Slashdot Mirror
3D LCD Screen without Glasses
Nomikos writes "
12" LCD screen giving a 3D view - no glasses needed. "Designs
created, float between the screen and the user."
They don't give any specs, only mention a head/body tracker (clip-on emitter, screen receiver) which makes it only a tiny bit plausible IMO.
Anyone know of a technology which'd make this possible? "
Anyone know of a technology which'd make this possible? "
yes...but even worse...was Mission Impossible (Improbable) with the cumputer harger asking for a Intel RISC AI chip with xxx Mhz. Now that was funny.
It is possible to create artificial holograms by inputting an FFT of the image into an LCD and then using the LCD to modulate a coherent light source, but the link you posted is to a YASD (Yet Another Stereoscopic Display). In this case, they select alternating LCD rows with a holographic element, but there is no holography. The LCD contains an image, not a transform.
/. readers. The bottom line is that LCD stereography is cheap, easy, and works for one person, while LCD holography is still in development, the image occupies a volume, and works from multiple (ideally all) angles (thus multi-user).
:-)
I'm sure the distinction is lost on most
I'd post a URL, but I'm applying there for a job and they might not appreciate being slashdotted, yet.
Here is a link to some people who are trying it out: 3D LCds
Randy Maas
randym@acm.org
It was Dresden University of Technologie. Have a look at http://www.inf.tu-dresden.de/D4D/ for more Information.
Jean
Here a link for a quick overview of how it could be done. http://www.cs.berkeley.edu/~jfc/MURI/LC-display/sl d008.htm
Yikes! Perhaps it could be done, but the required sweep linearity would require a costly precision deflection yoke (like a Celco), most likely, to begin with. The alignment between the CRT pattern and the lens array would have to be held to really-tight tolerances, maybe .002" or so. It would be not only linearity, but size and horizontal position. Then, there's faceplate optics... Hopelessly costly without some corrective feedback; an engineer's nightmare; sorry!
Having worked in the field of 3d displays (computer generated holography at MIT), I can fill in some background on 3D display technology.
First here are some pointers:
http://www.media.mit.edu/~h alazar/autostereo/index.html (A basic 3d display description).
http://spi.www.media.mit.edu/groups/spi (research on different 3D displays at MIT)
Whenever you see a display that looks approximately like an LCD display, it's almost always based on what's called a parallax barrier (stripes that keep some pixel columns from being viewed from different directions), a lenticular lens sheet (a bunch of lenses that directs light in different directions) or holographic elements (HOEs, that do the same thing). Each is a variant on the same theme. By tracking the viewer, you can shift the "sweet spot" where the two views appear at a location an eye-span apart so the viewer sees 3D. This limits this type of display in almost all cases to one viewer only. If you can track in depth, you can also change perspective of the rendered views. The technology's been around a while (Ives in 1928 or so); the implementation details of the electronic versions (lens alignment, fragility, cost, rendering speed, tracking, blur) are the (very) hard details.
The "girating image" display that some people have referred to used what's called the "kinetic depth effect." Close one eye, move size to size quickly. The parallax gives you depth information. Put both views on one screen and look at it, and you get the same kind of information (only a much weaker effect, since you don't have a motor connection between your position and the view you're seeing).
The MURI display from MIT/Berkeley projects images into space using higher quality lenses. Same basic idea, but very different optical components and lens train. Definitely not the same people. The MIT electro-holographic display, by the way, is also a completely different technology (holographic, i.e., based on diffractive optical elements).
There are many other types of 3D displays. All use mostly the same principles. Most have been pedalled by charatans sometime in the past. Just like in computing: 1) if something seems like magic (or is sold that way), be skeptical, 2) if people won't tell you how it works, be skeptical, 3) if technology is being sold by calling it something it clearly isn't ("a colour hologram", where holographic displays are very different beasts), be skeptical.
Shift the picture (a couple of pixels) instead of the lens & you're home!
The episode aired on That's Incredible about the 3D system (that jiggled a bit) would work even for a person blind in one eye. I couldn't believe it myself, so I closed one eye, and it worked! I was stunned!
