SID LA Chapter One-Day Conference Highlights Applications
The Los Angeles chapter’s annual event featured presentations on quantum dots, high dynamic range, human vision, head-up displays for cars, direct-view LEDs, light-shaping technology, and smart windows.
By Ken Werner
Last February, the SID Los Angeles Chapter held its 14th annual themed one-day conference, titled “Display Technologies and Applications.” As Program Co-Chair Michael Moyer said, “The conference’s title ... was chosen to suggest that visual applications and sub-systems will succeed to the extent that their displays contribute to an effective
Moyer, who is group lead for advanced cockpit displays for the Electro-Optics System Division of the Physical Optics Corporation in Torrance, CA, continued, “But thinking of all display technologies as commodities and assuming a particular off-the-shelf display cannot be improved at low cost is a self-defeating strategy. The human visual system is not totally understood, and there are un-explored display opportunities in the complexities of the human visual experience.”
Leading Off with Quantum Dots
The conference opened with a welcome from SID LA Chapter Chair Larry Iboshi of Iboshi Consulting in Fullerton, CA, and introductory comments from the author, Ken Werner of Nutmeg Consultants in Norwalk, CT.
The first speaker for the technical program was Jennifer Colegrove, CEO and principal analyst for Touch Display Research in Santa Clara, CA. She led off with an overview of the OLED vs. quantum dot (QD) market through 2027. Colegrove predicts that roughly 25 million QD-enhanced LCD TV sets will be sold in 2017 compared to about 2 million OLED TV sets. These numbers will grow to roughly 210 million for QD-based LCD-based TVs and 40 million for OLED in 2027, Colegrove said. Even in 2021, OLED TV’s overall market share remains under 6%.
Colegrove noted that there are 97 QD materials suppliers, component suppliers, and adopters in 2017. Forty-three of these are in the US; 14 in China; 8 each in Germany and Korea; 7 each in Japan and Taiwan; and 10 in “other regions.”
I will editorialize here and note that projecting market percentages 10 years out requires more self-confidence than even I possess. But Touch Display’s five-year numbers are in substantial agreement with other sources, and the inevitable conclusion is that OLED TV is going to play primarily in the very-large-screen, super-premium segment for a long time to come. Unless, that is, solution processing becomes viable more quickly than most of us anticipate.
HDR Present and Future
Next, Gerard Catapano, director of quality assurance for Samsung QA Lab in Pine Brook, NJ, presented “HDR, Today into Tomorrow.” Catapano introduced high dynamic range (HDR) as “the latest and most innovative technology that helps film studios deliver a better expression of details in shadows and highlights to the consumer,” and he quoted the Consumer Technology Association’s definition of an HDR-compatible display as one that has these minimum attributes:
• Includes at least one interface that supports HDR signaling as defined in the Consumer Electronics Association’s CEA-861-F, as extended by CEA-861.3.
• Receives and processes static HDR metadata compliant with CEA-861.3 for uncompressed video.
• Receives and processes HDR10 Media Profile from internet protocol (IP), high-definition multimedia interface (HDMI), or other video delivery sources. Additionally, other media profiles may be supported.
• Applies an appropriate electro-optical transfer function (EOTF) before rendering the image.
The HDR standard has been endorsed by a variety of organizations, include the Blu-ray Disc Association, the Moving Picture Experts Group (MPEG), the Ultra High-Definition (UHD) Alliance, and the United Nations’ International Telegraph Union (ITU). Although HDR is currently a premium feature, Catapano asserted that in the future it will be a basic feature of TVs across all screen sizes and display technologies.
Samsung TV sets are supporting only one HDR media profile, HDR10, because it is an open standard that does not require licensing fees, and also permits customization within the profile. Since use of at least HDR10 is required by the CTA definition of an HDR-compatible display, it will be supported by all major manufacturers. Although Catapano didn’t say so, some of Samsung’s competitors also include Dolby Vision, and in its new “Wallpaper” OLED TV, LG includes hybrid log-gamma (HLG) HDR for a total of three HDR profiles.
Catapano noted that at NAB in 2016, the major encoder manufacturers were offering 4K HDR as an option, and that the major mastering and editing tool sets were implementing it. He also noted that although the CTA definition only requires the support of static HDR metadata (metadata that is constant throughout the entire film or video), even more impressive results are possible with dynamic HDR metadata (which changes scene by scene).
