This year’s rapid adoption of liquid crystal on silicon (LCoS) visual display systems by some of the biggest players in the industry is hot news. For the longest time, LCoS technology was jealously eyed for benefits, especially spectacular brightness, ultra-high resolution and low maintenance costs that simulator operators prize. But the biggest drawback — a blurring or smear artifact associated with image motion — ruled it out for the critical demands of real-time training and simulation.

Recent technology fixes, however, have substantially reduced motion-induced smear associated with liquid-crystal-based systems. The new breed of LCoS systems entering the training world uses smear suppression technologies that range from simple shuttering techniques to unique and complex fixes patented and closely guarded by company engineering departments.

Blur on a movie screen might be regarded as merely annoying (or possibly even as a free special effect to convey speed), but on a flight simulator, it can be an unrealistic distraction that can lead to negative training. So the training industry has mostly forgone the benefits of LCoS in favor of the cathode ray tube (CRT) systems that have been industry mainstays for decades and which, although dimmer and more expensive to maintain, do not have blur issues.

The recently developed smear fixes have changed the simulation scene and ushered in a new era of LCoS-based visuals with many operating and cost benefits over rival display technologies (see chart).

So imagine the sharp intake of breath that a NASA research engineer caused when she questioned the suitability of LCoS systems for certain flight simulation applications in front of an audience of visual system specialists, many of whom had invested substantial time, expertise and money to bring LCoS to the simulator market.

Barbara Sweet, an aerospace engineer at the NASA Ames Research Center, told attendees at the IMAGE 2007 conference in Scottsdale, Ariz., in July that motion-induced blur “is a highly salient and potentially objectionable issue.” And in NASA Ames tests using an LCoS helmet-mounted display, even the lowest levels of perceived blur were above that experienced with CRT and would “likely produce very salient and objectionable differences in perceived resolution as a function of image motion,” she said.

“Other technologies offer advantages beyond CRT projectors that make them attractive,” Sweet concluded. “However, these projectors need to improve their moving picture response to validate the continued interest by the simulator industry.”

Is the crystal dream an illusion? Is it too soon to start retiring CRTs? The industry is voting for crystal.

This year, CAE, one of the most established experienced providers of training and simulation systems, made the switch from CRT to LCoS. CAE selected Barco as its LCoS projector provider, adopting Barco’s new SIM 7 as its standard projector system. As of June, all CAE simulators that were delivered arrived with LCoS systems.

Meanwhile, Lockheed Martin, one of the world’s largest defense contractors, has selected U.K.-based SEOS to provide its out-of-the-window display systems for the F-35 Joint Strike Fighter pilot training devices. Its simulator displays will incorporate SEOS’ new Zorro LCoS projectors. SEOS will start shipping its first Zorro LCoS projectors by the end of the year, and the company has a substantial backlog, predominantly for military applications, but also for planetarium and airline customers.

Clearly, the movement toward liquid crystal systems is gaining pace. “CRTs have essentially stopped their progress and CAE has elected to use Barco SIM 7C LCoS for its Medallion 6000 image generators. They are 10 times brighter than CRTs and produce better night-vision-goggle stimulation,” said Philippe Perey, CAE’s engineering director for visualization products. As a result, “The CRT workhorse is being actively phased out faster than many of us thought.”

But everyone agrees that the smearing issue can’t be ignored.

WHY BLUR?

In simple terms, LCoS displays are a cross between liquid crystal displays (LCD) and silicon processors. In an LCoS display, the silicon is an electronic back-plane with a highly reflective top layer that controls the orientation of the liquid crystal molecules. The size, complexity and weight of electronics required for LCoS systems is drastically reduced, making them easier to maintain. And LCoS systems are associated with good black levels. Black does not reflect light — a useful attribute in simulators. However, time is a problem with LCoS systems because there is a latency of about 4 or 5 milliseconds between crystals switching “on” and “off” and illuminating pixels.

This latency, or longer “hold time,” is the chief cause of blur in LCoS and LCD systems because it produces a perceived effect on the human eye — an excellent and instinctive tracking system — when the image being viewed is moving, such as an aircraft, missile or other weapon. CRT and laser displays only briefly illuminate a particular pixel, whereas LCoS and LCD systems illuminate a pixel for several milliseconds longer (hence their brightness).

The longer illumination stimulates a larger portion of the retina when the eye moves (see graphic) and is perceived as blur. This does not happen with CRT systems because of their relatively brief illumination. “It’s only an interaction between eye movements and the display that produces a blur. Eye movement with CRT jumps to the next pixel. With LCoS, the eye moves past the pixel and produces blur on the retina,” Sweet said.

