Slashdot Mirror
Perspecta Walk Around 3D Display
Spinneyhead writes "New Scientist reports on the Perspecta display, a goldfish bowl like device that displays moving images in such a way that they seem to "float" within the display.
"To display the image, software inside the Perspecta chops a 3D model generated by the computer into 198 separate pieces, like slices of cake, which are then projected onto the screen in quick succession by a graphics accelerator that feeds image slices to an optical system mounted below the screen. The result looks to the viewer like a 3D image composed of 100 million "volume pixels" or "voxels".""
Doesn't look so impressive on my screen!
so now I'll finally get to see that Leia message?
You'll find the company here:
Actuality Systems.
user@host$ diff
NetInfo connection failed for server 127.0.0.1/local
With all these 3D displays coming out it sure would be nice to actually see some pictures of one. No that little CG pic doesn't count.
for the same price as one 25 cm 3d display, you can have 4 powermac g5's with dual 30" displays. that's 8 massive displays for the price of one 25 cm goldfish bowl..
:)
In order to make it appealing they'd have to produce it for about 400 $ methinks. and connect it to a telephone
B.
Every experiment which ends in a big bang is a good experiment.
The lady who was operating it said I'd meet a tall handsome stranger. Luckily she didn't ask me to cross her palm with $40k (that's a heck of a lot of silver).
I read the atricle, it sounds realy cool. But isn't 12cm kind of small for use in air trafic controll. "sorry, i can't see your blip because it's behind another blip" I think screen flicker isn't as bad of a thing as size.
We are the Borg...
True confidence comes not from realising you are as good as your peers, but that your peers are as bad as you are.
I definitly don't want it until they get up 85 revolutions per second, and probably more. And I thought that 60 was horrible, imagine what 15 Hz would look like.
If I have nothing to hide, don't search me
Just don't forget and reach for the pretty picture. . .
Seeing bad movies only encourages them. Watch responsibly
I still can't figure out how this gives the appearance of 3D. Can someone explain this for the physics-challenged?
More can be found about the technique, as described by "To display the image, software inside the Perspecta chops a 3D model generated by the computer into 198 separate pieces, like slices of cake, which are then projected onto the screen" on Wikipedia
Now I can see the back of my email messages!
But I will have to mortgage my house to do so.
This is hardly new or innovative. I'm pretty sure the company itself has been mentioned before as well.
Projecting images onto a rotating plane or helix is old stuff.
http://www.actuality-systems.com/index.php/actual
How does it work? A spinning screen, must be transparent I guess... what's the sci-fi sphere for?
I don't want to sell you death sticks.
The product in TFA is only 25cm in diameter, and if the featured image of the display is up to scale, then its display dome is going to be about 20cm in height. I can't imagine the display of this thing can be too good from underneath with the bottom of this device in the way. It is also stated in TFA that, in order for this thing to work, the screen on the inside has to spin at 15 revs/sec.
You can imagine the complexity of this device as it grows in scale. Even having a version of it at double the width and height is going to cause issues in trying to control the stability of all of its components. The velocity of the outermost edge of the screen (closest to the enclosing dome) is going to increase significantly as the device increases in size. All the while, you're creating a tornado within the dome.
I'm afraid that large scale versions of this device are going to be infeasible in functionality, production, and especially cost (this baby version costs 40 grand) as many complications are going to arise.
Falun Dafa is good!
No more MS windows everyone will be using the new MS Bubbles OS. Unless Micheal Jackson has the copyright for that!!
At least for practical puposes. 100,000,000^(1/3) = about 465 pixels on an edge. That's about on-par with 640x480 on a regular monitor.
I'll stick with my "crappy" 2-dimensional monitor for the time being.
Pulp Audio Weekly - Geek News and Reviews
I dont know why it sounds like a jet taking off but heres a movie
"It's so convenient to have a system where everyone is a criminal" - A. Hitler
...as most of my programs tend to be floaters.
I recall reading about such a system in the late 80s or early 90s. It was made by TI and was much more ambitious - think a 2m x 2m x 1m tank used for air traffic control.
