Follow Slashdot blog updates by subscribing to our blog RSS feed


Forgot your password?

Projecting Data on a Sphere 139

necro81 writes "The NYTimes has an article in today's Science section that describes a four-projector system that displays images on a spherical screen. The Science on a Sphere system, developed at the Goddard Space Flight Center and used in some planetariums, can display and animate vast amounts of visual data from the Earth, Moon, Sun, and the other planets. The sphere is suspended by thin wires, and animating the image data gives the illusion of a free-floating, rotating world."
This discussion has been archived. No new comments can be posted.

Projecting Data on a Sphere

Comments Filter:
  • We are using this to plan our attack on the Death Star.
  • That's just what I need to show my pilots where the shield generator is located.
  • Could this be giving birth to the beginnings of 3D halographic techonology? Granted, they're only using spheres now, and they have to hang it from wires, but it's a start. Can they make a few cameras project a image of a person, so I can talk to them "face to face"? Maybe now I won't have to leave my house to go for an interview.
    • Wow, if that wasn't the weakest attempt to post early in hopes of getting modded up, I don't know what is.

      My cat sleeping on the keyboard has typed more insightful posts.
      • by Anonymous Coward
        My cat sleeping on the keyboard has typed more insightful posts.
        Is your cat halographic (sic)? If so, FRAG THE FELINE!
      • Wow, if that wasn't the weakest attempt to post early in hopes of getting modded up, I don't know what is.

        No, no, wait, I know! It's "imagine a beowulf cluster of these running Google Earth. SWEET."

        Or, "I wonder if this thing can run X Windows"
    • Re:Halograms (Score:4, Interesting)

      by w33t ( 978574 ) on Tuesday June 13, 2006 @03:39PM (#15526952) Homepage
      actually, by combining this technology with this technology [] you could have a pretty cool visual representation of the earth and it's current space traffic.
      • Pretty cool, now if they could just combine that technology with this [] and heat up the plasma a few thousand one would ever mod me down again for fear of a lightsaber enema.
      • Or, this technology [] if you want the starwars "Help Me Obi Wan Kenobi!"-kind of projector.

        Note to GP :
        it's hOlogram with an 'o'.
        it come from the greek holos : whole
        (and grama : drawing)
        because its the technology that is about representing a whole object (instead of only a flat drawing).
    • You misspelled "hagiographic."
    • by tepples ( 727027 )

      Could this be giving birth to the beginnings of 3D halographic techonology?

      No, that was the Xbox.

    • I don't think that's how you spell that word, Dave.

      This sort of thing has cropped up before, and it has always been atributable to human error.

      This comment can serve no useful purpose anymore. Goodbye.
  • Each projector covers a hemisphere.
    • by avalys ( 221114 ) on Tuesday June 13, 2006 @03:40PM (#15526962)
      Let's say you had two projectors, one pointing at the north pole and one at the south. The resolution would be terrible near the equator, because a very small part of the projected image would cover a large part of the globe surface.

      In order for the resolution to be consistent over the entire globe, you have to either intentionally reduce the resolution of the projection near the poles, or add additional projectors.
      • It depends if the projectors are inside or outside...
      • by Anonymous Coward
        .. and that's the best-case when you have the projectors positioned at infinity!

        If you have them any closer than infinity, then the equator will be in a shadow. Unless you cheat and have a slightly non-spherical globe.
      • Actually, I'm impressed they did it well with four. I expected six.
        • I'm guessing that by six projectors, you mean to put them along each of the axes (that is, at the points of an octahedron, or the faces of a cube). That certainly works, but if you put them at the corners of a tetrahedron, you can also achieve complete coverage, provided your projectors are at least 3 radii away from the center of the sphere. Using six merely allows you to reduce that margin to the square root of 3. (I think. Someone please check my math.)

          But, as it turns out, these folks aren't go

      • that would work if we could put the projectors at a distance of infinity. But digging a hole infinity feet deep or getting a pole infinity feet long to mount the projectors would be too expensive. Basically, if a projector positioned directly over north pole shining down, there exists a band in the northern hemisphere that the projector can't 'see' because of a point at a higher latitude that blocks it. put another way: Next time you are outside, look around and note how you can only see less than half
        • No, actually it's perfectly doable. This is different from seeing the horizon in that you eye is waaay smaller than the Earth.

