3-D Without Four Eyes
Nintendo and Toshiba will bring glasses-free 3-D to portable devices
Beautifully animated figures seem to be leaping out of the game player I’m holding. Planes and cars are swooping toward me so convincingly that I’m actually flinching. The graphics are detailed; the colors are natural. I’ve never had a better 3-D experience, and here’s the best part: This handheld, multidimensional marvel, a prototype from 3M, doesn’t require me to wear those clunky, chunky 3-D eyeglasses.
New glasses-free 3-D devices are about to hit the market, and their backers are hoping they’ll make 3-D spectacles as obsolete as Smell-O-Vision. These gadgets, described as “autostereo” to distinguish them from the kind requiring eyewear, will include not only game consoles like the one I’ve been playing with but also cameras, cellphones, and tablet computers. Among the first will be autostereo 3-D TVs, just now hitting stores in Japan, and Nintendo’s 3DS handheld games console, due for release worldwide early next year.
To perceive three dimensions, a person’s eyes must see different, slightly unaligned images. In the real world, the spacing between the eyes makes that happen naturally. On a video screen, it’s not so simple; one display somehow has to present a different and separate view to each eye. Some systems handle this challenge by interspersing the left and right views; they’re called multiplexed. Others, called sequential, alternate left and right views. Whatever the approach, the displays then use optical or technological tricks to direct the correct view to the correct eye.
For example, the bulkiest glasses used with currently available 3-D TVs are active-shutter glasses. They contain a set of miniature LCD panels that synchronize with the large LCD screen in the TV. When the main screen is showing an image destined for your right eye, a liquid-crystal shutter in the left lens of the glasses makes that lens opaque, and vice versa. This sequential system switches between images meant for each eye dozens of times a second, creating a smooth 3-D effect.
It works well. In theory, at least. According to a survey of 1400 Americans by the market research firm Interpret, a quarter of gamers got headaches from 3-D, a fifth complained of eyestrain, and one in six said that they felt disoriented or dizzy after playing. In a similar survey of 2000 Americans by the market research firm NPD Group, over half said that having to wear glasses would discourage them from upgrading to 3-D altogether.
And the glasses aren’t cheap. High-tech 3-D specs cost US $100 or more, and a pair bought from, say, Sony typically won’t work with a Panasonic or LG Electronics TV.
Nintendo last dabbled in 3-D 20 years ago, with Famicon Grand Prix II: 3D Hot Rally, a disc-based racing game using active-shutter glasses. Even in gadget-mad Japan, the heavy, flickery, first-generation goggles proved unpopular. “Since then, we’ve tried 3-D many times without announcing it, searching for a way to make it a product for the mass market,” says Satoru Iwata, president and CEO of Nintendo. “With the 3DS, 3-D effects finally give a much better sense of height, depth, and width. You get a much better ability to navigate and judge distance.”
The Nintendo 3DS’s autostereo screen, made by Sharp, uses a multiplexed “parallax barrier” technology. This method lays a second layer of liquid crystals next to a traditional LCD and its backlight. This extra layer creates thin vertical strips that block some of the light and direct the remaining light alternately to the left and right eyes, creating a 3-D effect for a single viewer at a set distance, usually around 30 centimeters. More on that later.
Parallax barrier technology does have a few problems. Because the multiple layers of crystals prevent some light from reaching the user, getting to an acceptable level of brightness means cranking up the backlight, sucking up power and quickly draining batteries in portable devices. And because each eye sees only half a screen’s total pixels, the technique cuts the effective resolution in half. So manufacturers must choose between a display of standard resolution and brightness—and suffer dull, low-resolution 3-D graphics—or upgrade to a brighter, higher resolution screen that’s also pricey and power hungry.
Nintendo split the difference with its new console, bumping the brightness up but keeping the resolution relatively low. The 3DS has an 800- by 240-pixel screen that delivers 400- by 240-pixel views to each eye. While this is a step up from the 256-by-192 screen of its predecessor, the Nintendo DSi, it is just one-sixth the resolution found on the similarly sized Apple iPhone 4. Put simply, you wouldn’t want to watch a movie—or even view a photo slideshow—on the 3DS.
Parallax barrier displays also have sensitive geometries that deliver optimum 3-D effects only at a particular eye-separation distance—as close as possible to the statistical average of 65 millimeters. These displays are also tuned for a specific distance from screen to eye, with the 3-D effect fading if that distance is off by as little as 5 centimeters. This distance sensitivity is less of an issue for handheld devices than freestanding displays, explains Michael Bove, director of the consumer electronics laboratory at MIT. “We know how long people’s arms are, so the viewing distance is bounded,” he says. “And people will naturally move the thing around until it looks right.” Still, says Nick Holliman, senior lecturer in engineering and computing sciences at Durham University, in England, “If you tune an image for an adult, it may not work so well for kids.”
Nintendo recognized these issues, says Shigeru Miyamoto, designer of the legendary Mario and Donkey Kong games and now general manager of Nintendo’s R&D division. “Everyone sees 3-D images differently,” he notes. “I wanted everyone to be able to instinctively adjust settings for themselves, as easily as changing volume.” So the 3DS comes with a slider that changes the differences between the left and right images to increase or reduce the 3-D effect, or remove it altogether.
