Abstract

One of the barriers to the construction of consistent computer-based color vision tests has been the variety of monitors and computers. Consistency of color on a variety of screens has necessitated calibration of each setup individually. Color vision examination with a carefully controlled display has, as a consequence, been a laboratory rather than a clinical activity. Inevitably, smart phones have become a vehicle for color vision tests. They have the advantage that the processor and screen are associated and there are fewer models of smart phones than permutations of computers and monitors. Colorimetric consistency of display within a model may be a given. It may extend across models from the same manufacturer but is unlikely to extend between manufacturers especially where technologies vary. In this study, we measured the same set of colors in a JPEG file displayed on 11 samples of each of four models of smart phone (iPhone 4s, iPhone5, Samsung Galaxy S3, and Samsung Galaxy S4) using a Photo Research PR-730. The iPhones are white LED backlit LCD and the Samsung are OLEDs. The color gamut varies between models and comparison with sRGB space shows 61%, 85%, 117%, and 110%, respectively. The iPhones differ markedly from the Samsungs and from one another. This indicates that model-specific color lookup tables will be needed. Within each model, the primaries were quite consistent (despite the age of phone varying within each sample). The worst case in each model was the blue primary; the 95th percentile limits in the ${\mathrm{v}}^{\prime }$ coordinate were $\pm 0.008$ for the iPhone 4 and $\pm 0.004$ for the other three models. The ${\mathrm{u}}^{\prime }{\mathrm{v}}^{\prime }$ variation in white points was $\pm 0.004$ for the iPhone4 and $\pm 0.002$ for the others, although the spread of white points between models was ${\mathrm{u}}^{\prime }{\mathrm{v}}^{\prime }\pm 0.007$. The differences are essentially the same for primaries at low luminance. The variation of colors intermediate between the primaries (e.g., red–purple, orange) mirror the variation in the primaries. The variation in luminance (maximum brightness) was $\pm 7\%$, 15%, 7%, and 15%, respectively. The iPhones have almost $2\times$ the luminance. To accommodate differences between makes and models, dedicated color lookup tables will be necessary, but the variations within a model appear to be small enough that consistent color vision tests can be designed successfully.

© 2016 Optical Society of America

Improving color reproduction accuracy of a mobile liquid crystal display

Eric Kirchner, Ivo van der Lans, Francisco M. Martínez-Verdú, and Esther Perales
J. Opt. Soc. Am. A 34(1) 101-110 (2017)

Assessing the effects of dynamic luminance contrast noise masking on a color discrimination task

João M. M. Linhares, Catarina A. R. João, Eva D. G. Silva, Vasco M. N. de Almeida, Jorge L. A. Santos, Leticia Álvaro, and Sérgio M. C. Nascimento
J. Opt. Soc. Am. A 33(3) A178-A183 (2016)

Color-motion feature-binding errors are mediated by a higher-order chromatic representation

Steven K. Shevell and Wei Wang
J. Opt. Soc. Am. A 33(3) A85-A92 (2016)

Large color gamut displays with diffraction gratings

Francesco Aieta, Peter Morovič, Ján Morovič, Marco Fiorentino, Charles Santori, and David Fattal
J. Opt. Soc. Am. A 33(6) 1133-1140 (2016)

Method to determine the degrees of consistency in experimental datasets of perceptual color differences

Samuel Morillas, Luis Gómez-Robledo, Rafael Huertas, and Manuel Melgosa
J. Opt. Soc. Am. A 33(12) 2289-2296 (2016)