Qualitative study on color clarity in color quality of LED lighting products

　　The International Commission on Illumination (CIE) defines color rendering as "the conscious and unconscious effect of a light source on the color appearance of objects compared with a standard reference light source"[9]. Currently, there is only one generally accepted method for evaluating color rendering in the lighting industry - the color rendering index (CRI). This evaluation method was proposed in 1960 through the joint efforts of some scientists and light source manufacturers. CRI is based on the "test sample color displacement method" and compares the color difference between the sample color under the reference illuminant and the test light source to evaluate the color rendering of the light source, that is, the "realism" or "naturalness" of the color[1]. The "standard reference light source" in the color rendering index is the spectrum of blackbody radiation below 5000K and the spectrum of average daylight above 5000K. The luminescence mechanism of incandescent lamps belongs to blackbody radiation. Therefore, it is obvious that no matter how lighting technology develops, the color rendering or color rendering index of incandescent lamps is the highest. As long as the spectral form of other advanced lighting technologies deviates from the spectrum of the reference light source, their color rendering index will decrease.

　　In 1947, PJ Boum described several reasons why daylight is an ideal lighting source: "The colors of objects illuminated by daylight (1) have very rich colors, (2) can easily distinguish subtle differences in color, and (3) make the colors of objects around us look very natural" [2]. From Bouma's description, we can know that in addition to the concept of CR, there are other concepts to describe the "color quality CQ (color quality)" of a light source [3], such as clear color, rich color, etc.

　　This paper investigates the current status of research on the color clarity index of lighting sources at home and abroad, and draws on relevant research methods to conduct visual ergonomics experiments, and conducts qualitative analysis on the relationship between color rendering index and color temperature and color clarity. The experimental results show that white light LED lighting products with high color temperature have higher color resolution than those with low color temperature, and the subjective perception of color is clearer, and the color rendering index has little correlation with color clarity, so the color rendering index cannot be used to evaluate the color clarity of LED lighting sources.

　　1. Current research status at home and abroad

　　Color clarity can be divided into two aspects: the first is the effect of the lighting source on the observer's ability to distinguish colors, and the second is the subjective feeling of the observer on the color clarity under the overall light environment. Many researchers at home and abroad have conducted a lot of research on the color resolution of LED lighting sources.

　　French researchers Elodie Mahler and others[4] studied the color resolution ability of white light sources composed of monochromatic LEDs. The test light sources they selected were RGB (red, green, and blue LEDs), RGBA (A is amber), WR (two cold white phosphor LEDs and a red LED), and WWARGB (a combination of two cold white lights, two warm white phosphor LEDs, and amber, red, green, and blue LEDs). The color temperature was maintained at 4000±40K, the illumination was 660±9lux, and a halogen tungsten lamp plus a filter was used as the reference light source. They used a homemade C32 color chess as a tool for the ranking experiment. The results showed that the RGB type had the worst color resolution ability, while the WR and WWARGB had better color resolution abilities than the reference light source.

　　Mark S. Rea et al. [5, 6] from the United States recommended using CRI as the evaluation index for natural color, GAI as the evaluation index for vivid color, and FSCI as the evaluation index for color resolution. For this purpose, they conducted many related visual experiments. In one experiment, they selected 5 types of white light LEDs and 2 types of fluorescent lamps, and divided them into warm white, cool white, and mixed color temperature groups, and set two different illuminations. The subjects performed the FM100-huetest color resolution test under these light sources. The experimental results showed that color resolution ability is closely related to illumination. As illumination increases, color resolution ability also increases accordingly. At the same time, the results showed that CRI has no correlation with color resolution, but GAI and FSCI have a good correlation with color resolution.

　　The current experimental design method for visual psychology experiments on light source color quality is to keep the desktop illumination of the light environment the same and group the test light sources according to similar color temperatures. Use the FM100 color chess test to analyze the impact of the light source spectrum on visual color discrimination ability　　.

　　In order to study the relationship between the spectrum of LED lighting sources and color clarity, this paper designed two experiments. The first one used the Farnsworth 100 hue test method to evaluate the changes in human eye color discrimination ability under different light source spectra. The second experiment used the subjective experience evaluation method. In the experiment, the subjects observed the given light environment in a dark room and then answered their subjective feelings according to the standardized scale.