Society of Motion Pictures and Television Engineers (SMPTE) ST.2094-40 provides dynamic metadata for tone mapping. Tone mapping is a key technology in HDR TVs, Catapano said. It is a color-volume transform that renders incoming HDR contents for a display with a dynamic range that is smaller than the contents were coded for. With static metadata, either every scene must have its color volume compressed so that the scenes with the greatest color volume can be fit into the color volume for which the display is capable, or the most demanding scenes can be insufficiently compressed. With dynamic metadata, each scene can be optimally compressed, which in some cases will mean no color-volume compression at all. Catapano observed that Samsung’s 2017 HDR TVs “are ready for ST.2094-40.” In the Q&A, Catapano said that HDR works best with movie mode, which he recommends for general viewing, at least with Samsung TVs.
The Human Visual System
In “Human Vision and Displays,” Karlheinz Blankenbach, a professor at Pforzheim University in Germany, outlined the characteristics of displays and the human visual system, discussing both the challenges of making them work together and the opportunities for exploiting the visual system’s characteristics to improve the subjective performance of displays (Fig. 1). He noted that between the image source and the display are 1) a signal processor that performs decompression, scaling, de-interlacing, frame rate conversion, and similar functions; and 2) an EOTF (see previous section) that performs dimming and interfaces with the panel hardware and software, etc.
Fig. 1: Karlheinz Blankenbach, who spoke about human vision and displays, listens to one of the other presentations at the LA conference. Photo: Ken Werner
The first of two major challenges is to allow the viewer to read a display in an arbitrary environment (think driving into the sun and looking down to read your center-stack display), given the eye’s adaptation to bright luminance levels. The second is to represent scenes with very high dynamic range on the display. For this we need “more and better pixels,” says Blankenbach. A better pixel is one that produces greater luminance, more gray-scale resolution, more gamut, and higher frame rate.
Among many of the observations in this extensive presentation was that the standard method of determining visual acuity, the Snellen test, which is what your ophthalmologist uses to prescribe eyeglasses, is only part of the story. The human visual system’s sensitivity to Vernier acuity (whether line segments line up or not) is 10 times that of Snellen. This is rather well known and is what leads to annoying aliasing on displays that would not show if Snellen were all that mattered. What was new to me is that our color-fusion resolution is 10 times greater than our Vernier resolution, which is why we see color fringes on lines or letters on pixelated displays when we don’t see aliasing.
HUDs for Cars
Gaia Dempsey, co-founder and vice president of DAQRI Laboratories in Los Angeles, CA, discussed her company’s approach to automotive head-up displays (HUDs) (Fig. 2).
Fig. 2: Gaia Dempsey, DAQRI Labs VP, talked about her company’s holographic technology for HUDs. Photo: Ken Werner
The automotive industry is enthusiastic – very enthusiastic – about presenting both instrument-cluster data and augmented reality (AR) information to drivers via HUDs. In fact, once HUDs become good enough, there may be no need for a conventional instrument cluster at all. That’s an intriguing idea for engineers who have had to stuff more and more electronics behind the dash as time goes on.
The demanding requirements for a high-performance HUD include large field of view, high resolution, large contrast ratio, and high luminance range. (If the HUD is going to be the primary, or only, instrumentation display, it must be visible against a very bright background: sunlit snow, for instance.) To these frequently stated requirements, Dempsey added multiple image planes.
DAQRI calls its approach “Software Defined Light,” by which the company means the image is created by phase-only holography. The audience was understandably curious about how DAQRI’s dynamic spatial light modulator (SLM) was implemented, but there weren’t many details about it in this presentation. Presumably, there is such an SLM because Dempsey showed a photo of the “first-ever solid-state automotive HUD,” which uses the technology. She said the HUD “has passed all automotive certifications.” The outstanding questions may be answered at SID’s Display Week, where DAQRI will demonstrate its technology.
The presentation by Grant Wylie, senior product marketing manager for NEC Display Solutions, was entitled “Direct View LED: How We Got Here, What’s Available Now and What’s to Come.” Wylie started with some history, commenting that the first large-format, digital-signage displays were ticker-like devices based on incandescent bulbs. He showed a version of the accompanying photograph, which was taken on June 6, 1944 (Fig. 3).
