Name
A Model for Very Wide Gamut HDR Displays that Accounts for the H-K Effect
Date & Time
Wednesday, November 11, 2020, 8:00 PM - 8:30 PM
Timezone
UTC
Session Type
Technical session
Dale Stolitzka
Description

 


We have developed an advanced computational model for predicting the brightness achieved by HDR (high dynamic range) televisions and computer monitors, in particular those with materials that feature primary colors having high spectral purity. The technological integration of quantum dot materials and emissive pixel lighting is expanding the color gamut level to attain a degree of performance unmatched in today’s consumer and professional displays. In fact, the migration toward 8K ultra-high definition (UHD) products, in most cases, is being closely coupled with displays that use a wide gamut for elevated performance. These displays exhibit highly significant improvements in saturation and brightness due to the H-K (Helmholtz-Kohlrausch) effect, where people perceive some colors—notably reds and blues―as far brighter than any current model would predict. Our experiments thoroughly tested a new model and compared subjective findings from television sets with high luminance and a very wide color gamut, which were ideal for re-assessing the H-K effect. We achieved a suitable model and quantitative metric by blending the iCAM06 computational model originally designed for HDR image evaluation with a much improved treatment of vivid colors. Our data methodology relies on a one-of-a-kind 2D spectroradiometer that captures up to 1.4 million samples of spectral data across any image, an achievement impossible to attain manually. Our approach coupled nicely with the filter image-processing steps in iCAM06. Finally, we extracted image brightness for determining the SEMI Perceptual Contrast Length (PCL) improvement level of our model, then matched those results to subjective experiments. As a result, we have a high-accuracy model that can be used by display makers to predict set brightness as a function of color saturation and optimize luminance—notably for reducing power—in displays with spectrally pure primaries.

Technical Depth of Presentation
This talk is being presented at an intermediate to advanced level. Our extension to the SEMI D75-0118 perceptual contrast length model contains essential new technical details around color appearance models and 2D image filters. However, each perceptual effect will be discussed in a way that audience members with any degree of familiarity to the material are not left wanting for prerequisite information.

The authors note this paper can be a fit in the "Small Screens" track, and we look forward to the review committee's feedback.
What Attendees will Benefit Most from this Presentation
This paper will educate television engineers, vision scientists, color scientists, technical leadership and executives about our extension of the ‘perceptual contrast length’ metric for evaluating grey scale as well as colorful HDR images in better assessing perceptions of HDR technology. This is particularly important to technical executives and staff interested in more accurately addressing the ways in which the human visual system can process brighter and richer colors. In addition, scientists conversant with software modeling – particularly Matlab or Python – may want to access our computational models sometime over the next 10-12 months.
Take-Aways from this Presentation
• Computational models based on the industry standard CIECAM02 or CAM16 will fall short when characterizing many of the new displays that will become available over the next few years. • Accounting for the current bias favoring the Helmholtz-Kohlraush effect (H-K effect) is becoming an increasingly important imperative in television characterization. Subjective testimony reports brighter HDR highlights and stellar display performance results not quantifiable by present-day methods. • Acceleration of global 4K and 8K television market penetration is being coupled to sets having more highly saturated HDR colors and higher luminance than in years past, which adds urgency to our imperative.