Another Dimention

Philips research

Funding is an obvious stumbling block when it comes to developing new technologies, so when a company the size of Philips gets involved, there's always a chance it may grow beyond the research lab. The company has been working on various 3D LCD monitors for some time and has shown its work to the press on a number of occasions. Although the first designs date as far back as 1995, Philips still isn't marketing the technology.
The glasses-free system uses a sheet of cylindrical lenticular (biconvex) lenses to direct images towards the left and right eye. The company claims its technology provides a more realistic 3D experience by showing several views, adding that the extra views contribute to a better perception of depth.

'The Philips display is optimised to show natural 3D content by allowing more views, up to nine or even 15, so that depth cues provided by head motion parallax – viewing a different perspective when moving your eye position – is also provided,' said Dr Marc Op De Beeck, principal scientist at Philips Research in Eindhoven. Philips' nine-view prototypes are based on 15in XGA (1,024 x 768) and 18in SXGA (1,280 x 1,024), but the lenses reduce the resolution to 341 x 256 and 427 x 341 pixels respectively. The company hinted that it was waiting until screen resolutions increased before it would consider commercialising the technology.

'As the lenticular sheet is designed as an add-on to existing matrix (LCD, for example) displays, the quality of this particular display follows the steady increase in resolution and decrease in pixel size,' said Op De Beeck.
The company also claims that slanting the lenticular lens at a slight angle across the pixels ensures the change in views is a lot smoother.

The red, green and blue subpixels for each pixel are grouped vertically, rather than the side by side alignment of a typical LCD screen. The lenticular lens is slanted so that, in a nine-view system, the first view would pick up the red subpixel from row one, the green subpixel from row three and blue subpixel from row five. The second view would start with a blue subpixel several columns along the display, then the red a couple of rows down, with the final green pixel another couple of rows down.

 This explains the drop in resolution, because if you're using nine subpixels (3 x RGB) instead of three for each complete pixel, you're dividing the amount of information displayed by three. It's a complicated technique to comprehend, but one Philips insists is essential for a smooth 3D effect.

'Displays with vertical lenticular lenses separate the views with a disturbing black band due to the non-light-emitting black matrix in between the pixels.
'In short, we win on many fronts: more views, impression of solid objects instead of discretely filling views, and more natural imaging,' said Op De Beeck.
www.research.philips.com

SeeReal Technologies

Understanding human vision is central to the 3D display movement, and a group of researchers from Dresden, Germany has been working on maximising the intricacies of visual perception in a bid to produce the first widely-available screen in the field.

SeeReal Technologies was set up in 2002 to market the D4D screens that came out of the Dresden University of Technology in the mid-1990s. The company's CEO, Lars Povelsen, claims the new impetus given to the project will propel the technology to the fore.

The D4D technology, which has been on show at IT exhibitions for years, exploits cleverness in the brain that fuses two separate images to create the 3D effect. The company's displays are autostereoscopic, meaning they send a slightly different image to each eye without the use of special glasses. The D4D monitor interlocks two images – each vertically aligned – on the screen at once. This would create a confusing jumble if it wasn't for a prism immediately in front of each column. The prism refracts the light rays to direct each image toward either the left or right eye, ensuring each image is separated enough so that only one eye will pick up each picture.
Essentially, it's aligned according to the viewer's movement, which is determined by a head tracker on top of the screen that observes the position of the user. If the person looking at the display moves, the prism refocuses the light to ensure the monitor is directing the image in the correct direction. However, this technique means that, currently, only one person can view that 3D effect at any one time.

'Multiple viewing is something that's being discussed at the moment, but it will come – it must come,' said Povelsen. Another potential limitation is that D4D only supports two views, so that the ability to look around an object is minimal – you only ever see one 3D perspective. However, SeeReal describes this as a plus point. The company says this allows its product to better support realtime applications by limiting the amount of rendering required.

Most 3D display manufacturers are unwilling to discuss the price of individual units, but Povelsen was forthcoming in this respect. A single unit would cost in the region of €15,000 (AU$28,000), although the price would obviously fall considerably with volume.

The Dresden developer had largely restricted its outlook to the German market until SeeReal was born last year. Now the company is actively looking for business outside its home nation, targeting Europe first and then the US. Visit SeeReal's Website, where there's a RealPlayer animation showing the principle behind SeeReal's 3D technology.
www.seereal.com

Deep Video Imaging

One company avoiding the stereoscopic approach is Deep Video Imaging. The New Zealand-based company's multilayer display was conceived by Power Beat International, a research and development company that claims the world's 'first intelligent car battery' among its line-up of successful patent applications.
Formed in 1999, Deep Video Imaging has already started selling multilayer displays to the military industry, but its ambitions stretch much further than that.

'Our vision is to replace single-layer displays with multiple-layer displays. That's for all applications, from watches to wall-sized desktops – everything,' said Gabriel Engel, the company's CEO.

Independent display scientists regard the approach as an intriguing method of creating 3D, and it's undoubtedly an innovative approach. Rather than fooling the brain into seeing a single image that's a combination of two, Deep Video Imaging uses two distinct screens positioned at different distances from the eye.

There's a physical space between two LCDs, so one image plane is actually closer, and one further away. The nearer display is translucent, meaning the viewer can see images on both panels, thus creating the 3D effect. The only requirement of the PC that drives the display is that it has two graphics cards or, more simply, a multimonitor card from the likes of Appian or Matrox. The card drives the display as two distinct screens.
'With multilayer displays, you have actual physical depth, so there's no illusion, which means any amount of viewers can see the depth from any angle, from left, right, above and below,' said Engel, 'Many of the other companies are doing stereoscopic. You show one image to one eye and another to the other, and then your brain is tricked into thinking it's seeing 3D. The problem with that, besides the limited viewing angle and the head-tracking requirements, is that because there's no actual de