Ah... reminds me of that Lost World movie when the little girl is searching for the correct file (or executable?) by flying over a vast landscape of icons and maps... and what did she say "Oh, I know this! This is a UNIX system!" ;-)
Since then have I had nightmares about that dodgy windowmanager... seemed to take ages to do _anything_ since you had to _fly_ there b4 doing your stuff.
Having said that, one can probably come up with a better design than that... Just thow some HCI on it..
I've seen a couple different demonstrations of this technology at SIGGRAPH. They just cover an LCD display with that sort of ridged material that they use on the covers of the big childrens' books that have 3-D covers. (Or see the covers of the Pixar _Bug's Life_ or _Toy Story_ books to see what I mean.) It's a little blurry, but it requires no external viewing apparatus and I imagine it must be dirt cheap to produce.
Of course, I mean't alternating columns in the
previous post. Ack!
Heh, and that program actually exists!
Its called fsn or something. Its for SGI machines...
Hmmm.. they probably got the idea to use that program in the scene when they were looking at the 3d animators' machines....
There are any number of technologies that can do this (from viewing a reflected screen off an oscillating mylar mirror-film to lenticulars to making the screen itself oscillate -- for that matter, what's wrong with glasses?).
The real issue is the same thing that killed 3-D movies the umpteen times they've been revived: content that uses the technology in an interesting enough manner that would attract enough use to bring down costs and perfect the technologies.
Maybe try checking into the Pulfrich Illusion.
===
Nicholas Bodley
nbodley@tiac.net
NO!!! YOU ARE NOT THE ONLY ONE! I actually have (somewhere at my parents house) a videotape with this segment on it - the 3D is still there, last time I viewed it (about 8-10 years ago). The only thing I remember is that these guys weren't backyard experimenters, but rather a couple of people (fauculty, I think) at some university (I know that doesn't help much).
Yes folks - this is a true story - it was a little black box that had something to with alternating scans or something - it jiggled a bit, but the effect was amazing. I found a book at my library that discussed this method - it has something to do with eye latency as well (similar to the single eye dark lens method of 3D)...
That is all I know...
After reading and posting, seems that's one way it could work. Head tracker + software calculates how much to shift the lenticular screen; doesn't have to move much.
===
enby
Nicholas Bodley
nbodley@tiac.net
There's a research institute here in Berlin that's working on something like that (Autostereoscopic display and 3D GUI). http://atwww.hhi.de:80/~blick/
ME TOOOO!!!!!
dam, I've been trying to find out about these guys....
I'm VERY intrested in getting a copy of the tape.
All the alta-vista searching I've done, and back issues of some video/av/ smpte journals/ readers guide has not turned up anything.
Please at least view the tape again and post all the info. I don't remember their names, but I kinda remember that they had a van with the equiptment, and were at a NJ university.
I think it is unlikely that this was the Pulfrich effect as the Pulfrich effect relies on the user putting something over one eye (like a darkening filter from a pair of sunglasses) and is still about the delivery of different images to each eye. The claim that this effect could be see with _one_ eye suggests that this is something else entirely.
I am probably wrong, but wouldn't this reduce the horizontal resolution by a factor of 2 ?
I remember the artical on this. It was developed by a couple of holographic artists. The holographic lens that goes over the LCD will only add about $5 dollars to the cost. Every other column on the LCD are directed to the right eye and the others to the left eye by the holographic lens. The lens does not move. The head tracking system is so that when the person moves their head, the image can be recalulated/redrawn so that it looks like you are looking at the side etc. All you need is the software to divide up the image. Games are going to be great...
I saw an earlier incarnation of this device at Macworld Expo in either Jan 98 or Jan 97. The version I saw did not have the head-tracking gadget; it was just the LCD panel.
The booth personnel were sales flacks, not able to explain how it worked in technical terms. I got the impression that it was a similar principle to those plastic 3-d images like the "Lost World" movie poster. They superimposed two 640x480 LCD screens somehow and used a plastic lens to divert images to the appropriate eye. (Maybe it had something to do with polarisation?) The effect seemed to work fine as long as you looked at it dead on, but it got disorienting if you moved out of the sweet spot.
I seem to recall this technology first being developed in a movie-theatre-sized application by some Japanese researchers five or six years ago. Wish I could remember the specifics.
