360-Degree 3D Imaging

CompSurfer writes "Yesterday, 3Dsolar released information about a new 3D imaging system, it works by somehow projecting onto a rotating screen. According to the story the "Prototype debuts at NVIDIA Oct. 25 in Santa Clara, California and [at] Innovact in Reims, France, October 5-7" I wonder where they are hiding the force field emitter..."

Comments

Re:In depth technical analysis

[jepaton]

Found their website:

3dsolar.com

But good luck in finding technical details there

Re:In depth technical analysis

[aklix]

All the images look photoshoped...

How would an image project on to a rotating screen work? I just can't see it happaning, it would have to be spinning pretty fast. My poor monitor...

I am sceptic

[uncl_bob]

I still remember those "super-cool" (according to the salesman) 3D-glasses with two tiny lcd-displays. After standing in line for like an hour I was allowed to try them on, and playing the coolest game there was at the time - hexen. And I still remember how worthless they were and how ripped-off I felt waiting that hour. This new thing is probably nothing like that piece of shit, but salesmen are still salesmen so I am sceptic.

Re:I am sceptic

[Anonymous Coward]

To feel immerseed in a 3D world using the shutter glasses you need a projection screen large enough so that the eyes field of vision doesn't see the edge of the screen, which acts as a "depth" clue to the brain that its not really 3D. Thats the reason why IMAX cinema works. If you ever look at the IMAX screen you can't see the screen boundary because its HUGE.

On a monitor you'll see an object popping out initially. But if your brain notices the edge of the screen, you'll suddenly see it less 3D because it sees the depth of the object being the same as the depth of the monitor screen.

One of the IMAX shows "Misadventure in 3D" eplains this better visually, very interesting if you get a to see it.

Won't This Need Tactile feedback?

[StateOfTheUnion]

"Imagine a 3Dsolar screen displaying selected store products 8 inches away from the actual retail store window and offering the passerby the possibility to interact with the virtual objects."

Without tactile feedback when interacting with virtual objects, is this really a practical thing that users would embrace? Or is "interact with virtual objects" just a euphemism for using the mouse and keyboard?

Old Hat

[FoboldFKY]

This isn't exactly new. It sounds like this system I saw a while back. I'm sure I'm oversimplifying it, but it basically has something akin to a thin pane of glass that spins around really fast.

What they do then is project a different image at each angle onto the glass. The images themselves are kinda like slices of a 3d volume (think volumetric textures). It spins just fast enough to fool the human eye into thinking it's seeing something hovering in midair.

I saw a few QuickTime VR demos of one of these a while back... they showed it with some wireframe terrain and a little purple jet thing in midair, as well as showing the sugar molecule.

Also, if I remember rightly, Nintendo had one of these puppies up a few E3s ago, with a model of that Star Fox character in it. Quite cool.

As for manipulating it by hand, I sure hope I'm wrong about the spinning glass bit...

Solar, sun .. Helios?

[joe user jr]

Noting the semantic compatibility between "solar" and "helios", could this be a development based on (or just a re-hash of) the heliosdisplay technology that's been mentioned here before? (to get the nyt article without a login, just google this and click on the link, at least that works for me.)

At least one blogger seems to be equating them.

Re:In depth technical analysis

[eMartin]

Well, it does work.

http://www.inition.co.uk/inition/product_stereovis _volumetric.php
http://pc.watch.impress.co.jp/docs/2004/0224/hitac hi.htm

But in neither of those case can you interact with the image with your hands, because either the image is enclose in a glass sphere or it would be pretty dangerous to try to "interact" with an screen spinning fast enough to create the illusion. And in both of those cases, it's clearly based on multiple images, which 3DSolar claims they don't need.

Re:From the 3Dsolar site...

[Anonymous Coward]

There is some info on the heliodisplay in an old article in the New Yok times:

Pressed for more detail on the nature of the conversion, Mr. Dyner referred to it electronic and as thermodynamic. After air is drawn into the machine, he said, it ''moves through a dozen metal plates and then comes out again.'' No moving parts are involved, he added.

He said the device works by creating a cloud of microscopic particles that make the air ''image-friendly.'' The machine, he asserted, uses no harmful gases or liquids, but he would not say whether it uses water. ''The ambient air is bottom-projected and illuminated, generating the free-space image that floats in midair,'' he said. At the demonstration, there was no odor in the air, and the area onto which the images were projected seemed dry to the touch.