:-)
This ones looks more practical, even if much less useful. At 15Hz and a mere 200x768x768 pixels, it is requires a mere 1/3GB but a whopping bandwidth of 5GB/s, and the quality is like that of a Dr. Who prop. Scale it up to 512x1048x1048 at 60Hz and you'll need an acceptable 1.5GB of memory but unrealistic 90GB/s memory bandwidth to drive the thing.
While this might be possible to resolve using massively parallel interfaces or something, I bet we'd still need Moore's law to hold for another decade or two before the quality of this type of display can rival that of current 2D ones.
In the meanwhile, this will remain a gimmick or be limited for very special applications where the low quality is acceptable (hint: this probably rules out medical applications
This type of display could revolutionise the way we do 3d modelling. With 3d controllers you could model as you would with clay.
A spinning display will finally justify our IT depts aversion to neck-ties. Now if we can just find a good excuse for hygeine...
So I have to walk around the display to see the image in 3D? This is progress?
--"It's Bradford Company, slash your last name, dot your first name"
bring new life to all those classic pornos!
In 1998, when I saw it on "C|Net".
-- 'The' Lord and Master Bitman On High, Master Of All
Kinda reminds me of the early televisions, with the huge spinning color wheel.
But when you become entranced by the image floating in space in front of you, and then slowly reach your hand forward to touch it, you'll be glad that you're standing around zoning out on a million dollar 3D flipbook with a half dozen neurosurgeons.
Quantum materiae materietur marmota monax si marmota monax materiam possit materiari?
But can't afford it yet, so... *Everyone*, show this technology to your CEOs and CIOs. Use some visuallisation bullshit to get them to buy some of them.
Then when the price comes down to a reasonable level we'll all be able to get one.
Deleted
If i remember right, it was about a year or so ago that this company started getting press for their 'upcoming' product.
Was mostly just a spinning disk inside a globe to 'simulate' 3D.
---- Booth was a patriot ----
I remember seeing descriptions of rotating projection plane 3d systems more than a decade ago in magazines.
If someone want to "wow" me, set up a system based on multiple scanning lasers in a transparent medium where two or more beams intersecting cause the medium either to glow or to become opaque depending on the combination of beams intersecting and the non-linear optical properties of the medium.
*That* would be cool.
will be swimming fish.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
Based on the actual product's website, I think this is a way cool device with some Pros and some Cons.
Pros:
Looks really neat. Seems right out of sci fi.
Could really be useful, especially with med imaging.
Cons:
Too small
Too expensive
Glass sphere seems too glossy and reflective.
apparently only 'glowy', translucent, laser-light-like images can be produced.
Big spinning piece inside: wear and tear? noise factor?
Tetris just got a lot harder.
The Boston Chapter of SIGGRAPH http://boston.siggraph.org/ had a 'factory visit' last year. Very cool tech, nice very smart folks. The main problem is display bus bandwidth, if you start cubing the required data over any existing wire the technology just does not exist yet.
We can project a 3 dimensional object onto a two dimensional plane (a standard monitor), and get a good idea of what the object would look like if it was in 3D. I've seen animations of 4D objects on a 2D display, and I just don't grok the nature of the objects. If someone writes a visualization that projects a 4D object onto this 3D display, will it be much easier for the average person to grasp and understand?
LS
There is a fine line between being a cultivated citizen and being someone else's crop. - A. J. Patrick Liszkie
Here's an idea. Instead of providing a surface to scatter off by rotating a surface in space, fill a vessel full of some gas and focus two lasers at the point you want to scatter the light. Arrange for the freqencies of the two beams to add up to the frequency of a transition from the ground state of the atoms in the gas to an excited state. Photons should be produced where the beams intersect. Then you could make an image by just scanning through the volume intersecting the beams in a grid. Conceivably color could be provided with a mixture of gases and various lasers.
The world is everything that is the case
Basically, this is just a spinning projecter & screen. To make it bigger, just build it with a bigger screen and brighter projecter. The rest of the hardware and software would remain the same.
As a side note, the flicker probably comes from viewing the back of the screen when it's turned away from the viewer. They need to add a second screen & projector to the back of the first.