-Mars
I think Nintendo should grab hold of this technology and put it to good use in lets say their color gameboy?
Although the gameboy uses a reflective lcd and not a backlit one, details, details.
I would be interested what it looks like in real life.
I once saw a demo of a 3D no-glasses technique that gave the illusion of depth by causing the background to jitter. Imagine a shot at the beach where the close object is close and the background perspective is horizontally angled in a fluttered fashion. My mind recognized there was something to the depth, but I would imagine a full length film would be a headache. This was a demo I saw back in the 70's, so if it was patented, its free now! I'm not sure how to easily create the effect with two cameras and they must have used some heavy film tricks to get that effect.
i am still frustrated about 3d ... when you are blind in one eye everything just comes out blue ... i can't see those stupid pictures of squigly lines that suddenly become a picture ... and hand eye coordination should be hand eye*s* ... but i digress ... does anyone know if this 3d will work for me?
Hand-eyes coordination? Nice way to treat one handed people ;)
Honestly I doubt that most (if any) 3d displays will work for you, as they all tend to exploit stereoscopic vision (trying to fool sensory inputs into accepting a false image as a true one...)
Also, you're not missing much in the way of those magic eye pictures - I can't use one without crossing my eyes, and it's not like they're all that amazing in the first place.
-- This and all my posts are in the public domain. I am a lawyer. I am not your lawyer, and this is not legal advice.
I believe it shouldn't be terribly difficult to mount a camera on top of the 3D display, use pattern recognition to search the video signal for something which looks like a face and adjust the 3D image accordingly.
Comments, anyone?
Stop worrying about the risks of nuclear power and start worrying about the risks of not using nuclear power.
Back in the 1970's, Byte Magazine (RIP) had a feature article on a glassless 3d display, which used a rotating mirror synchronized to a vector display.
The object displayed appeard to float in front of the CRT surface, but, because of the mirror, would only be visible from the sides and top of the display.
The article had sufficient amount of (assembly language) code to enable any (serious) hobbyist to start his own...
-- ----------------------------------------------
Vive le logiciel... Libre!!!
Some French guy has made this 3-d display system... http://www.micronet.fr/~emuller/alio2.ht ml. According to them, is sounds like the greatest thing since flying toast, but I can't remember enough about optics to determine how real this is.
"I control the Mouse, the Mouse controls me."
Is the one in Disclosure, which featured the interesting shell command
do it/kill all
and a great scene in which the hero is watching a video as it is being deleted by Demi Moore, and the movie player displays a nice graphic of the video being deleted on the hero's screen as it is playing. More dramatic, I suppose, but awfully funny.
So incorporate the tracker into a set of headphones.
Work is for people who lack the imagination to play.
Cheesy to follow up my own reply, but incorporating the tracker into headphones would also assist 3D sound efforts, by allowing the user to experience 3D sound depending on the position of the user's head, not the position of the character's head, so a user could "look around" for a sound source while keeping the character's weapon trained on a particular location, etc.
Work is for people who lack the imagination to play.
If this company had enough money to produce such a product, why would they not have enough to set up a site that allows the use of credit cards? Or hire a decent designer to give a face lift to that site? It doesnt seem quite there to me.... I think I will wait for a little outside confirmation on this one.
The "L" shaped tracker on the monitor, does it not remind you of the tracker placed upon your TV in order to break out with the Power Glove in Mike Tyson's Punch Out? I don't think glasses free CD will be a big deal. Without total immersion, there will not be a large home market. I could see them being used for displays etc. but not much else. If I'm modeling something I want to be immersed in it, if I'm playing a game, likewise.
Dinyar
I remember reading about it in 'Electronic Design' about a year ago ...
LCD panel with a "Lenticular" screen over it. The screen filters out even and odd vertical pixels between the left and right eyes. The head tracker knows the position of your head so it knows how far to turn the screen "fins" so it separates the two.
Anyhow, in the future, there will be "eye trackers" which will allow the display to know what you are focusing in on. As such, it will allow you to "input" and read your eye "feedback." E.g., by focusing on a window in a GUI, it will come to the foreground!
Neat huh?
