Apple has been granted a patent for a projection system that can enable multiple viewers to simultaneously view 3D images without the need for those dorky 3D glasses. The patent, succinctly entitled "Three-dimensional display system," is fiendishly complex, but its goal is simple: to provide "highly effective, practical, …
No good for the hone market
All right if you've got an audience of people trying to win school posture badges but not for people who actually like lie on the sofa or rest their head on the person next to them - or generally relax while watching contentless visual effects.
Did you actually read the article? The system tracks people as they move.
Ok look at the picture then
left sub image right sub image
singular not plural so it tracks A person using unspecified technology - through the back of my sofa? No capability for tilting heads.
left leg right leg=pants. with the important bits hanging out
The system tracks people as they move and assumes that their eyes are in a left-right arrangement as opposed to up-down when one is lying on a couch.
What can I say? I don't want to sound negative, but "this almost certainly won't work properly".
Tracking multiple faces, and angling projecters to bounce light off micro-sized-hemisphere coated screens into differing eyes? (Thats a hell of a trick shot, especially when you have to make several million of them 60 times a second) Fiendishly complex but uncomplicated?
I guess this is one of the ones Apple are patenting just in case it becomes even slightly possible in the next x years, like the patent they put in for a motorised refractive lens per pixel on a HD display.
To be honest though, I can't imagine this ever really working outside a lab - it just all sounds too sketchy.
Not so fast
They did not claim to have built such a system, let alone tested and proved it. A patent ca simply be for a theoretical process. This CAN work, its within the bounds of technology, it may just require a supercomputer and a few dozen sensors for now until chipsets and sensor technology catch up. The original PS3 was more than a rack full of CPUs and boiled down to a simple console in a few years.
As far as pixel altering speed, this is not ridiculously complex. DLP systems have been doing that at 1080p for almost 10 years (as a retail product, not just in labs). They've recently been doing it at 120 and 240hz. It should not be unreasonable to use a similar tech to target sub pixels on a screen a larger distance away in similar variation of speed. Yes, much greater accuracy, and probably 400+hz is required, but it should be possible.
I wasn't under the impression they had built or tested it. I know patents can be applied for without any proof of concept, but I don't think they should be able to.
I don't envisage the main problem with this as being the processing power required, as thats just a few million basic math operations - still a heft load especially if those calculations need to be done per pair of eyes, as the article insinuates. And yes, it can work - the basic physics is sound.
I think the main problem will be the angling of the pixels themselves. It's not as simple as moving the entire pixel field slighty, each pixel needs to be moved by a differing amount, and possibly even sent to several different sub fields, again per pixel. In a non-lab environment, the projecter will need to be aware of it's location in relation to the screen in incredible detail, in order to callibrate itself accurately enough to pull off the exact angles required. It would ony take the screen moving or flexing by sum-millimeter amounts to cause a complete loss of the 3D effect in that area.
I would also question whether the beam of light each pixel comprises would be anywhere near precise enough to even survive the journey without dissipating to the point of terrible cross talk, or loss of the effect all together. The angles required to bounce a beam of light precisely to 2 targets mere centimeters apart over 10 feet are minute, and this is twice as problematic as current "glasses free" 3DTVs, as the light has to make 2 journeys. The beam of light would have to cover a tiny fraction of the surface of the pixel to be reflect accurately enough, most likely easily less than 1%. This would have to be more similar to an array of a million lasers than a current projector.
Maybe, in time, this could be made to work outside of a lab, as with pretty much anything. I would be suprsied if a much simpler and more effective method hadn't been discovered by then.
This is so retarded i cannot believe they granted this ....
Why the fuck cant they just turn around and say Unless you can make the technology you cannot patent it.
Or at least say when granting a patent you have 2 years to manafacture it and if not manafacturing you lose the patent.
At one time in the UK you used to have to deposit a model of your device with the Patent Office Library.
Eventually they made a new building to house them - The Science Museum at South Kensington.
In those days a Patent gave you a limited monopoly as a reward for publishing the details of your invention.
I can't help feeling we've lost our way somewhat.
"In those days a Patent gave you a limited monopoly as a reward for publishing the details of your invention."
