five different reference conditions, i.e., 3500 K (Duv 0.000), 5000 K (Duv 0.000), and 6000K (Duv shifts of +0.015, 0.000, and −0.015). Each reference condition had a total of 24 to 48 test colors, and each test color was evaluated in hue using a magnitude estimation method. Second, the warmcool experiment was performed at six different reference conditions at 3500 K (Duv shift of −0.010, 0.000, and +0.010), 5000 K (Duv 0.000), 6000 K (Duv shift of −0.010 and +0.010) using two different experimental methods. In both hue and warm-cool experiments, 10 to 20 participants were recruited for different experimental conditions. The results of the experiments showed a strong relation between hue and warm-cool feelings, so the warm-cool model for lights was proposed depending on hue. For estimation of hue, the hue prediction model (HPM21Oh) was proposed, and its main feature was the development in the degree of adaptation and unique hue settings for lights, which performed significantly better than the current color appearance model, comprehensive CIECAM02 (CAM16). Using the proposed warm-cool model, the warm-cool based correlated color temperature (CCTWC) was assigned as a new way of describing color of lights.