-- Bryan "TheBS" Smith
Independent Author, Consultant and Trainer
fine.
maybe it's just me.
the only price i could find on the entire site was for some of the software that runs on it.
has anyone got any idea how much this actually costs?
i've needed an extra monitor for a while,
and if that monitor could also do some of the 3d devel i need,
it would be great,
but i cant shell out but so much.
maybe one of the HMDs....
-Tannin Kal
It comes from the Mac's puzzle game. :)
Now if they can get Tetris in 3D, I'm there !!
Pope
It doesn't mean much now, it's built for the future.
The company was started by two people out of their own pockets (sound familiar?)
So I doubt they've got the $10,000 to pay a professional web site development firm to make a pretty web site.
I'm pretty sure I've read about this sort of thing several times over the past few years. It's just like those 3-d postcards or movie posters (among other things)...yes? I remember Cracker Jack prizes where the sticker would change depending on the angle you looked at it from. It doesn't strike me as too much of a stretch to stick it on a monitor. Also...couldn't you just as easily put the plastic "lens" on a CRT to get 3-d really cheaply? The only problem would be lining it up properly, which wouldn't be too hard with some calibration software.
It may not have quite reached the API level, but ATM customer recognition by iris pattern is a coming thing. If they can locate the eye in
the face well enough to take measurements of
patterns in the iris, figuring out what you're
looking at (and where from) should be a piece of cake.
So, where can I download libeyetrack.so ??
-- Alastair
Yeah, I remember that. Somebody had developed a special camera rig to film using that technique, and they showed a clip of it on TV. My eyes about bugged out of my head and my jaw hit the floor - it was damned convincing (imagine watching conventional 2-D, NTSC broadcast TV when suddenly the image goes 3-D on you...), but did have that background jiggle.
This was, oh, early 1980s. I often wondered what happened to it.
-- Alastair
Am I like the only one that remembers the bit on That's Incredible in the early 80's where *on a standard television broadcast* they ran a 3D segment using some setup invented by a couple of backyard hacker types? Basically it used two cameras and a switcher box to cut back and forth between every other scan (or something along those lines). It jumped around a bit (read looked wiggly) but damned if it wasn't 3D. They showed some sports footage and some frizbee tossing that looked straight out of Jaws 3D. Very cool and no special equipment required by the end user. Whatever happened to these folks?
Ok, so the holodeck of StarTrek may be a ways off into the future, but I find it interesting the the Star Wars films uses grainy blue holographic projections of images (people). These new projection technologies will bring the imaginations of Hollywood closer to our reality ....
... in our lifetimes
Forget webcam- we'll simply export DISPLAY=myhome:0
...some days you're the dog, some days you're the hydrant...
Actually, you could make the pixels half as wide, twice as bright, and use the optics to make them appear full size to each eye. QED
:-)
Well if you could somehow have some kind of channels in the screen that directed light towards a certain source and were able to manipulate their direction (chemically most likely), then by having 2 different sets of channels you could present a 3D image.
I saw them quite a few (4?) years ago at a electronics show in London. As others have pointed out, they focus two different images in slightly different directions.
The model I saw looked to be based on a phospor CRT, in fact, and it had a 'sweet spot' - you only got the effect properly if you were standing in the right place.
Jules
-- Any sufficiently advanced technology is indistinguishable from a perl script.
if it aint oss i aint interested, dont tell me about it!!!!
-- your knees hurt, don't they?
i am still frustrated about 3d ... when you are blind in one eye everything just comes out blue ... i can't see those stupid pictures of squigly lines that suddenly become a picture ... and hand eye coordination should be hand eye*s* ... but i digress ... does anyone know if this 3d will work for me?
purplestarr69@yahoo.com
if knowledge is power, the internet is god - me again
Hmm... a technology that would make this possible..
How about lenticular lithography? - Some recent advances
in this field provide 3D animated images.. considering how
LCD panels are made now, a marriage of the two creates some
interesting possibilities...
I have to say that web pages looks a little suspicious, if this technology works they should have something first class. /. For a look at how that works, have a look at this paper ( http://www.brunel.ac.uk/depts/mes/Research/Groups/ vvr/vrsig97/proceed/008/hasdpape.htm .