What do you think this is doing? What are the above if not the details of the invention? (And trust me, there's a lot more than is in the Reg article.)
Also, it'd be kind of hard to put 'a model of your device' somewhere if your patent is on *part* of a device. Or if you're not a large company, and your patent involves something that requires significant investment to implement, but is a solid idea.
The patent is not about judging practicality, or marketability; it's about judging novelty.
The point being made was that it used to be the case that 1) models were deposited - leaving a lovely historic record (I love the science museum!) and 2) patents used to be about SHARING technology in return for some ccontrol over how it was used, compared to now where they are used to block others from bothering to research that field.
You cannot patent a wild idea
Each nation has its own standard of proof of possibility.
in the UK this is that an embodiment of the invention must be described sufficiently for one skilled in the art to make it. This satisfies the primary purpose of the patent which is to advance technology for all to enjoy.
I don't know what the rule is in the US. Tell us someone.
follow the dollar
But think of the children/lawyers. How could they afford their Porsche if they can't put a tax on the rest of society for doing basically anything? Don't worry about the fact that the laws are largely written by lawyers for lawyers and everyone else a distant second.
That's right thought-crime bytch, you can't patent an idea good-or-bad. Ideas/concepts/algorythms ... belong to the COMMONS ... every last one of them.
You can only patent a concrete object with novel performance aspects. -- a physical machine. Tuff tit about the thought-crimes I know how it must aggravate you to lack power over anothers pure thoughts.
because in order to actually feel safe in investing the money to R&D an idea, including what might be many years and many millions of dollars, we have to ensure someone ELSE doesn't take the same idea, make it work faster using more money invested, and cut us off at the patent line.
You patent IDEAS, not products.
Nice idea but...
Let me get this straight... they're intending to send a different image to every single eye looking at the screen? 4 people would therefore require 8 different images - I can see a processing bottleneck appearing.
Now, to get a beam of light to be so focussed that you can target a single eyeball with $x pixels, I can only imagine that they're planning to use laser beams, however I can't imagine the sharks keeping still enough to maintain the beam angle!
This will require a VERY rigid mount/screen... unless maybe they incorporate markers in the screen so the projector system can track the screen as well as the viewers?
Sounds like interesting tech though. Unfortunately, the most accurate name for it would be 'retina display', but some muppets have already used that phrase elsewhere!
Its the SAME image, just to different sub-pixels based on where you stand in order to ensure the correct eye is receiving the correct image at the right time.
It's not on-the-fly sending 8-10 different renders, its sending the same render to multiple subpizel locations. The decoder for the video does no extra work, just a lense (though, there would I assume be 1 lense set for each pixel, enabling subpixel targeting).
That would only work if the projector was the exact same size as the screen, mounted entirely dead on (meaning no-one could sit in between), and magic.
If the projector isn't set up in this way, then you can't just move the entire pixel subfield, each pixel would need to be individually targeted. Draw the raytyrace if you don't believe me - current pixel subfield technology does not help in the slightest.
I've also realised since, the effect this would provide would be utterly horrible unless you sat at the exact "right" distance away from the screen - the projector cannot adjust the convergance for the viewers, as the convergance required is dictated by the projector to screen distance.
You might be able to bounce a point off a curved surface onto an accurate target, but attempting to bounce a beam is going to result in the worlds most expensive glitter ball.
"highly effective, practical, efficient, uncomplicated, and inexpensive autostereoscopic 3D displays"
So I'm okay as long as I make it expensive or really complicated?
Yep, but if I make it expensive *and* really complicated I have a different version to yours as well.
I'm highly sceptical. While it sounds ok in theory the critical part seems to be determining exactly where the observers eyes are. I'm assuming a reasonably high level of precision is required here for a believable 3D image to work and any imprecision could break the whole effect or, even worse, make the user quite sick.
Frankly I don't think it's possible, expecially if you allow the observers to move about.
I should have patented my idea then.... :)
Thought about this a while back when 3d came back in and was thinking what would be a way to do this without glasses, but didnt think it would be possible due to requiring the precise location of the eyes, and the extreme refresh rate needed (would increase for each viewer added). This would be only for large screens, smaller, I thought about and that would be a standard LCD/PLASMA display with a secondary screen, directly attached infront, and aligened (could be LCD again??) which would then deflect the light coming from the screen to the person, but again requiring higher refresh rate for the number of viewers. It would then have all the required hardware in 1 box, much neater for smaller screens.