That slider doesn’t solve all of parallax barrier’s problems. If you move from a head-on viewing position, your left eye will see the image destined for your right—Pinocchio’s nose, for example, will suddenly seem to grow inward. Move a little farther and the 3-D effect might flip-flop. What’s more, light leaking through the LCD barrier layer could cause cross talk, a mixing of left and right views that creates ghosting—the blur that’s one of the biggest complaints about autostereo gizmos. Together, these effects can cause that dreaded 3-D nausea.
“Your brain is receiving ambiguous messages,” says Bove. “It might interpret them as two objects at two different depths, or as mostly a 2-D screen with something funny going on. Often, it just gives up, and you get a headache or a stomachache.”
Engineers are trying to fix these problems. In Sharp’s latest 3-D displays, the transistors that control each pixel use the company’s proprietary continuous-grain silicon technology; these transistors are thinner than ones fabricated from traditional polycrystalline silicon and they allow more light to pass through, increasing the brightness of the layered screens.
Researchers have also experimented with autostereo displays that generate multiple sets of 3-D images, either to accommodate several viewers simultaneously or to reduce the flip-flopping effect when your head moves relative to the screen. There’s a trade-off, of course: Each added set of 3-D images divides the screen’s effective resolution and requires more processing power. According to Holliman, “If you use multiview, you get too low a resolution per view for a game to be acceptable or possibly even playable.”
There are emerging alternatives to parallax barrier screens. Some are new; others are reinventions of technology that has been around for a long time.
You’ve certainly seen lenticular postcards, first popularized in the 1950s. They generate a 3-D image or a changing 2-D scene by overlaying a rippled sheet of transparent plastic on two interlaced images—a classic multiplexed 3-D system. The rippled sheet acts like a set of lenses, throwing an image of one series of image strips to your left eye and the other series to your right. Lenticular LCD screens work in a similar way with moving images—this is the technology that Toshiba uses in its recently announced autostereo TVs, which it expects to have in stores in Japan by the end of the year. However, lenticular 3-D effects reduce a display’s resolution and are visible only from specific, narrow viewing angles. That might make it pretty tough to, say, watch a 3-D football game with more than one friend. “People want to put chairs where they’re comfortable, not necessarily where the right view zones are,” notes MIT’s Bove.
More innovative are 3-D screens based on so-called wedge lens technology, pioneered by Adrian Travis when he was at the University of Cambridge, in England. (Travis moved to Microsoft’s Applied Sciences Group in 2007, bringing his technology with him.) “The wedge is better than parallax barrier and lenticular because it doesn’t throw away half the resolution,” Travis says.
If that makes the wedge sound like a sequential 3-D display, you’re right—but it’s a very special one. Instead of throwing light in all directions like a TV screen, the wedge lens shines focused beams in specific directions. If you know where the viewers are sitting, and alternate those beams precisely to their left and right eyes, you’d have an extremely efficient 3-D display, because you wouldn’t be wasting light on empty chairs. “We use maybe half the power—and in principle much, much less than half—of a normal 2-D display,” says Travis.
The wedge is basically a flat lens, twice as thick at one end as at the other. It covers the entire viewing screen, which is an ordinary LED-backlit LCD. The thick end is curved and faceted so that it lines up the beams from the LED backlights, creating a solid beam in just one direction. Activating different LEDs in the backlight creates beams that project in different directions. A prototype using a fast (240-hertz) LCD panel produces smooth autostereo 3-D, with little cross talk and no flip-flopping.
The toughest trick for the wedge is figuring out just where to point those beams. But it may get some help with that from another development at Microsoft. “Head tracking using Microsoft’s Kinect technology is the direction we’re traveling in,” says Travis, referring to a new $150 accessory for Microsoft’s Xbox 360 console that uses near-infrared cameras and sophisticated software to follow gamers’ motions.
Although Travis expects some form of the wedge to feature in consumer gadgets within two years, he admits, “Head-tracking 3-D is further out. However, I think there’s a good chance we will ultimately be able to get it to fit into a phone.”
But Microsoft’s struggles are good news for 3M.
Not only has the St. Paul, Minn.–based company managed to get sequential autostereo working well, it’s gotten it small enough to fit in the device I’m holding right now. 3M’s senior technical manager, Bill Bryan, doesn’t mince his words: “There hasn’t been a bigger improvement to the viewing experience since the introduction of color screens.” Of course, he would say that, wouldn’t he? But there’s no denying that 3M’s new Vikuiti 3-D technology [PDF, 323 KB] is something special.
I viewed the Vikuiti 3-D effect on both 3-inch and 9-inch screens, and I found it to be startlingly bright, crisp, and vivid when seen head-on. If you do move to one side, the images fade gracefully into 2-D. At a trade show last May in Seattle, 3M’s LCD business director, Erik Jostes, showed off a 9-inch model. The company was running a show reel of animated gaming characters and also some smooth, full-resolution video.
