It is well known that color, brightness and contrast are the primary drivers of digital video quality. For some video displays and media, and some viewing environments, these factors are even more important than resolution. Televisions and video projectors have progressed significantly in the last ten years to address improved brightness and contrast, but until recently color quality has remained relatively unchanged and digital color standards have not changed since the 1970s.
In addition, the advent of high bandwidth communication technology has allowed video entertainment media to be targeted for all sorts of new display modalities including computers and small, portable devices that generally have much lower quality capability than televisions and projectors.
The recent color improvements in televisions and video projectors, the desire to enjoy entertainment media on smaller, lower quality devices and the static digital color standards have created a unique need for new video color processing technology.
Entertainment Experience is a solution company that integrates all aspects of display type, screen size, room lighting into its products. For the best color, contrast and brightness processing to optimize the full solution quality in these changing times, Entertainment Experience believes the answer is the full multiâ€dimensional visual science of eeColor™. In this white paper excerpt, we will present the need for visual modeling followed by a discussion of video color mapping.
To illustrate visual modeling, consider a bed of flowers on a bright sunny day. The flowers look very colorful. As the sun goes down the flowers appear less colorful and have lower contrast. Did the flowers lose color? The answer is no. The physics of flower light absorption and reflection still apply, and the flowers have the same purity of color. What makes them less colorful? The answer is the adaptivity of human vision. As the brightness decreases, the perception of colorfulness and contrast decrease. The human eye adapts. Perceived brightness, colorfulness and contrast are all interrelated. If an image is brighter it looks more colorful. If it is more colorful it looks higher contrast.
A second key element of visual modeling is the perception of memory colors like sky blue and flesh tones. Human vision is highly sensitive to changes in these memory colors with viewer reaction being quite negative as they do not look accurate and real. The perception of these memory colors is not only adaptive but the three dimensional volume of memory colors is difficult to define in any standard color space across all brightness levels. For example white, brown, red, olive and yellow skin at different brightness levels cover a fairly large three dimensional color volume and can only be well defined mathematically in perceptually adaptive color space.
Visual models are not only adaptive but they are nonâ€linear. Commonly used linear light models for video displays employing 1D gamma tables and linear color matrices are insufficient. These tools do not have the mathematical degrees of freedom to optimize color quality for any of the applications discussed below. eeColor™ technology uses patent pending, multiâ€dimensional lookâ€upâ€tables to implement these models and achieve the highest video quality.
Video Color Mapping From Current Standards to Larger Color Gamut Displays
Digital color standards and display color quality have been relatively constant for years. Digital color standards such as CCIR709 (HDTV), sRGB, EBU and SmpteC are very similar and sufficient to define the color intent of digital content for display devices that have color capabilities well matched to the standards. With the recent integration of bright RGB LED light sources in projectors and televisions, however, this has changed. RGB LED’s can produce much more saturated colors than the standards due to their similarity to laser color purity.
Figure A shows representative color gamuts for the HDTV 709 standard, digital cinema projectors, cinema film, OLED displays, RGB LED displays and lasers which bound all physically realizable colors.
Figure A. Color Gamuts for Video Standards and Various Display Technologies
The larger color gamuts for OLED and the Entertainment Experience TruVue RGB LED Projector are significant. (The reader is cautioned to note that LED televisions may not be RGB LEDs but rather white LEDs with filters that have approximately the same color gamut as standard televisions.) Except for small color regions, the TruVue RGB LED Projector actually has a larger color volume than film and professional cinema projectors that are forced to use lamps to achieve brightness. This is a unique and exciting situation in color display allowing creative artists to explore much large color palettes in homes than even those available in professional cinemas.
Figure A also shows an example of a 4 color television with RGB and Cyan illustrating that the color gamut is larger than the HD709 standard, and the Hollywood Digital Cinema standard, DCI, used for professional cinema digital releases. DCI is the first color standard based on the color signals of vision, namely triâ€stimulus values, and not a particular color rendering device. Hollywood studios and manufacturers established this standard to be truly futureâ€proof because it totally encompasses laser colors meaning there is no physical display device that could ever produce a color outside this standard. Unfortunately this futureâ€proof feature causes a significant increase in the required bits/color to 12 to accommodate the larger color volume that includes nonâ€physical colors. It is also not compatible with current televisions and projectors and as such is unlikely to be adopted for consumer video distribution or capture in the foreseeable future.
In addition to more colorful light sources, additional colors beyond RGB such as yellow, Y, and cyan, C, are now being added to new video displays to further expand available colors. With these new larger color gamuts there are three simple approaches that can be implemented with the color processing tools of current televisions and projectors. They are to:
clip the larger display colors to the input standard using conventional matrix processing,
ignore the difference in input RGB and display R’G’B’ and map Râ€toâ€R’, Gâ€toâ€G’ and Bâ€toâ€B’, or
something in between.
Unfortunately, none of these produce anything close to the highest visual quality. The first approach often recommended by engineering purists dramatically deâ€saturates the available display colors which mutes color expression and prompts the question as to why we should have these extended display colors in the first place. Artists generally do not like to find out that their artistic expressions are limited by outâ€dated standards.
The second approach and any combination of approach one and two alters every input color away from the standard to a new output color. Sky blue and all memory colors like flesh tones will change. For LED projectors with the color gamut shown in Figure A, flesh tones become a disturbing red or orange color and sky blue takes on an unacceptable greenish caste. Experienced reviewers of the home theater industry have said that these projectors are unsellable with this color mapping. Actors will not accept the altered colors and Hollywood colorists and directors who spend many hours getting just the right color tones for their movies will not either.
An additional factor in driving optimal visual quality is to consider the viewing environment. It is well known that as brightness changes and the human eye is adapted to a surround that is brighter or darker than the viewed image, that perceptual contrast is reduced. This has been used in the design of motion picture film viewed in dark theaters since its inception. Including the contrast reduction caused by ambient light adding to the emitted or reflected video image, and one can see that increased brightness and color contrast is needed to produce optimal viewing.
eeColor™ technology integrates all these factors, white point selection and full display color calibration into the solutions for optimal visual quality. It uses visual models of flesh tones and other memory colors, the interrelationships of brightness, colorfulness and contrast and viewing environment to maximize the perceived color quality and preserve key memory colors. For implementation, these models can be sampled and integrated into multiâ€dimensional lookâ€upâ€tables in all TruVue products.
To learn more about color mapping for displays, creating a cinema color experience and more, click here to read the entire white paper or visit www.eecolor.com. Also, we are looking for custom electronic installers to join our exclusive Early Adopter Program, email email@example.com for details.