But it's still a solitary experience.. I can't see how the system could work with more than one observer
so let me see
Am I right in saying that a patent on this idea now will stop anyone thinking about implementing it or a similar solution, and Apple don't even have to have any idea of how to implement this using current technology?
Doesn't this stifle technological progress?
In a word....
The fact that I can't see it working (even poorly) with this decade's technology for more than a handful of folks under near-ideal conditions and using a whole rack of processing power to do so means that the most you're likely to ever see is one demo rig in a children's science museum in California, someday.
Meanwhile, I'm going to put out here, for all the world to see, the simple plan of twist-polarization CONTACT LENSES for both corrective and non-corrective use. Since somehow the world has gotten the idea that glasses are such a TERRIBLE BURDEN, this eliminates the glasses, and should be compatible with the RealD systems already in use in many theaters.
As a wearer of glasses
It's the stupid paper / plastic designs that are the issue. Make them out of carbon fibre or a light-weight metal and you hardly notice them (in a future where everyone wears them for 3D and some for correction). But I agree. If you never wore glasses before, you will notice them. Perosnally, contact lenses are very annoying and I would opt for glasses systems being developed alongside them to give choice.
But even that is unecessary. What's WRONG with 2D tv? It's clearly very entertaining and the 2D element keeps it seperated from the reality of your living room. (horror movies would be a lot more scary if the zombie could reach out of the screen, but not everyone would like that...)
I don't feel 3D adds to the medium of TV.
As a very short sighted wearer of glasses...
... I'll opt for Contact Lenses any day, but I wouldn't want to swap to a different pair just to watch TV!
Stereo is not 3D
No matter HOW you do it, Stereo viewing doesn't work for nearly 20% and gives eye strain to everyone.
Maybe all LCD TVs will have active shutter "3D", because it costs nearly nothing to add. Cost is in the glasses. But other forms of so called 3D other than lenticular handheld are unlikely to ever be very popular.
Still nice to see Apple try to innovate. Even it is prior art dressed up to look new. The USPO should chuck this out. It really fails originality.
Yes it is.
> Stereo viewing doesn't work for nearly 20%
Stereo vision gives 3D for every animal with two overlapping fields of vision. That includes all humans with two working eyes.
You appear to be talking about dot autostereograms - but that isn't what is described in the article.
Stereo vision gives 3D for every animal.
That's not entirely how 3D vision and Depth perception works,
If one eye is enough weaker (almost no-one has matched eyes), then pure stereo vision doesn't work. I.e. if one eye sees box and other eye sees animal, with perfect stereo vision the animal is in the box.
It doesn't work for significant number of people.
Also REAL 3D has depth. The eyes focus at different distances according to the object. With FLAT Stereo vision that doesn't work. You get a headache.
So not only is this not going to work for some people, as even Active shutter doesn't work (which is surprising, I thought active shutter might work for slightly "lazy eye", but apparently doesn't)
But it's horribly complicated and not really an original idea.
The guy is wearing Stereo goggles obviously
3D or not 3D
To clarify, stereopsis gives depth relative to your point of focus, which could be centimeters from the end of your nose or or many meters away. Its great for picking out an object camoflaged against its background, which is why predators have forward facing eyes.
Really, it is.
> That's not entirely how 3D vision and Depth perception works
Yes, it is.
You have two eyes.
Any depth perception effects you have are a result of the images coming from those two eyes.
Whatever you might be trying to say about various methods used to attempt to create a 3D display is irrelevant - whatever you see, it comes from a stereoscopic view created by the two eyes in your head.
Any alternative to this proposition would mean that you have 3D view with a number of eyes other than two - and you don't, even if you can infer something about distance-to-object from its size.
Stereoscopic images are only part of the whole picture (pun intended).
There are two other features of depth perception that are also important. One is the act of focusing the eye for the correct distance, and the other is that each eye moves to focus the object of interest on the center of the retina, where there are more light receptors.