So, any guesses on how long before we see "porn bowls" 6 feet high projecting full sized 3D porn?
"That's so plausible, I can't believe it!" - Leela
Put the video processing in the device itself and develop a 3d language (opengl?) to allow the main computer to describe what's to be displayed.
Plus you only have to send what is to be displayed, empty space doesn't need to be sent or held in memory.
Deleted
Cool pictures!
...something like that some 10 years ago...
If you want practical 3D displays for something like an airport, you CAN use stereo goggles along with a trackball to rotate around the y axis. A setup like that would work well sitting at a desk, rather then this setup which would require everone standing around the center of the room, getting in each other's way. (Or each person with their own table to walk around)
Stereo goggles only require 2 3D graphics cards, which can easily be run by a single computer, and they end up giving you more freedom, since you can rotate an object quicker with a knob (or trackball) then you could walking around the display.
So why haven't they done this yet? Price sure isn't an issue. Though the price of reprogramming their systems might.
It may be because they want them to be able to see other stats other then just the view of the planes. But if they give multiple views on a single "screen", then use a device like the Nostromo SpeedPad to switch between them, then they'd have just as much control as they have now. Left hand on the screen controls, right hand on the rotation knob.
3D technology has been around for a long time. The problem isn't with the hardware, it's with the software. Hopefully, that's why the PS3 has two screen outputs: for 3D graphics in games. (If the PS3 has built in 3D goggle support, I'm getting a 3rd mortgage)
"That's so plausible, I can't believe it!" - Leela
Anyone else remember this game from the early 90's? It featured a time traveling cowboy that appeared to be 3d because the video was projected on a parabolic mirror with a dark filter.
Does this mean I can sell my fishy and bowl for $40k?
My BETTER suggestion/solution: a screen than can be gimbaled 360 degrees in the horizontal and 90 degrees in the vertical. Project 2 polarized HIGH-resolution images on it for high-resolution stereo imaging (only a cheep pair of polarized glasses required for viewing). Now track the head of the viewer with an infrared beacon worn on the user's head and automatically swivel the display to be perpendicular. Display the appropriate image pair at whatever frame rate you wish.
I have seen near VR systems that track your head with a pole to accomplish the same effect, but only at a fixed distance and not a full 360. Head tracking in my suggestion could be accomplished at a variety of qualities.
1. 360 tracking only (no need to gimbal in the vertical).
2. Gimbal in both horizontal and vertical, but assume a constant distance (just track the head's direction from infrared beacon). Note: if there ins't a need to actually peer directly down on the object then no vertical gimbal would be needed. Vertical adjustments could be made from head tracking and image adjust for a reasonable range of angles, certainly upto 45 degrees.
3. Detect distance to head also with some kind of ranging based on the infrared beacon, adjusting the stereo pairs to maintain proper parallax.
4. Track head tilt with a tilt detector in the infrared beacon's housing, modulate this information into the beam so the image pair can also take head tilt into account in maintaining proper parallax for stereo viewing.
5. If you want to get really exotic you could track eye accommodation (focus), you would then have lenses counter adjust focus to keep optimum focus on the screen's actual location, then have the computer blur those parts of the 3D image not at the focus distance detected.
Two doctors consulting on one 3D image set -- two viewers. This method would be orders of magnitude cheaper for any given resolution. P.S. if anybody wants to run with the ball on my suggestions I wouldn't mind a kickback. Though in all likely hood all my suggestions are probably already being worked on in the lab.
Letter To Iran
This technology is nothin new.
It was already used during WW2 - but instead of projector, there was single light beam to the mirror, so it was capable to display only simplistic vector graphic. But the idea was the same.
And with this technology you surely can NOT see any image when you are looking from top at this spere (as stated in the article). You will see only upper side of revolving mirror.
What scientist wants to walk half way around just to be able to see the back of something? I think in practice they will just use a mouse to rotate the thing so they don't have to get out of their chair, in which case why not use a regular flat 3d or even 2d display? Also for meetings or times when there are multiple people around a display, they will be all be seeing it from different viewpoints, not the best thing to have happen, wouldn't it be more effective to also use a regular flat 3d display for this also? I mean this is cool technology and all, and it would be cool to have a 3d fishbowl but it just seems like a waste of money. Also keep in mind this is 15 fps at low res, i hope they have some eye drops on hand.