With regard to 3dsolar, the article said

IO2 does not yet have a manufacturer for the Heliodisplay, but Mr. Dyner says he hopes production will begin in 2005

Here is another article with more info

Dyner bought a digital projector--the same kind used to display PowerPoint presentations--and took it apart. Inside was a micromirror system, a single chip that relies on a million tiny mirrors that tilt back and forth to create images. Dyner spent "seven days a week, 18 hours a day" trying to figure out "how to make the light stop in free space" using the micromirror system.

The key lay in using a fan to create a sheet of air that would reflect light projected at a given angle by the micromirror system. Dyner won't be too specific since his patents haven't yet been issued. But his first prototype made images from a computer hover in midair, something like a two-dimensional hologram. The nifty part: Sensors built into the box can tell when a user's hand (or an object used as a pointer) "touches" the image, allowing a finger to serve as a mouse.

This is somewhat confusing. If the air was onized, then I would expect the room to smell of ozone. Is ionized air reflective? Can someone explain the physics?

Re:In depth technical analysis

[Anonymous Coward]

Did anyone else notice that if you go to http://3dsolar.com (without the "www."), you get redirected to a bizzare news-reading avatar called "xinua":

http://www.xinua.com/xinua/xinuaus/index.htm

Apparently xinua and 3dsolar are all part of the same UK-based "AI" company:

http://www.audiotrack.org/xinua/audiotrack/index 2. htm

Strange mix of products & names to say the least!

Re:Smoke and mirrors?

[rpcxdr]

There are already existing technologies that create the illusion of 3D without fog, mirrors, or any moving parts. Bug Eye already holds a patent on a display that:

• Uses standard LCD screens
• Causes the illusion of 3D with no eye-ware, and
• Can be adapted for full surround 3D

The only way to do true holography is with (realtime) holograms. With everything else, you have to make some sacrifices.

Enhanced product image gives it away

[Animats]

The product image shown in the article is very dim. But if you bring it into Photoshop, do gamma correction, scale it up by about 150%, and filter the JPEG artifacts, it looks like this. Now you can see what it is.

That looks a lot like the DL-1 digital light projector, which is a video projector on a 2-axis tilt mount. "Using the motion control feature, project your imagery anywhere in a 3D space". It's used for nightclubs and stage shows.

It's a cute stage effect, but not a breakthrough.

A spinning spiral in an evacuated chamber?

[fyngyrz]

A spinning spiral (think Archimedes screw) places a targetable surface at every XYZ location in the entire volume it occupies during the time it takes to make one rotation.

So by modulating when you illuminate it (and what color, and how bright) you should be able to create a very nice representation of multiple objects; since the illumination only plays at the appropriate location in space, no glasses or other viewer-end tricks would need to be employed. It wouldn't look 3d, it would be 3d.

There would be size/material limitations, of course - the larger this thing is, the faster the outer edges have to go to get the same update speed. This would be a better approach for a personal viewing device than a group viewing device, I think.

Or, if you used a white surface, so that exact front-view objects could be created, put these spinners into "glasses" that would physically travel with the user, then sensing the viewers position would allow the display to change appropriately.

You could do some interesting things with a non-glasses version by making the screw something that varies in transparency as well - LCD or electrically controlled polarization. It'd have to be fast, though. LCD might not like the kinetic energy applied by the spinning, either.

It seems to me that all of the really hard complexity is in the illumination end, not the display end. Where do we get 3D information about the world? It's fine if you have synthetic scenes, for instance the output of a ray tracer, but to record 3D information about a real scene... ouch.

Re:In depth technical analysis

[NanoGator]

"How would an image project on to a rotating screen work? I just can't see it happaning, it would have to be spinning pretty fast. My poor monitor..."

What you do is mount a screen on a spinning doohickey. As the screen rotates, different images flash on the screen. Make the pixels flash at the right time, and the speck of light will appear to float in space. Use enough dots, and you can make an image.

I've actually seen this work at the 01 Siggraph show. It's not the greatest display ever (think vector graphics like in really old video games), but it's a start I suppose. Ever go to Sharper Image and see the clock that uses spinning LEDs? It's sort of like that.