Nice Overview of HMD and PD technologies and markets...
Displays: Up Close And Personal
By Charles McLaughlin
Ever since the days of Dick Tracy's teleconferencing wristwatch, futurists have touted the vision of a personal display-a wearable, wireless device geared for a single viewer. Before the age of the PD finally arrives, however, one critical puzzle piece must still be put into place:a miniature flat-panel display that has the price and performance to make PDs competitive with conventional displays. Most industry leaders believe three-dimensional imaging and virtual reality will fuel consumer demand for personal displays, but the prospects for explosive growth are dim unless the price/definition hurdle is overcome. To build $200 virtual-reality goggles, for example, OEMs will need a full-color, computer- and TV-compatible mini-imager that costs less than $25-and that's no small task.
There's also an ergonomic hurdle that needs to be overcome before PDs hope to become a mass-market item:The clunky designs of current head-mounted PDs restrict their use to nerds and coneheads. For 3-D and virtual-reality applications that demand interactivity with the real world, the "head in the bucket" feel of some of these displays is a big disadvantage. What's more, the high-bandwidth interconnect the display requires means the user must be tethered to a large and expensive computer. Current systems offer little freedom of movement, and portability is out of the question. Until the imaging system can be integrated into the size and weight envelope of eyeglasses, few of us will choose to don a personal display when a conventional display can do the job.
A PD Market Taxonomy
More than 10 companies are energetically addressing the issue of more pixels for fewer bucks, devising display technologies that promise to dramatically change the price/performance equation of PDs. Full-VGA devices with improved efficiency and performance will be available from multiple suppliers next year at about the same $50 that a one-quarter-VGA device demands today, and prices could approach $10 within a few years if high-volume markets develop. However, innovation in optical design and technology to address the ergonomic hurdle is less apparent.
The personal display comes in three basic incarnations:the handheld viewer, adapted from the imagers and optics used in camcorders (See "The pocket personal display," below); the head-mounted display (HMD), a binocular device favored for entertainment, simulation and visualization; and the monocular HMD, the PD of choice for most U.S. companies that are developing wearable computers for industrial, professional and point-of-sale markets.
These one-eyed PDs target data-intensive applications where the user needs an unobstructed view of the real world while also viewing graphics and data on the computer. They demand improved optics and higher brightness over handheld viewers but, because the field of view is limited, low- to mid-definition imagers suffice.
For binocular HMDs that strive to immerse a user in a virtual experience, full VGA and NTSC definitions are a minimalist requirement and full-color, 3-D images with wide fields of view are preferred. For 3-D/virtual-reality applications in simulation and professional markets, it takes high-definition XGA and SXGA capability to achieve a wide field of view. Operators of location-based VR entertainment venues, such as Disney and San Francisco's Cybermax, say mid-definition VGA and NTSC mini-imagers don't cut it their markets. A handful of high-definition PDs are being made today, but prices are out of sight and performance not notably better than conventional direct-view or projection-display alternatives.
All in all, comparing the slate of candidate miniature displays against the conventional units they'd like to displace, the only consistent advantage PDs can boast today is their size. Conventional liquid-crystal displays and cathode-ray tubes deliver higher-definition images at a lower price, and the PD frequently uses more power than a standard LCD. Today's PD applications are more or less restricted to low-definition or monochrome displays, and early adopters have to pay a pretty stiff price even for these. A typical quarter-VGA mini-imager-a polysilicon active-matrix (AM) LCD-costs more than $50, for example-and that's too many bucks for too few pixels.
Displays at Play
Despite the challenges, an aggressive handful of pioneering companies have set their sights on commercializing PDs. Most of the early work has been in head-mounted devices, such as the i.glasses from the Seattle-based Virtual i-O, the most successful commercial HMD available. The i.glasses use low-definition polysilicon AM LCD imagers to attack the mother of all high-volume markets:consumer entertainment. They can be found on the shelves of leading electronics retailers, as well as for rent at Blockbuster video stores, and Virtual i-O has fashioned a series of strategic partnerships including a deal with Sony Corp. to sell the glasses with its entertainment systems. Another deal with investor TCI looks toward providing 3-D cable broadcasts on a PD.
The 200-odd-line definition of the Virtual i-O imager is clearly inadequate for
a genuinely immersive experience, and Virtual's HMD is priced beyond the
range that a high-volume entertainment peripheral can reasonably
demand:about $400 for a video-compatible model and $600 for a
PC-compatible version. What's more, the i.glasses are uncomfortable. At a
total weight of 8 ounces, they start to feel heavy after an hour or so of use.