　　A total of 6 LED light sources were selected for the experiment, all of which were blue LED chips that excited phosphors to produce white light. In order to obtain the light source spectrum of the field of view actually observed by the human eye, the spectrum reflected by the center of the desktop in the light box was measured by a spectroradiometer. The measuring instrument was a calibrated Konica Minolta CS2000. The horizontal illumination in the light box was measured using a TES1330A illuminance meter. The relevant parameters of the light source are shown in Table 1. Light sources 1 to 3 are 4000K color temperature groups, and 4 to 6 are 8000K color temperature groups. Each color temperature group is also divided into three levels of color rendering index 70, 80, and 90. A total of 25 subjects participated in the visual psychology experiment in this paper, of which 11 completed the FM100 color chess test and 25 completed the subjective experience evaluation experiment. The average age of the subjects was 20.7 years old, ranging from 19 to 25 years old.

　　1. Farnsworth 100 hue test experiment

　　Previous studies have found that changes in illumination have a significant impact on human color discrimination. Therefore, during the experiment, the illumination was measured in real time using a TES1330A illuminance meter to maintain it at 500 lx. The experiment only considered changes in human color discrimination caused by differences in color temperature and spectrum.

Figure 1. Color chess sorting experiment in a light box lighting environment

　　The subjects first practiced color chess sorting under the first test light source, completed at least one complete test, and then officially started the experiment. Before starting the sorting, the subjects had to fully adapt to the current lighting environment for 15 minutes. During this period, the subjects could only observe the lighting environment in the light box and could not be disturbed by other light sources. After the adaptation was complete, the sorting test began. After the experiment was completed, the next lighting source was changed, and the subjects adapted to the current lighting environment in this environment. At this time, the experimenter recorded the experimental results and re-shuffled the color chess. After the subjects completed the sorting under the two light sources, they rested for 30 minutes and then continued the experiment.

Figure 2 Average error scores of subjects' color chess sorting under six light sources

　　The experimental data is shown in Figure 2. It can be seen that the error score of color sorting is higher at 4000K color temperature than at 8000K correlated color temperature. Therefore, the increase in color temperature can improve the color discrimination ability of the human eye. In the 8000K group, the color sorting score of the light source with a color rendering index of 80 is higher than that of the light source with a color rendering index of 70. Therefore, it can be seen that the color rendering index is not always related to color discrimination ability.　　2. Subjective evaluation experiment of color clarity

Figure 3  Comparison of subjective evaluation of lighting color clarity and selection of experimental scenes

　　In order to enable the subjects to evaluate the color clarity of the light environment, it is necessary to present a simulated light environment to the subjects, so simulated fruits, canned drinks, magazines, and MCC color cards are placed in the light box. This experiment uses all 6 COB light sources, divided into 4000K group and 8000K group. In the experiment, two light boxes are respectively selected from two color temperature groups to light up. The subjects will compare the subjective feeling of color clarity of the light environment on both sides and choose the light environment with clearer subjective feeling. A light source needs to be compared with three light sources in different ways, and repeated three times, so one light source will be compared 9 times, and 25 subjects will compare, so one light source will be compared 9*25=225 times. The number of times selected is divided by the total number of comparisons to obtain the percentage of comparison selection evaluation of the subjects. The experimental results are shown in Figure 4. From the test results, it can be found that in the subjective comparison of color clarity, almost all subjects choose high color temperature light sources, but the relationship with the color rendering index CRI is not significant.

Figure 4 Results of subjective comparison test on white LED color clarity

　　3. Conclusion and Outlook

　　The two experiments conducted in this paper can draw the following two qualitative conclusions: First, white light LED lighting products with high color temperature have higher color resolution and subjective color clarity than those with low color temperature; second, the color rendering index has little correlation with color clarity, so the color rendering index cannot be used to evaluate the color clarity of LED lighting sources.

　　The International Commission on Illumination (CIE) Technical Report 177:2007 “Color Rendering of White Light LEDs” points out that “the results obtained when CRI is used to evaluate the color rendering of a series of light sources including LED light sources are not satisfactory”, and the technical report also points out that “a new color rendering evaluation index or a set of evaluation indices should be developed as soon as possible to replace the existing CRI” [3]. According to the research results of this paper, the color quality of lighting sources is multifaceted and cannot be evaluated by a single evaluation model. In the future, a color quality evaluation framework with multiple evaluation indices should be established to systematically and completely evaluate the color quality attributes of white light LED lighting sources and provide scientific and objective evaluation standards for LED lighting.