Doug is probably *the* leading expert on what will be the most groundbreaking technological revolution of the decade, and perhaps the century. This will be the start of the era where the barriers between the digital world and the real world break down, and the human perception changes on the most fundamental level.
+Tom Fahey magic leap blew this out of the water already.. even though it's slightly different and it's not a 100% lightfield truly but still.. it's muuuch better
This is the long term solution for comfort issues like Headaches and eye muscle twitches that happen with today's HMDs not to mention basically turning them into sunglasses... Can't wait..
This is 2024 and we still don't have a commercial Lightfield HMD. Sadly, people are obsessed with high resolution displays and Lightfields cut the resolution in half. which is why Oculus is focusing more on the Varifocal approach (though those suffer from brightness loss, require eyetracking and cost alot) Rendering lightfields isn't very fast either and with the market shifting to standalone headsets they need all the performance they can get :((
if he stops working at nvidia .....he should just post this on kickstarter ......and he will raise way more money than Palmer p.s. now he works at Oculus research
Glad to see people working on VR's weak link. For me personally, as a Vive owner and VR enthusiast for years...I've come to the conclusion that the broad acceptance of VR completely hinges on this issue. Shoebox=VR failure, or at least VR will stay a niche enthusiast toy until this changes. How much sense does it make to pipe our highly evolved visual cortex through a cheap plastic fresnel lens? VR headsets in their current incarnation are just awful optical compromises. So the headsets are blurry, have variable sweet spots, need to be adjusted for each user, are variable eye-to-eye. They're top heavy, they're hot, bulky, and have very limited resolution and not to mention many other flaws (the convergence and so on). The current paradigm of HMD's; fresnel lenses and screens are a simple dead end evolution. Every other aspect of the industry is evolving fast, the processing, the software. But in the visual realm, the vital and most critical component, we've gone almost nowhere. Big lenses aren't really the solution, just pay attention to how camera lenses have only made minor strides in the last 50 years. Look at the history of the refractive telescope and how that too was a dead end evolution. Regular lenses are expensive to fabricate, are heavy, and refract light, among many other weaknesses. Light fields and some completely new visual approach, that's the vital future for VR (assuming it has a future). The HMD's have to comfortable to wear for hours on end with natural effects on the eyes. Pretty innovative approach but I'm not sold that micro-lenses are the answer. Lenses still have refraction limitations no matter how small they are. And they are an obstruction the field of view (i.e. no AR). Why not take a Newtonian approach? Toss the refraction in favor of reflection and invert the whole system? I think the possibility to have a single display that blends AR/VR a reflection approach might work better. This is a lot easier said than done, but it just seems straightforward to me that placing a lens anywhere in the field of our view is the wrong approach.
Struggled to watch because of the way he kept saying interesting wrong, for some reason it got incredibly grating. Technology looks like it has promise.
how does this only have 15k views...
Amazing presentation.
Doug is probably *the* leading expert on what will be the most groundbreaking technological revolution of the decade, and perhaps the century. This will be the start of the era where the barriers between the digital world and the real world break down, and the human perception changes on the most fundamental level.
+Tom Fahey magic leap blew this out of the water already.. even though it's slightly different and it's not a 100% lightfield truly but still.. it's muuuch better
This is the long term solution for comfort issues like
Headaches and eye muscle twitches that happen with today's HMDs not to mention basically turning them into sunglasses...
Can't wait..
This is a great talk and what a great way to prototype HMDs.
Hello this is 2019 calling. Can we have the technology now?
It has gone nowhere so far :(
This is 2024 and we still don't have a commercial Lightfield HMD.
Sadly, people are obsessed with high resolution displays and Lightfields cut the resolution in half. which is why Oculus is focusing more on the Varifocal approach (though those suffer from brightness loss, require eyetracking and cost alot)
Rendering lightfields isn't very fast either and with the market shifting to standalone headsets they need all the performance they can get :((
if he stops working at nvidia .....he should just post this on kickstarter ......and he will raise way more money than Palmer
p.s. now he works at Oculus research
Like his humor.
ingenious
Right ? :D
Forget depth of field, your eyes do that for you
+Pul5ar With a light field yes. With a flat image, they can't. That's the goal of this to passively allow your eyes to do so.
Gavin Jenkins Yeeah, that was my point.
Glad to see people working on VR's weak link. For me personally, as a Vive owner and VR enthusiast for years...I've come to the conclusion that the broad acceptance of VR completely hinges on this issue. Shoebox=VR failure, or at least VR will stay a niche enthusiast toy until this changes. How much sense does it make to pipe our highly evolved visual cortex through a cheap plastic fresnel lens? VR headsets in their current incarnation are just awful optical compromises. So the headsets are blurry, have variable sweet spots, need to be adjusted for each user, are variable eye-to-eye. They're top heavy, they're hot, bulky, and have very limited resolution and not to mention many other flaws (the convergence and so on). The current paradigm of HMD's; fresnel lenses and screens are a simple dead end evolution. Every other aspect of the industry is evolving fast, the processing, the software. But in the visual realm, the vital and most critical component, we've gone almost nowhere. Big lenses aren't really the solution, just pay attention to how camera lenses have only made minor strides in the last 50 years. Look at the history of the refractive telescope and how that too was a dead end evolution. Regular lenses are expensive to fabricate, are heavy, and refract light, among many other weaknesses. Light fields and some completely new visual approach, that's the vital future for VR (assuming it has a future). The HMD's have to comfortable to wear for hours on end with natural effects on the eyes.
Pretty innovative approach but I'm not sold that micro-lenses are the answer. Lenses still have refraction limitations no matter how small they are. And they are an obstruction the field of view (i.e. no AR). Why not take a Newtonian approach? Toss the refraction in favor of reflection and invert the whole system? I think the possibility to have a single display that blends AR/VR a reflection approach might work better. This is a lot easier said than done, but it just seems straightforward to me that placing a lens anywhere in the field of our view is the wrong approach.
Oculus does not have a 40 degree field of view jack nugget. More like 100+
Struggled to watch because of the way he kept saying interesting wrong, for some reason it got incredibly grating. Technology looks like it has promise.