Sales figures for this private company are not available, but competitors
estimate its 1996 volume at well under 100,000 units.
The second biggest player in the entertainment and computer-games market
is Forte Technology, whose VFX1x HMD, available for computers only, is
pricier still. It sells at above $800 with a low-definition, 200-line imager. The
VFX1x is totally immersive, clamping the user's head in a clamshell-style
mechanism. Heavy and bulky, it is clearly for serious games aficionados only.
According to recent quarterly reports, Forte's sales are failing to live up to
expectations, but the Rochester, N.Y., company hopes for a holiday-season
bounce in demand. In addition, Forte plans to use the Smart Slide LCDs from
parent company Kopin Corp. in future products. These AM LCDs are based
on single-crystal silicon, rather than polysilicon or amorphous silicon. Kopin,
a Taunton, Mass., display maker, has one other HMD initiative, Kopin
Innovision, based in Los Gatos, Calif. Its focus is on designing HMDs for the
wearable-computer market.
At the high end, Kaiser Electro-optics, based in Carlsbad, Calif., leads a small
group of firms focusing on military, simulation and visualization markets. Its
most popular professional-class HMDs sell for $4,000 to $10,000 and offer
VGA definition at best. A conventional workstation monitor, in contrast, can
be equipped for 3-D at a total cost of a few thousand dollars.
Coming contenders
Driven by demand for camcorder and camera viewfinders and
desktop-projection systems, the mini-imager market is expanding at an
annual rate in excess of 30 percent. Many of the companies vying for a piece
of this market, which is largely dominated by polysilicon-on-quartz
technology, are also eyeing PDs as a near-term growth prospect. That means
the current inadequacy of minidisplays for PD applications will probably not
last for long.
These companies are devising a number of new imaging technologies, all
based on substrates that are fabricated using conventional integrated-circuit
processes. The very fine pixel pitches required for high-definition PD image
sources are well within the capability of semiconductor fabrication facilities.
These new displays potentially offer more than a four-times improvement in
manufacturing efficiency at one-tenth the cost of the current market-leading
technology. Some of them aim primarily at projector applications and require
further development to work for PDs.
New Kids on the Block
The approaches fall into three categories:AM LCD imagers that are
derivatives of the leading polysilicon approach; imagers based on
micro-electromechanical devices, aka MEMs; and mini-emissive imagers,
which are useful only in viewfinders and PDs. The AM LCDs and MEM
devices, in contrast, can be used in projectors as well.
Several companies are offering miniature displays based on unconventional
LCDs. Raychem Corp., based in Menlo Park, Calif., has teamed with Japan's
Hitachi Ltd., for instance, to offer a reflective imager using a
polymer-dispersed LCD on a crystalline-silicon substrate for the projection
market. And Kopin has announced a frame-sequential color system based
on its 17-micron-pitch Smart Slide AM LCDs.
Displaytech, based in Boulder, Colo., has announced a crystalline-silicon
ferroelectric AM LCD whose 12-micron pixel spacing results in a VGA
imager with a tiny viewing area:0.4 inch on the diagonal, compared with 0.7
for viewfinder-class displays and 1.4 inch for today's commercially available
monochrome VGA projection imagers. With the Displaytech method, a
six-inch wafer will yield some 10 times more imagers and, in high-volume
manufacture, the cost of the AM substrate could be well under $10. Several
other developers plan to decrease their pixel pitch to 12 microns, which can
yield an SXGA device with a diagonal well under an inch.
Three minidisplay initiatives use micro-electromechanical devices on
crystalline-silicon substrates to produce reflective or diffractive imagers for
projection and PD applications. The well-publicized Digital Micromirror
Device (DMD) from Texas Instruments Inc., with 17-micron pixel spacing, is
now being promoted for projection systems. However, the DMD may not
prove to be an effective personal-display imager because of its very narrow
viewing angle.
Meanwhile, Silicon Light Machines (formerly Echelle), based in Sunnyvale,
Calif., is developing a diffractive display technology that could result in a very
low-cost device. An alternative diffractive design is being developed at the
Fraunhofer Institute in Germany.
As for the emissive mini-imagers, Planar Systems, based in Beaverton, Ore.,
is taking prototype orders for a 24-micron-pitch AM electroluminescent
display that emits white light and uses a color shutter to generate full color. For
polysilicon LCDs, resolution this fine is only being discussed in research
papers. Planar last year announced a strategic partnership with Virtual i-O,
which will build AM EL-based PDs for the entertainment market while
Planar fields its own PDs for the medical arena. Because AM EL, unlike LCD,
requires no lighting source, it may have an edge in size, weight and power.
