At first, the idea of stacking LCD panels sounds utterly crazy — I mean, aren’t LCD panels opaque? In actual fact, while the gaps between the pixels are opaque (that’s where all the control circuitry goes), the pixels themselves are fairly transparent — otherwise illumination from the backlight wouldn’t be be able to shine through. Realizing this, Nvidia decided to test whether you could increase spatial resolution (i.e. pixel density) by quite literally stacking LCD panels on top of each other — and sure enough, with a bit of software wizardry to account for the rather odd setup, you can.
Nvidia started by buying two fairly standard 7-inch 1280×800 LCD monitors, yours for just a few dollars from the nearest Chinese supply chain. The LCD panels are removed from their casings, and the backlight is removed from one panel (which will eventually be the front panel in the stack). Due to a conflict in polarization (the rear polarizer on the front panel is crossed with the front polarizer on the rear panel), a quarter-wave film is placed between the two panels. The two panels aren’t placed exactly in front of each other, but rather with a slight (quarter-pixel) offset.
Virtual Reality technology is widely accepted in the healthcare field. From diagnostics to treatment to practicing difficult surgical procedures, healthcare institutions are incorporating the technology in all the facets of the industry.
This offset is important as it means that each pixel on the front display actually acts as a “shutter” for a cluster of four pixels on the rear panel. This is why the screen offers quadruple the spatial resolution (multiplicative), rather than just double (additive). It takes some special software to drive this new cascaded display, of course, but Nvidia is obviously well suited to such a task. As you can see in the images throughout this story, this setup really does add a significant amount of resolution. Both the front and rear panels can be used in unison to provide refresh rates above 60Hz, too.
The primary application that Nvidia appears to be targeting with this new approach is head-mounted displays (HMD), such as the Oculus Rift. The Rift, by virtue of it being a cheap device, uses a fairly low-resolution screen — and when the pixels are that close to your eyes, low resolution can be really unpleasant (it’s a lot like the low-resolution arcade machines, if you remember those). Rather than quadrupling the spatial resolution by putting an ultra-expensive 3840×2160 (4K) display in there, why not just stack a couple of cheap LCDs? By cramming more effective resolution into a HMD, you also have the opportunity to spread the pixels out a bit, to provide a wider field of view (human vision is closer to 180 degrees, rather than the 100 degrees offered by the Rift and similar HMDs).There are some caveats to Nvidia’s approach, of course. Stacking one display in front of another does decrease the overall brightness of the image — but this is more of an issue for laptop or smartphone, rather than a HMD where the image is very close to your eyes. There is also a fair bit of distortion (mostly corrected in software) and limited viewing angles (not correctable) caused by separation of the two panels — again, though, the impact of these issues is much reduced in a HMD. [Read: Virtually perfect: The past, present, and future of VR.]
The First Head-Mounted Displays – The Telesphere Mask and the Headsight. You might think that strapping a display on a person’s head is a relatively new idea, but it is not. The first head-mounted displays were developed as early as the 1960s. The Telesphere Mask was the first example of a head-mounted display, which provided 3D stereoscopic and wide vision with stereo sound. However, the device lacked certain immersion, because of it being a non-interactive medium. In 1961 two Philco Corporation engineers, Comeau and Bryan, came up with the Headsight. A head-mounted display, much like the Telesphere Mask, the Headsight featured magnetic motion tracking technology, which was connected to a close circuit camera. While the goggles can be named a precursor to modern virtual reality technology, they were not developed for entertainment purposes. Instead, they were developed for the military with the idea that a person would be able to immerse themselves in the remote viewing of dangerous situations.
For a lot more details on how Nvidia created a cascaded display (and a less successful attempt at doing the same thing with two LCoS pico projectors), read the research paper. The video (above) is also well worth watching.