I work in a company that uses 3D display a lot ( http://www.reachin.se ) and we have just had a look at the display talked about in the article a month or two back in
This display was the best passive, i.e. no LCD glasses, display that I have ever seen. The lenticular lense displays that are around are poor.
The big problem with all these displays is that they always wind up halving the resolution. The active stereo system give you full res at your monitors display rate. The problem with these is that they need CRTs running at 120 Hz, which rules out LCD's as they have a maximum refresh rate of around 30Hz.
is, need I say it... Hackers? =D
"Short, tall, fat, skinny, from the highest king to the lowest man, everyone uses the potty." - Brak
http://www.cs.berkeley.edu/~healey/MURI/ It may depend on this screen or not.. I'm not sure. It is interesting though.
If this is what I think it is, it's entirely plausible.... it's simply a standard LCD display that uses a holographic lens to alternately focus a vertical line to each eye, and has a point source of light in the back. It is ingeniously simple, and will be very cheap too... now, all we need is to make an X server/Window manager that can allow us to use the third dimention for better organization. :)
...si hoc legere nimium eruditionis habes...
Unfortunately pattern recognition for something as variable as a face has not yet reached the API level. To do something like you suggest one would have to 'roll your own' with fuzzy logic, genetic algorithms for improving the weights, neural nets for feature extraction, etc. Doable, but not an evening programming project. It would be a lot easier to pick an IR diode or filter microphone input to get an ultrasonic signal. Once we have packages that encapsulate proven facial pattern recognition, sure. Just ask for the face in the field of view and compute it's position.
On older LCD panels, the contrast controls the optimum vertical viewing angle. How about this.... Put a chip like a quad-bilateral switch on the contrast control, with two pots, so you have two contrast controls that switch in and out on a flip flop. Now, hook the flip flop to something like the vertical sync line of the monitor output, or something slow and controllable like the parallel port. Program it to switch between contrast controls so that every other frame is controlled by one of the contrast controls. Now, TURN THE SCREEN 90 DEGREES and adjust the controls so that each eye is just inside the viewable angle for that eye... Hold your head very still... It's not perfect but it just might work.....
-rMortyH
___________________________________________
I have no use for hardware with a purpose.
I saw this technology on a Philips stand at the CeBit. Quite spectacular. Typically Philips: they really make very high tech stuff!!
c ial/3dlcd/proto/14inch/index.htm
http://www.research.philips.com/generalinfo/spe
enjoy!
Ewoud
ejjansen@xs4all.nl
http://www.xs4all.nl/~ejjansen
Definitely a RingMouse (had one, took it back). It can tell head distance and position (through a small transmitter worn on the head), but can't tell head orientation (i.e. tilt, direction). A better, but more expensive, solution would have been to use an infrared light source and two infrared CCD cameras to triangulate and get the exact position of each pupil. The eye's pupil reflects infrared light and could be easily tracked. This could automatically account for factors such as differing eye-to-eye distances as in an adult or child, head tilt, head direction, etc.
This is basically how I envisioned the design. The screen tracking the movements of your head/face seems unlikely... You would think that in order for the images to sync up properly it would have to focus more on where your eyes are. If it were some kind of head gear that you lined up say near your temples (the side of each eye) or something like that...
-=Redir
http://www.cl.cam.ac.uk/ Research/Rainbow/projects/asd.html
Research project in the Computer Lab and Dept. of Engineering.
Hi,
I saw a prototype of a device like this at a stand at this year's Cebit. The stand was from some German university, I forgot which one unfortunately. The demonstration display was limited to 640x480 pixels and it was also about 12 inches in size. Two eyetrackers at the top of the display note the position of your eyes. The display's columns have two sides that are at an angle, these are rotated so that both eyes see a different side of the column.
I was underwhelmed by the effect at first when the demonstrator was just showing rotating cubes and other geometric shapes but later he put on a simple 3d landscape built out of some iso-lines (not solid) and there the effect of zooming in and out and circling around was quite impressive.
One disadvantage is of course that only one person can use it at any one time.
Michiel Denie!
There was an article posted here on Slashdot a month or two ago about doing it. Its probably using the same technology. (Ie, a holographic lens over an LCD screen to divert segments of the image to each eye) Alternately they might be doing it the old fashioned way, having a plastic lens over the screen like you'd see on those 3-D baseball cards.