Another emissive display comes from Micron Display Technology of Boise,
Idaho, which has a mini-imager that uses field-emission-display technology.
Making higher-resolution pixels and smaller imagers is not the only way to
improve price/definition, however. Both Sony and Optical Shields, based in
Menlo Park, have announced techniques to effectively double definition by
partitioning images and displaying them sequentially on a low-definition
imager. High-speed light shutters and a refractive element are used to
interlace the high-definition image.
From a performance point of view, the minidisplays that really shine will be
those that use a frame-sequential approach to color generation. With
frame-sequential addressing, each frame is divided into three fields-red,
green and blue-which are displayed sequentially in synchronization with a
multicolor RGB light source. The alternative spatial-color approach, which
divides each pixel into RGB subpixels, may use an array of RGB phosphors, as
in the Micron display, or color filters, as in the LCDs.
The two color approaches differ dramatically in efficiency, size and cost. By
way of comparison, the field-sequential Displaytech device is more than four
times as efficient as a polysilicon AM LCD with color filters. By using three
RGB LED lamps, switched synchronously with the display, the Displaytech
device suffers no inherent light losses due to the color system. The fill factor is
also three times as great, since each color uses 100 percent of the active pixel
area. The aperture is nearly twice that of the AM LCD because of the
inherent advantages of a reflective imager, in which the mirror can overlay the
underlying circuit elements that decrease the aperture efficiency of a
transmissive mini-imager. A beam splitter is required, however, which
reduces efficiency. Prototypes of the Displaytech device demonstrate
excellent, clear and bright images.
Through the looking glass
Beyond improvements in imagers, the human-factor issues for
head-mounted PDs are more difficult to quantify. The PD adds a new
dimension to the ergonomic aspects of displays. The eyestrain of monocular
and binocular viewing for extended periods, for example, has not been
analyzed extensively. Minimally, HMDs require some acclimation and are
frequently stressful when first worn. They also need to accommodate a range
of head and eye geometries:The eye must be precisely positioned, but
adjustment of the headgear to get optimal results is usually not very
straightforward.
Though a variety of approaches are being pursued to improve the
price/performance of miniature imagers, there is a dearth of programs aimed
at overcoming the shortcomings of PD optics in yielding comfortable HMDs.
Today's optics may be satisfactory for handheld viewers and monocular
HMDs, but 3-D/virtual-reality HMDs need dramatic improvement. As
smaller, higher-definition and higher-resolution imagers enter the market, the
optics challenge only becomes more severe. Optics will quickly become the
performance and cost stumbling blocks for success in the entertainment
market.
Most of the current systems use simple achromatic doublets-lenses based on
a standard optics technique for preventing color distortion-or more complex
multi-lens eyepieces. Both result in significant distortion and a small sweet
spot, or area of focus and brightness. Virtual i-O's catadioptric reflective
system-in which the viewer sees a virtual image in a mirror-offers a
significant improvement in performance and a bigger sweet spot, but at the
expense of larger headgear. The high-end Kaiser Electro-optics PDs use a
proprietary lens array called the pancake lens, which offers a wide viewing
area but is relatively inefficient in light throughput.
Some designers see the next wave of innovation coming from diffractive
optics. A new family of thin, lightweight lenses is envisioned that would
improve fields of view with little distortion. No major commercial developers
have come forward, though, and it appears that diffractive lenses are destined
to remain on the shelf for another decade as a technology with a lot of promise
but little commercial reality.
Other innovators are developing alternative optics for head-mounts. In a
promising but highly unusual HMD system developed by Optical Shields, for
example, the mini displays project an image seen only by the user on a
remotely mounted retroreflective screen. With the Optical Shields system, the
viewer in effect is watching his own private movie screen. And since the only
thing in the way is a clear beam-splitting half-mirror, he also gets an
unobstructed view of the real world.
But with few new concepts for PD glasses on the horizon, it appears that we
will be stuck with clunky head-mounts for at least another five to 10 years.
Despite the much improved imager definition and significantly lower prices,
the ergonomics of current HMD designs will prove a major limitation for
market penetration. They will work for pilots, designers, doctors, mechanics
and other professionals who get paid big bucks to wear them. They will be
embraced by gamesters and virtual-reality enthusiasts. But the vast majority
of the market will wait for the kinks to be worked out. |