Working with the U.S. Air Force Research Laboratory’s (AFRL) Mesa Visual Systems Development unit, the NASA Ames research engineers came to several conclusions.

First, reducing hold time is an effective mechanism for reducing motion-induced blur. According to Sweet’s paper, research indicated that hold times of 11 milliseconds or less can significantly reduce perceived blur, but hold time reductions of below 8 milliseconds provide negligible additional improvements; therefore, hold times in the 8 to 10 millisecond range are probably optimal.

Researchers also agreed that shuttering is a highly effective fix for blur. “On fighter air-to-air missions, applying shuttering produced a dramatic reduction in blur. It’s very promising. It appears that shuttering is a worthwhile strategy,” Sweet said.

Agreeing to a standard for measuring and comparing different technologies and displays is also important, Sweet said. “A common measurement standard must be developed to allow side-by-side comparisons of the different models and technologies of today, and to allow more cogent discussions on projector performance,” Sweet’s paper concludes. “For these standards to be usable to inform design decisions, a clear correlation needs to be established between the sensor-based measurement metrics and human perception, across varying levels of contrast, luminance and resolution.”

FRAME RATE INCREASE

More controversially, Sweet proposed doubling the current typical frame refresh rates from 60 Hertz to 120 Hertz — 120 cycles per second. Increasing the refresh rate rather than shuttering would reduce hold time without loss of luminance. But it’s expensive and technically difficult to achieve with LCoS. Geoff Blackham, chief technical officer at SEOS, said he was not aware of any 120 Hertz LCoS solutions available today, although he said he believes they will become available in the future.

“It would need double the IG capacity, and that would increase the cost,” he said.

According to Blackham, the SEOS blur fix leverages a four-panel architecture that reduces motion blur only on the content that changes, without loss of light. He said that he found Sweet’s paper interesting because it was based on relatively independent research that tallied with some of his company’s research findings. One conclusion was that as long as you reduce motion blur width to half a frame motion-blur width, it meets the user requirement. “The motion blur reduction technique we employ is about half a frame, which is consistent with that and will, by and large, meet the simulator requirements for the military,” he said.

Barco, which provided much of the hardware for NASA’s and AFRL’s research work, uses a shuttering technique on its SIM 7 projector and also applies true motion reproduction to compensate for response-time smearing effects.

Paul Lyon of Barco’s Simulation & Training Group based in Xenia, Ohio, agreed that brightness was lost with shuttering but said: “At 2,500 lumens unshuttered, we are still the best in the industry even after shuttering — we are still over 1,000 lumens.”

He also pointed out that the SIM 7 was not in the lab anymore. “It’s working and functioning and we demonstrate that on a daily basis in [CAE headquarters] Montreal,” he said. “There’s an alternative CRT business, and we are moving into that business. This will replace CRT, although we are not abandoning support of our CRT business.”

TRADE-OFFS

Ultimately, the simulator visuals industry has always been about trade-offs and fixes. Back in the 1970s, phosphors used in early CRT systems suffered “ghosting” or “tailing” issues, mostly with night scenes that called for intense calligraphic light points to simulate runway and airport lights. Different companies devised unique and often company-secret solutions, just as has happened today with LCoS smear. “It was a trade-off between color saturation and intensity,” said James Davis, a member of the American Institute of Aeronautics and Astronautics Modeling & Simulation Technical Committee.

Similarly, according to Davis, today’s LCoS fixes are about trade-offs. “You probably could not do LCoS at 120 Hertz, so you would have to go back to CRT and then you would lose all the lifecycle costs and brightness benefits that you get with LCoS. You have to ask yourself, which benefits are important and which are just good to have? But for a dome display, for example, CRTs don’t give you enough light,” he said.

“LCoS is very affordable and it’s good technology that’s being aggressively pursued. It will only get better, even if at this point there might be some high-end military applications where it will be found wanting,” Davis added. “But if pilots are complaining, two things might be happening. First, they might be using a cheap LCoS system where better systems are available. And second, separate issues could be happening that really good engineers could work a fix for.”

And the lower operating costs that the new breed of LCoS systems offers are significant. In a world of tight budgets, that is a tough fact to ignore. “Unique smear-reduction hardware makes LCoS compatible with fast jet and helicopter simulation at 60 percent lower operating costs,” Perey said.

LCoS is not zero-maintenance, “but it’s an awful lot less than the burden of CRT,” Blackham agreed. å