Effect of Lighting Variability on the Color Difference Assessment
Noor, Kashif. Effect of lighting variability on color difference assessment (under the supervision of Dr. David Hinks.)
The purpose of this study was twofold: a) to quantify the degree of lighting variability in selected large and medium retail stores, and to compare the measured area lighting to the quality of lights used in selected standard light booths, and b) to assess the performance of the current ISO and AATCC recommended color difference formula, DECMC, to a new formula, CIEDE2000 (or DE00), recently recommended by the Commission Illumination de l?Eclairage (CIE).
The effect of lighting variability was assessed using two pairs of metameric dyed cotton samples. Spectroradiometric measurements of several large department stores were taken at various locations around the store, including areas in which clothing was displayed, changing areas, in front of full length mirrors, at the check-out counter, etc. Similar measurements were made at several medium sized retail chain stores. The lighting variability was assessed using key factors, including illuminance (lx), correlated color temperature, metamerism index and color inconstancy index. Using the measured spectral data at each location in the store, and the reflectance factors of the two metameric pairs, the variability in key colorimetric data was calculated and compared to standard illuminant data. Also, as a new color difference formula has been recently adopted by the CIE, the performance of the new formula was assessed using 19 color difference sample pairs (100% polyester) around 5 color centers. The colors of four of the color centers were selected to be in regions of color space that the new formula is reported to perform better than DECMC, namely blues, dark, and near neutral colors. The performance of each color difference formula was assessed against visual pass/fail data for thirty one expert shade matchers were using each color difference pair.
Considerable variability was found within each store measured, and between stores, for each of the colorimetric and radiometric variables studied. For instance, the illumination levels varied from 50 Lux to approximately 1800 Lux and very often did not comply with the levels recommended by the Illuminating Engineering Society (IES). DEcmc values ranged from .4 to 7.5. In general, the lighting variability at the point-of-sale indicates strongly that protocols for selecting dye recipes should be developed to minimize color inconstancy between the light sources used in the store in order to insure that the color perceived by the consumer is close to that intended by the product designer.
Using the limited set of color difference pairs, the performance of CIEDE2000 was found to be the same as DEcmc with optimum correlations of 87%. It was also demonstrated that the performance of CIEDE2000 will likely vary significantly as a function of the selected illuminant. This factor brings into question the prudence of the textile industry recommending the new formula as a result of the significant variability in lighting that exists: a) at the point-of-sale for garments, b) between the store lighting and in standard light booths, and c) that the performance of the new formula does not appear to be significantly better than DEcmc.
Advisor:Dr. David Hinks; Dr. Brent Smith; Dr. Warren Jasper
School:North Carolina State University
School Location:USA - North Carolina
Source Type:Master's Thesis
Date of Publication:05/19/2004