Slashdot Mirror
Google Leads $542m Funding Round For Augmented Reality Wearables Company
An anonymous reader writes: After rumors broke last week, Magic Leap has officially closed the deal on a $542 million Series B investment led by Google. The company has been extremely tight-lipped about what they're working on, but some digging reveals it is most likely an augmented reality wearable that uses a lightfield display. "Using our Dynamic Digitized Lightfield Signal, imagine being able to generate images indistinguishable from real objects and then being able to place those images seamlessly into the real world," the company teases. Having closed an investment round, Magic Leap is now soliciting developers to create for their platform and hiring a huge swath of positions.
Is it going to require a Google+ account?
No I will never let this one go.
TFA's tech is all an extension of Google Glass in a way
Google's after the defense contractor market now...developing/marketing Glass as a consumer product was an afterthought and mostly for PR, imho
Thank you Dave Raggett
The limits in augmented reality wearable technology are processing power, weight,comfort and variable focus. Something as simple as whether you can actually see reality directly through the display or whether reality needs to be captured first, processed and redisplayed with the augmentation added to it. If you can see through the display is greatly simplifies processing requirements, just requiring a portion of the wearable display to block light at the appropriate focal point and display an alternate image. This reduces the potential for far greater visual lag, something we already deal with to a degree.
Variable focus of the display is trickier as that requires distortion of the display field to cause the eye to shift focus upon various elements within the display.
So logical method would be to arrange a grid of mirrors within a curved array and either allow the normal visual field to pass or block it and reflect an alternate light source at that point. That light being delivered via, say fibre optics and light channels formed by the curved array. Distortions if the array allow for variations of focus, with the base level for the array being initially set for the specific user. Tricks with polarisation can also be used to achieve this. Still the big choice is whether to capture the environment process it and alter it or just alter a transparent display.
There is a safety factor, altering a transparent display which fails to clear is safer than altering a processed image which of course fails to no visual input and that time required to realise the loss of image is extended, reach up and take off the glasses in conjunction with the loss of possibly needed information being provided by the augmented display.
Chaos - everything, everywhere, everywhen
Your post started off good, noting the issue of focus (the actual depth cue related to focus, though, is not focus itself, but accommodation--the response of the eye to defocus) which is often ignored and people commonly only address stereopsis and motion parallax. However, it then took a couple of wrong turns. First is the issue of color reproduction. Humans are trichromats. We have three types of cone cells in our retinas, and a variant of RGB (read: color space based on only three primaries) can reproduce the full CIE perceptual color gamut. Prototype display systems that do just that exist, and I've seen more than one demoed at SIGGRAPH over the past decade. Perhaps your confusion stems from the very different situation with lighting, where the full spectral response of the luminaire matters because the resultant perceived colors are the product of the light spectrum and the lit surface spectral reflectance, convolved with the retinal cone cell response curves, and so two lights with the same whitepoint but different spectra can result in a situation where two surfaces are perceived to have the same color under one of the lights, but different colors under the other. However, this does not apply to emissive displays, where three primaries is all you need, as long as the choice in primaries is right and the maximum saturation achievable is sufficient. The second wrong turn is in regards to contrast. High dynamic range displays exceeding the dynamic range of the human eye have been around for quite a few years--just search for BrightSide technologies. I still remember the initial prototype at the graphics lab at UBC (where I did my masters). So all the issues have been individually resolved; it's just a matter of putting it all together.
"Politicians and diapers must be changed often, and for the same reason."
The variable focal depth issue was solved long, long ago by microlens arrays. Current ultra-high resolution displays make this approach very practical, as you can have 8x8 patches behind each microlens and still have decent overall 2D resolution. As for the variable opacity, just use DLP tech, as was (is?) fashionable in some projectors.
"Politicians and diapers must be changed often, and for the same reason."