Basically, this is the 3-D display we’ve been promised in science fiction films for the last 50 years. And an early version of it, which suffers from slightly more cross talk than the 3M models, is available today, in Fujifilm’s W1 3-D digital camera.
To understand the 3M system, start with the fact that most LCD screens are transmissive; they have a backlight that shines through the liquid crystal panel to form an image. In handheld devices, the backlight LEDs are located at the display’s edge to save space. A single row of LEDs constantly shines into a plastic light guide that, as its name suggests, directs and disperses light in the correct direction—up and out of the display. For years, 3M has been creating optical films that help move this light evenly from the sides into, across, and out of the liquid crystal panels, increasing their effective brightness. The company has now come up with the Vikuiti optical film, which can direct images to one eye or the other.
This autostereo 3-D system has a light guide with a column of LEDs mounted on either side of the screen; the columns flash alternately 120 times a second. Each burst of light travels through the light guide as usual, then up into the Vikuiti optical film. Inside the film are microscopic bumps—finely engineered features that act as tiny lenses. When light from the left-side LEDs shines on them, they direct light to the user’s left eye, and when light shines on them from the right-side LEDs, the image is directed to the right eye. Flashes of left and right light are easy enough to produce, but for the screen to form an image, the LCD panel above the film must be perfectly synchronized with those flashes, showing only the left image when the left-side LEDs are on and the right image for the right-side LEDs.
The Vikuiti screen has just one 3-D sweet spot, straight out from the screen. At this point it’s definitely intended for personal viewing, not Super Bowl parties.
In order to direct light reliably to each eye at the correct viewing position, 3M needs to manufacture the microscopic lenses—the bumps—with extreme precision. The film is just 150 micrometers thick (around twice as wide as a human hair), with the lenses measuring a mere 50 to 70 µm. Perfecting this film took 3M over three years, and developing the techniques necessary to produce the film in commercial quantities took another year.
With no optical barriers or screen-mounted lenses to reduce the brightness or resolution, the 3M film creates autostereo 3-D video that looks just as sharp as traditional 2-D images. But 3M’s Vikuiti film shares some of the problems of other 3-D systems. Like the Nintendo 3DS’s parallax barrier system, the Vikuiti 3-D effect works in only one direction, typically landscape. Rotate the screen, as cellphone users are in the habit of doing, and the 3-D effect simply disappears. The 3-D effect is also sensitive to interpupillary distance.
Finally, the effect requires a high-speed (120-Hz) LCD panel to deliver those sequential frames without jerkiness. These are more expensive than standard 60-Hz LCD panels. “The panels have to catch up. The graphics chips have to catch up. And because you’re driving the LCDs harder, the power consumption will be higher in 3-D mode, too,” admits 3M’s Jostes. “If you want 3-D, you essentially have to double the processing speed.” In short, it needs the kind of high-powered smartphones and tablets that are only just starting to arrive in the marketplace. These are unlikely to add pricey 3-D features until manufacturers perceive a real consumer demand, although Chinese company Rockchip has demonstrated a prototype autostereo Android tablet.
“With the first panels, there will be a slight cost premium,” says Bryan. “There has been a chicken-and-egg situation around 3-D—how can you make a device when there’s no content? And why make content if there’s no device to view it on?”
Today, however, the Nintendo 3DS and its high-profile games are hatching chickens and laying eggs simultaneously; the industry could finally be gearing up for a handheld 3-D revolution. Unlike 3-D movies that require expensive multicamera setups, 3-D games can be developed at a premium of just 10 to 15 percent over their 2-D equivalents, according to game analysts at Futuresource Consulting. “Now that Nintendo has come out and said we’re going to do 3-D, other companies will think about moving towards it as well,” says Jostes. “Our expectation is that demand for 3-D is going to be driven by gaming.”
Steve Vrablik, a director of business development at Toshiba America Electronic Components, is thinking even bigger. His company supplies LCDs to smartphone manufacturers and showed the 3M Vikuiti film in several devices at a recent trade show. “Last year, people looked at the prototypes and walked away,” he says. “This year, they’re saying, let’s talk.”
Glasses-free (and headache-free) 3-D could be the new must-have upgrade for cellphones—like GPS location, digital photography, and music playing before it. Industry research association DisplaySearch predicts that by 2018, mobile devices will have leapfrogged televisions to become the most popular 3-D gadgets, selling over 70 million units a year. MIT’s Bove won’t predict which autostereo technology will triumph, but he is convinced that the most successful devices will all be handheld.
“There are lots of things you can do on a screen the size of a business card that you just can’t afford to do for a larger display,” he says. “Whatever your magic component is in this domain, you don’t need a lot of it.”
And without those headache-inducing spectacles to contend with, you shouldn’t need a lot of Tylenol, either.
About the Author
Mark Harris is a British technology and lifestyle reporter based in Seattle. Although wowed by 3-D View-Masters as a child, he credits the 1953 classic House of Wax, starring Vincent Price, for triggering a lifelong fear of 3-D glasses (and of Madame Tussauds). It took interacting with the latest glasses-free gadgets to finally cure his phobia. He writes regularly for The Sunday Times, The Economist, and Wired UK.