If you just use separate images beamed into each eye, you can have the correct image for an object close to you. The brain says that it should be focusing close, but in actual fact you need to focus on the screen further away. Ditto the depth related parallax issue of the two eyes. These are what causes people to have headaches.
If you want a demonstration, try the following.
Hold both one hand about 6" in front of your right eye. Hold the other at arms length in front of your left eye. Try to focus on both hands at the same time. Don't do this for too long, or else you will get dizzy.
Then find someone you know quite well, hold your index finger about 2' from their face and ask them to look directly at it. Watch their eyes, and move your finger to about 6" from their face. You will see that they go slightly 'cross-eyed'.
Stereoscopic vision together with both of these other effects are required for the brain to correctly determine depth. If you only get one of them, some people can ignore the fact that the others are missing, and some can't.
Oh, and one more thing. With proper depth perception, moving your head will alter the image due to parallax (again), but the current 3D TV cannot do this. If the eyes are tracked correctly, this system *could* be able to do it, but I agree with most people that this is unlikely using any tech. we are likely to see soon.
Re: Eye matching
Mismatched vision is only going to become more common. Vision correction in anybody aged 40+ frequently intentionally mismatches focal depths so that one eye dominates for distance and the other for reading.
> Stereoscopic images are only part of the whole picture (pun intended).
They're actually pretty much the whole of the picture.
> One is the act of focusing the eye for the correct distance
That is a very short-term effect, if it happens at all. This is why contact lenses are often prescribed for monovision as people get older - the eye adapts *very* rapidly to changes from "the norm" in terms of focussing. This is why you can put on a pair of glasses with only very temporary disturbance to distance estimation, even if the prescription is really quite strong...
> the other is that each eye moves to focus the object of interest on the center of the retina
And, by positioning each entity within the image, the angle between the eyes is set for that entity, thus creating the 3D view. That is exactly how stereoscopic vision works - and it matters not whether the image is real or synthetic. The eye moves to place the object correctly...
> If you just use separate images beamed into each eye, you can have the
> correct image for an object close to you
You can also have the correct images for an object far away from you.
Here's an experiment to show you how it works: pick an object - any object, at the distance of your choosing. Can you see it in 3D? Good. You're now interpreting that 3D image according to two images - one in each eye. It really is that simple.
> Try to focus on both hands at the same time.
That has nothing whatsoever to do with 3D vision. It is how the human vision system ignores items in the vision path. And it's a temporary effect - wait a while, and you can focus on both.
> Don't do this for too long, or else you will get dizzy
Nope. Do this for a long time, and you acclimatise to it.
> You will see that they go slightly 'cross-eyed'
That's how stereoscopic vision works. The angle between the eyes is used as a reference. That angle is set by correlating the two images, and it happens whether the images are real or synthetic. It's also the reason why aliasing causes strange effects - which is why Ford Escort headliners used to look so odd, and why random dot autostereograms work.
> Stereoscopic vision together with both of these other effects are required
> for the brain to correctly determine depth
Absolutely untrue. Eyeline angle is an *effect* of stereoscopic images, and focussing is a feedback loop - it alters over time.
> If you only get one of them, some people can ignore the fact that the others are missing
Anyone looking at something with two working eyes is using sterescopic vision. That is the beginning and end of it. Look up the term if you don't believe me - it's "stereoscopic vision". Everything else you say implies some deep-seated misconception about how human (and most animal) vision works. But don't take my word for it - look it up. You will see that the separate images are all that is necessary.
> the current 3D TV cannot do this
It's a good job I wasn't talking about 3D TVs, then, isn't it? I'm talking about stereoscopic vision - the sort I'm using right now to look around me - which isn't reliant on any display technology. .
3D or not 3D again
Its a good job deer and rabbits don't have eyes on the sides of their heads then, or they would always be bumping into trees.
Seriously, stereo vision is a latecomer in evolutionary terms. One eyed people manage fine without it and they don' t think they see the world in 2D.
Nothing I have said indicates that stereoscopic vision is not a major part of depth perception, and I don't think that anybody would think that I said anything different.
It's just not the only thing that matters. You can ignore that the other effects exist if you want to, but that would not alter the fact that they do exist.