I'd guess its the former. Using head tracking you could shift position of the holographic lens to keep the images meeting the proper eyes as the position of the head changes. Having experimented with those plastic lenses over monitors before, that's definately an issue -- you'll get the correct image at certain "sweet spots", but they alternate with areas where the image gets reversed. Looking at it in one of those spots gives me a headache.
I believe this was on slashdot alittle while ago... The technique involves using two lcd screens (one on tope of the other) and alternating scan lines on each (i believe). originally you needed to keep your head in the correct possition to make the effect work... it seems that they've just added a head tracker to allow users to move their head.
This thread has it right. The technology is new, and newly patented, and dirt cheap. It uses fixed hologriphic lenses on normal LCD screens, and is supposed to look VERY good.
There was an article about this in EETimes about 2 months ago (I saw it in the paper version, so unless I can address my recycle bin there is no URL available)
The only limitation now is that it halves your resolution. To get an 800 x 600 display, you would need an actual LCD resolution of 1600 x 600. (but that is an easy problem to solve). This will make for some very good games, and will meet the price point.
The original developers (the holographic artists mentioned in this thread) are poor now, but I believe are currently negiotation rights to their patent for millions of dollars. They probably do not have credit card processing information on their site because they are not particularly interested in selling to jane consumer, they are negiotiating with the big guns for the big money. They can knock out prototypes pretty cheaply now, but they are not interested in large scale production.
All this is from memory of an article I read over lunch two months ago, so take it all with a grain of salt.
Bill
Mathematically impossible requirements are technically not against policy.
Here is the old story from Slashdot. This article happens to be on a completely different product. As some other people already mentioned there is no information about how it works exactly. The previous article actually had some stuff about how it worked.
Well, if anybody finds that useful...
License managers are the bane of my existence
This may just be a shot in the dark, but here's how I imagine that it works:
... ...
You build the screen so there are two pictures interleaved pixel by pixel, and physically, only one of the two pixels is visible to each eye. That way your eyes see different pictures at what appears to be the same place, thus giving the illusion of 3D.
The grid might look like this:
R00 L00 R01 L01
R10 L10 R11 L11
.
.
.
Where R and L refer to right and left eye images. Then you have to construct the screen such that from the angle of the right eye, only the R pixels are viewable, and from the angle of the left eye, only the L pixels can be seen (simple optical filters can do this).
The screen has to track your head because you have to face the screen directly for it to work.
Any other ideas?
The EE Times article is here.
At Live'96 (!), a consumer electronics show in the UK, I saw something similar from Sharp, IIRC.
The 3D effect was only viewable from a few defined angles, and to check you were in the right position there was a green LED on top of the screen which could only be seen when the angle was right.
I *think* that a polarising filter was used to display a different image for each eye. It worked, but of course if you moved your head the 3D effect disappeared...
If this uses tracking, perhaps it changes the polarisation angles dynamically, or is that just wibble?
It was a few months ago, may have even been here (on slashdot). The monitor has little polarized 'fins' much like modern trafic lights or 'walk/dont walk' signs; each eye can only see the image from the 'fins' polarized for it so each eye sees a different image and you've got 3d. It also seemed the process for creating these was afordable and the technology works with LCD screens.
Slashdot, would a spell-checker for posting be too much to ask? It's not rocket science!
The tracking device looks like a slightly modified RingMouse (now called the Owl by some resellers, I think.) It uses ultrasound time of flight, emitted from the tracker and received by three speakers, to track 3D position, and an infrared receiver for button clicks.
I sincerely doubt that the tracking is being used to adjust the lens (be it holographic or lenticular, I don't know.) Even with a perfect lens, you want to track the position and orientation of the head so that you can adjust the 3D rendering view accordingly. Without the headtracking, the image of the object will distort as you move away from the sweet spot, i.e., the point in real space that corresponds to the virtual view point. With the head tracking, the object appears to remain still, and you can move back and forth to get parallax, look at the sides, etc. With a full 6DOF tracker, you can even tilt your head and the view remains perfect. With the tracker used here, you will not be able to do that.