Managing color accuracy in next-generation displays: a human-centered, future-proof approach

Zoltan Derzsi^1,2,3 (), Robert Volcic^1,2,3; ¹New York University Abu Dhabi, ²Center for Artificial Intelligence and Robotics, New York University Abu Dhabi, ³Center for Brain and Health, New York University Abu Dhabi

The problem of maintaining color accuracy across the entire luminance range is as old as color displays themselves. The phosphor and electron gun assembly differences were compensated for electrically with cathode ray tubes (CRTs), because the nature of the triode characteristics were not fundamentally different. The color accuracy issues were made worse with the advent of Liquid Crystal Display (LCD) panels, where not only the viewing angle is a factor, but the optical response of the LCD assembly is dissimilar to what is observed with CRTs. The drive circuitry of these panels was adapted specifically to mimic a CRT, but this is often impossible due to electrochemical limitations. Other technologies, such as the Digital Micromirror Device (DMD) often used in projectors, or when driving semiconductors as light sources and displays using Pulse-Width Modulation (PWM), should provide luminance linearity by principle, but are often practically limited: by the maximum momentum of the micromirrors on the DMD chip, and by the limited spectral bandwidth of the driver electronics. In addition, if discrete semiconductors are used to generate each primary color, the difference in luminous efficiency degradation also affects the final rendered color. Here, we propose a calibration method that takes human vision into account, and assesses luminance linearity along with color accuracy in perceptually equivalent steps. The results of the measurements could be implemented at the electronic drive level, without the need for application-level compensation that would reduce the active dynamic range of a color channel. We demonstrate this concept with addressable light sources and PWM dimming with an embedded device, which can be integrated into the firmware of existing displays and lights.

Acknowledgements: We acknowledge the support of the NYUAD Center for Artificial Intelligence and Robotics and the NYUAD Center for Brain and Health, funded by Tamkeen under the NYUAD Research Institute Awards CG010 and CG012.