We develop an approach for combining illuminance and spectral power distribution of the LED and ambient light and apply our technique for developing an LED camera flashlight balancing the illuminance contrast between object and background. Our method uses the closed loop, multiobjective optimization comprising: (1) characterizing the lighting task by illuminance, correlated color temperature (CCT), and statistical color quality indices that include a set of Statistical Color Quality Metrics and the Color Rendition Index (CRI) implemented with indexes of S (saturation) or D (dulling); (2) measuring the illuminance and the spectrum of the ambient light on the target lighting surface, which might depend on all the sources proving illumination and on the reflected light; (3) determining the desired illuminance of the LED source on the target lighting surface; (4) calculating the desired luminous flux of the LED source according to the desired illuminance; (5) constituting the SPD of the LED source; (6) calculating the relative spectra counts of the LED source and the ambient light on the target lighting surface (7) calculating the CCT and statistical color quality indexes of the combined light; (8) repeating the above steps until the resulting SPD is close enough to the expectation. Using the above method, an LED camera flashlight has been designed, which works together with usual fluorescent ambient light and generates working lighting environment with high fidelity and high CCT (6000K). The spectrum and luminous flux of the LED lamp is automatically tunable with a change of the ambient light.
The color rendition engine based on the statistical metric allows us to uniquely quantify the characteristics of color quality of illumination and assess the color rendition preferences. We now report on using the color rendition engine for revealing individual and cultural differences in color quality preferences of 205 American and Chinese subjects. Our study demonstrated that the majority of individuals preferred the color blend with the same statistical figures of merit on the average but with a much larger spread of blends for Americans. For both groups, the color rendition preferences depended on the object being illuminated. This was demonstrated by illuminating a set of common colored objects and three different paintings. We conclude that the color quality of lighting can be optimized and enhanced using the feedback to change the spectral power distribution of the illuminating source depending on the object being illuminated and on the preferences of an individual observer and a cultural group.