OK, it's kind of wacky, but here's the scheme. The screen is just a flat piece of Scotchlite. This is a remarkable material that reflects almost all light directly back to where it came from. Obviously, this is just part of the scheme.
The other part is the wacky goggles. These have projectors mounted above them with tiny LCD or OLED screens that project down through half-silvered 45 degree mirrors in front of each eye.
So, the light from the projector is as if it's coming straight out of your pupil, and so reflects back right at each pupil. This way, each eye gets only its own image.
To make this work well, you'd have to get some kind of head tracker as well, to move the image as you move your head.
The nice thing about this is that you could have multiple people viewing the same screen, and so sharing the experience.
The devestating problem with the perspecta display is that there cannot be hidden surfaces, not a problem for this (or any other) goggle based system.
The bandwidth problem of the perspecta display comes from needing to display an image from all points of view -- even though 99% of those points of view don't actually have anybody viewing them. Goggle based systems don't have that problem.
But! If you wanted to have a shared goggle-free environment and had a large amount of money, you could do that too! You'd want to have a very large half-silvered mirror with many projectors above it -- enough so that from any point in the field of view you could have a projector more or less lined up with each viewer.
Anyway -- back to reality.
Thad
I love Mondays. On a Monday, anything is possible.
Also, where would a company put such a thing? It certainly wouldn't do any good sitting on my desk - you'd have to have a dedicated 'analysis' room with this thing sitting on a little desk all by it's lonesome.
I do like programming things that work super quickly, especially when they work super quickly, super quickly.
I don't understand all of these negative posts. This equipment is obviously still in the very early stages of development and impractical for most people.
This is an extremely common pattern in technology. A new product comes out and is only useful in a few niche markets, often analysing the very important (medicine), very small (molecules), or the very large (climate). Eventually the product matures and becomes useful to the larger society.
I am delighted that people are working on 3D displays; I certainly don't want to working on a freakin plasma or LCD screen in 30 years.
I imagine how many of these posters would have reacted to inventions in the past. "WOOOW, a device that adds numbers? Why would anyone spend $20000 on this 'computer' when you can just add the numbers in your head? These machines are lame."
Finally! My goldfish screensaver has gotten that much useful.
so really, this is a $40,000 crystal ball
j/k
it does seem really cool. I just can't think of a practical use for it for myself, though I'm sure that one day they, or technology inspired from them, will be very important in some sectors.
We have this in out school , Oregon state University , its way cool ..although the resolution is quite less ..and you can directly plug in to your 3D output. Things like wireframes come up nicely ... but textures etc suck.
The result looks to the viewer like a 3D image composed of 100 million "volume pixels" or "voxels".
What is the dead pixel policy on that one?
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
So it must be perfect for playing Delta Force in 3D!
Moderation Totals: Flamebait=2, Troll=1, Redundant=1, Insightful=6, Overrated=1, Underrated=1, Total=12. (not mine)
Mmmmmmmmmmmm spinning lollipops in transparent polycarbonate shells..
*drool*
https://www.eff.org/https-everywhere
You can bet the military have such things on their long-term shopping list.
Do not mock my vision of impractical footwear
We've had a Perspecta for a year or two now. The biggest problem we had was in getting the damn thing perfectly level. Think of what happens when the tub in a top-load clothes washer gets out of balance, now imagine your $40k display doing the same sort of banging shuffle across the floor.
Once it was level, all the criticisms leveled above apply. Not enough resolution, not enough colors, not enough refresh. The room has to be dark to prevent glare from the bubble and so that you don't see the screen spinning. Stare at it long enough and get retro VirtualBoy-style headaches. Makes for a fun demo, but I can't imagine a doctor or air traffic controller seriously using one.
I don't understand the "feature" of being able to walk around the display. Why do I have to get out of my chair? Why can't the computer just show me the back side? It would be great if you could walk inside the image, even interact with it somehow, but you can't with a Perspecta.