Fig. 3: D-Day: Early digital displays used incandescent bulbs. NEC’s Grant Wylie showed a photo of this ticker-type display on the New York Times Building, June 6, 1944. Photo: Howard Hollem or Edward Meyer for the Office of War Information
NEC Display Solutions’ goal, said Wylie, “is to show full-color video to our audience, no matter how close or far they are from the display.” Clearly, incandescent bulbs did not turn out to be the technology of choice for video walls, but subsequent technologies have also not been ideal.
Front projection, said Wylie, lacks brightness and contrast, doesn’t function well in bright ambient environments, and must be set up so there are no obstructions in the projection path. Rear projection has its own problems: limited brightness, difficult color calibration, a deep display module, and lines between individual displays.
Cathode-ray tube (CRT) solutions (Sony’s Jumbotron and Mitsubishi’s Diamond Vision) were thick, bulky, and heavy; power hungry; low resolution; and expensive. For a time they were the only viable full-color solution bright enough for outdoor displays, but the world was waiting for a better video-wall mousetrap.
That better mousetrap was the LED display, but it didn’t come all at once. At first, only red LEDs had sufficient luminous efficiency, and green was really yellow-green. Blue didn’t exist yet. But a lot of monochrome LED displays were deployed. Now, red, green, and blue (RGB) LEDs have arrived, and full-color video walls are proliferating.
There are, Wylie said, three main differentiators between indoor and outdoor LED signs. First, the luminance of indoor LEDs is typically limited to 800 to 2,000 candelas per square meter (cd/m2), while outdoor LEDs range from 2,000 to 6,500 cd/m2. Next, indoor LEDs have ingress protection (IP) ratings of IP30/31, which means they can’t get wet. Outdoor LEDs have ratings of IP56 or better. And third, indoor LEDs have a temperature range of 32° to 104°F, while outdoor units are –4° to +122°F, and are also designed to withstand solar heat gain.
Minimum LED pixel pitch has decreased remarkably over the past few years, to the point where most makers have modules with 2.3-mm or 1.9-mm pixel pitches, and even finer pitches have been shown. But less may not always necessarily be better.
Wylie said a “retina display” – one in which the eye cannot resolve the individual pixels – requires that pixel spacing does not exceed 1 mm of pixel spacing for each 2.5 meters of viewing distance, and this is called the “optimum pixel spacing.” But 75 to 80% of the substantial cost of an LED display is in the LEDs themselves, which has stimulated makers of LED displays and their customers to think carefully about the cost/image-quality trade-off. They have come to a generally shared conclusion that a 1-mm pitch for each 1 meter of viewing distance is acceptable for most applications. So, for instance, the recommended minimum viewing distance (RMVD) for a display with 6-mm pitch is 20 feet and
the RMVD for 1.9 mm is 6 feet (Fig. 4).
Fig. 4: NEC’s Grant Wylie presented this chart of recommended minimum viewing distances for LED displays with different pixel pitches. Figure: Grant Wylie, NEC Display Solutions of America
Wylie had a few predictions for video walls. Printing of OLED displays will bring the cost down for signage as well as television, but inconsistent pigment lifespan, color shift over time, insufficient brightness for outdoor applications (despite high contrast), and a manufacturing process not ideal for very large formats are likely to limit video-wall applications. HDR LED video walls will be attention-grabbing and they will come, but new hardware and high-bandwidth digital content production (HDCP) 2.2 are required first. On the positive side, LED displays are emissive and thus capable of extremely high contrast.
Chip on board (COB) refers to LEDs being applied directly to the printed circuit board. Eliminating the current surface-mount devices will allow designers to place sub-pixels closer together. “Almost all LED display manufacturers are working on a version of this technology,” said Wylie.
In the Q&A, Wylie had the opportunity to add:
• RGB LEDs all degrade at the same rate, with most products today specified as having a lifetime of 100 K hours to half luminance.
• The market growth for LED video walls is expected to be exponential through 2020.
• The refresh rate for these walls is roughly 3,000 frames per second, varying somewhat by model.
Seth Coe-Sullivan is well known as a co-founder of the quantum-dot company QD Vision, although that company’s assets were acquired last year by Samsung. He is now VP of technology at Luminit LLC – and looks far more relaxed than when he was CTO at QD Vision!
Although Coe-Sullivan’s presentation was devoted to Luminit’s use of holographic technology to fabricate light-shaping diffusers, the organizers had asked him to start off with some comments about quantum dots and QD Vision, and he did.