I did look it up before posting. Maybe you would like to look up the monocular cues Accommodation, and Blurring, both of which are real documented features of depth perception, along with Motion Parallax. Wikipedia will appear top of the search list, but it is not the only reference for this on the 'net.
The universe is vast and complex. Anybody believing that we fully understand any part of it is either a fool, or deluding themselves.
So what if stero vision isn't quite a complete system of depth cues?
If you'd been around in the days when artists were starting to use perspective properly in their paintings would you have been telling them not to bother with that too? Proper use of perspective was a step forwards. Stereo based 3d is a step forwards. No sense in not taking one step just because we can't take them all at once.
@stucs201 re: one step forwards.
I'm not sure what you are getting at. I was responding to the comments that stated that some people get headaches, which they do, and I was trying to explain why that was.
I don't buy into 3DTV or other 3D displays of any type at the moment or the foreseeable future, but that is my opinion, and I don't try to force my opinions on anybody. I may discuss them especially with people who I believe have not seen all the sides of a problem, but that is what dialogue and conversation is all about.
To: Peter Gathercole (no nasty twitter style addressing from me)
My comments weren't really directed to you, though technically I think I clicked reply on your post as the last in a thread. The comments were directed more towards the whole 'its not exactly like the real world and therefore crap' argument that gets raised here everytime an article on 3D films/tv/displays appears.
Eyes to the right
If its tracking from behind then how does it know the distance between the users eyes? Different people have different eye separations - naturally and when turning to the left or right relative to the screen. I think there would be some real-life problems with this technology which I assume has not been actually tried yet.
Maybe better to use a front tracking Kinect, Move type thingy to see where the users eyes are relative to the screen. That idea is hereby disclosed on the reg ;-)
Man, that looks seriously expensive and breakable, but could get me interested in 3D if it actually works
Does this mean...
That they think they can do this or are they being cheeky and just describing what would need to be done. If it is the latter then I'd like to patent Teleportation through the use of a device that will allow me to do this.
Re: Does this mean...
"That they think they can do this or are they being cheeky and just describing what would need to be done."
Well, given that they're merely describing the basic optics required for people to see depth in projected images, it's what academics might call a review paper, or what people in other lines of work might call a position paper.
Combining basic observations of optics with widely considered "here's how you'd direct the rays" observations (and I wonder whether already deployed lenticular technology is referenced, because that's already used to do this kind of thing) and head-tracking (also widely understood and demonstrated) isn't worthy of a patent, but as usual this is a company wanting to have a monopoly on "this kind of thing".
Given the tendency of Apple and similarly-minded companies to erect a toll booth in front of any kind of business they think they must "own", I suggest that such patent-waving tactics be referred to as a "Stevejob".
You can get patents for unrealisable schemes
See GB patent 1310990 - the infamous British Railways patent for a flying saucer. It was actually granted in 1973 but lapsed in 1976 due to non-payment of the renewal fee:
Paris? Well, you haven't got a Barbarella icon so she'll have to do!
Can anyone ion the know explain how long one ca hold a patent on something that's a mere idea, without even a hint of a demonstrable technology or product? If someone were to invent something propperly that filles some of the "boxes" in the diadram with actual real technology that works, would Apple have anything on them?
By moving to America, land of the free and the litigious!
However it's more than an idea. You couldn't patent "a black box that does some magic and presents a 3D image", you would have to explain what it is the black box does.
Quite often ideas exist before the technology can be built (due to manufacturing constraints at the time etc). But if you come up with an idea and try to get it funded to chase the technology required, your investors will really want to be sure you "own" the idea in the first place.
given that a patent lasts for 25 years they essentially lay claim to anyone trying this, even if they make absolutely no effort whatsoever to pursue it themselves?
Yep. Isn't it great how our patent system benefits society and encourages innovation? I am sure the founding fathers would be proud of the Rube Goldberg special interest driven patent system we have in place today.
- Updated Zucker punched: Google gobbles Facebook-wooed Titan Aerospace
- Elon Musk's LEAKY THRUSTER gas stalls Space Station supply run
- Windows 8.1, which you probably haven't upgraded to yet, ALREADY OBSOLETE
- Mounties always get their man: Heartbleed 'hacker', 19, CUFFED
- Android engineer: We DIDN'T copy Apple OR follow Samsung's orders