On the subject of putting quantum dots directly on an LED chip (called dot-on-chip), which is very difficult because existing QD materials degrade under high heat and high luminous flux, Coe-Sullivan said, “QD Vision would have solved the heat/lumens problem in another two years, I firmly believe.”
On what true QLED should be called now that Samsung is using the name “QLED” for its new quantum-dot TVs, “since the QLED term has been co-opted: ‘electroluminescent QD,’ ” he said. Coe-Sullivan went on to say that he believes Samsung’s acquisition of QD Vision was driven primarily by its need for an alternative
to LG’s OLED.
Then Coe-Sullivan made the transition to his presentation on Luminit, a privately held, 11-year-old, 75-person company with headquarters and manufacturing in Torrance, CA. The company’s first application of holography to lighting and displays was light-shaping diffusers (LSDs), which used holographic recording to create a pseudo-random pattern that mimicked the function of a diffuser with surface relief. Since LSDs have no particles, there is no wavelength dependence and no loss from scattering sites. In addition the diffusing effect can be either symmetrical or asymmetrical – even highly asymmetrical. One of the company’s LSDs produces a pattern that is 60 degrees by 1 degree. Coe-Sullivan said that
Luminit has 50% of the market for light shaping in automotive HUDs.
The augmented-reality (AR) industry would benefit from holographic optical elements (HOEs), Coe-Sullivan said, but worldwide mass-production capacity is negligible today. Luminit is making plastic laser imagers using holography to fabricate the master, and believes it can leverage this technology for volume production. They expect to be making thousands of units per month by the time you read this. These HOEs have a transmittance of more than 90% and are approximately 25µm thick (plus the substrate).
“Volumetric HOE may have finally found its killer app – augmented reality,” Coe-Sullivan concluded. (For more about Coe-Sullivan’s work at Luminit, see this issue’s Business of Displays Q&A, which features an interview with him.)
As program co-chair, I introduced the conference’s final speaker this way: “Robert Miller has an impossibly long title: senior business manager for liquid crystals and advanced technologies for EMD Performance Materials, the North American specialty chemicals affiliate of Merck KGaA, Darmstadt, Germany. I just call him Merck’s display man in America.”
Miller’s presentation was not about displays, but his title was descriptive: “The licrivision Liquid Crystal Window Technology.” Smart windows for interior applications are not new, but Merck’s licrivision technology is intended for exterior windows, particularly the kind that sheathe modern high-rise buildings. Merck has taken substantial time and effort to develop a material that can survive heat, cold, and ultraviolet light, while still performing its required function for many years. Merck’s solution is a guest-host liquid-crystal display (GH-LCD) in which all components have been individually screened for performance and lifetime. In addition, each combination needs to be tested to ensure reliable operation, Miller said.
An architectural liquid-crystal window (LCW) must perform a variety of functions. The obvious one is switch from clear to opaque, but in its transparent state it must admit visible light, block other solar energy, and reflect interior heat back into the room. In addition to architectural applications, Merck hopes to replace existing polymer-dispersed liquid-crystal
(PDLC) products such as automobile sunroofs.
Field tests suggest that LCWs can produce 40% savings in energy for summer cooling, depending on floor area, window area, and extent of insulation.
Merck is making its own pilot facility to supply architects and builders with sample windows, but the company’s goal is to sell its GH-LCD material. The company is in active communication now with architects and glass makers, and several field installations – including the new Merck KGaA/EMD Innovation Center in Darmstadt -- have demonstrated the impact of the technology, Miller said (Fig. 5).
Fig. 5: The facade of the Merck KGaA/EMD Innovation Center in Darmstadt, Germany, uses switchable licrivision smart windows. Photo: Merck KGaA
Tabletop exhibits accompanied the technical program. The exhibitors were Colorimetry Research, Crystalplex, Gamma Scientific, Luminit, NPB Technology, RealD Me, TFD Inc., Touch International, Westboro Photonics, and Z Microsystems. Most of the exhibits were well attended during the breaks, and those exhibitors were happy. The speakers seemed very happy with the number of attendees and the lively Q&A sessions.
The 14th annual one-day conference was held February 3, 2017, at the Costa Mesa Country Club in Costa Mesa, CA. Plans are already under way for the 15th one-day conference in 2018. •