          For this scenario all you have to do is have a light source and lens bigger than the sphere you are projecting onto. With an infinitely large lens you could also project on the *whole* sphere, from any given (fixed) distance, not just half of it. That's if you manage to focus it correctly, though.
          • That isn't quite the way that lenses work, however. You can't put a huge lens some distance away from the globe, and use the top part of the lens to project over the north pole, and the bottom half to project under the south pole. Instead, pretty much the entire lens is used to focus the image of the light source to any given point on the globe. You can illuminate more than half of the globe with a very large lens, but the parts at the periphery won't be as well illuminated, since only part of the lens

        • by Anonymous Coward
          In other words, the North Star, which is quite close to being directly overhead at the north pole, is not visible from about 10 degrees north latitude to the equator. Sailors know this very well.

          And Polaris is 430 LIGHT YEARS away, or more than umpty-jillion Earth diameters. If your sphere were the size of a beach ball, the projectors could be REALLY FUCKING FAR away, and there would still be a three-inch band of darkness around the middle.

          (I got too lazy to finish working out the math.)
      • You could do it with two projectors in the center provided they had ultra wide angle lenses like an Elumen's projector [] but those are somewhat expensive. One could easily spend $24k on precision wide angle lenses that would blend well at the edges.

        The smaller "planetoids" globe in the lower depths of the Sci-Fi mueseum in Seattle uses a single Elumen's lens/projector hidden in the wall projecting from the hidden back side of the globe.

        I have access to one of these projectors and I've long thought about t

    • You'd get very bad distortion on the boundary between the two. Three has the same problem; four is the minimum you need to avoid "singularities" in the projection.

      I'm actually a little surprised they didn't use more. I'd be interested to know how they keep the projectors in focus over the whole surface...

      • They actually use 4 projectors mounted around the equator. So, the "typical setup" still has a problem at the poles (like a 3-projector setup would). However, presumably most people will be standing around looking at the equator, so that might not be such a practical problem.
    • obviously you are not familiar with the 'limb effect'
  • by Anonymous Coward
    This is pointless - don't they realise the precipitous drop in attendance at Plane-Ariums around the world? I dont see how this is any different.

    ""Laser"" Floyd in your own house, though... woot!

  • by WinEveryGame ( 978424 ) * on Tuesday June 13, 2006 @03:34PM (#15526908) Homepage
    Isn't best projection on a sphere done from the center of the sphere?

    Since the application of this thing can easily require building a custom sphere, this seems a more cost-effective way to me.

    • by Anonymous Coward
      Since the application of this thing can easily require building a custom sphere, this seems a more cost-effective way to me.

      Assuming a requirement was the ability to hang the sphere to make it appear to float, then you'd have to figure out a way to have a fairly strong internal structure to support the projectors without creating seams from within the structure itself. Also you need to get power and video signals to the projectors and still keep your "disappearingly thin wires". Then there is always th
    • the wires would have to go in somewhere, which seems nonoptimal.
      • Gotta hang the thing from something anyway, run them in the top.

        I don't think it could easily be done without shadows from the internal structure though.
      • The projection need not be from the center of the globe. It could be from the bottom. The problem with that approach is that the image quality in the bottom half of the globe would be relatively poor. The way the system is designed now, most of the globe, maybe 75 percent, is better than the entire bottom half would be with a projection from the bottom of the sphere, and the mediocre parts of the current projection are much more limited than they would be, projected from the bottom.

    • Except you couldn't change the project bulbs.
    • by Chirs ( 87576 ) on Tuesday June 13, 2006 @03:47PM (#15527021)
      How would you support the translucent screen? Any structural members would cast shadows.
    • That makes sense. I bet they place the projectors in corners of horisontal squares. If inside, they can be point to the corners of tetrahedron.
    • by Anonymous Coward
      I don't pretend to understand anything about this, but the most logical way to project from inside a sphere would probably be a spherical projector. The problem with that being that there would be a distinct loss of resolution as the distance increased from the projector to the screen or area of the projection. If you project from the outside this loss is reduced as you don't have to deal with the reverse conelike nature of the projection from inside the sphere. In other words, the distance between single d
    • You mean something like this []. I've never seen it in person but friends who have say it's cool stuff. If you are a true geek, that's something you should have in your livingroom to entertain the kids!. Don't know the price though but I haven't seen them at the dollar store yet!
  • by 9gezegen ( 824655 ) on Tuesday June 13, 2006 @03:36PM (#15526921)
    There are not one but two such displays at Seattle Sci-Fi Museum [] One of them uses four projectors to project movie clips on a large sphere. I'm not sure about the second one since there are not projectors around (I suspect there is one inside) but it shows the surfaces of famous sci-fi planets. You can see this one at ds/index.asp []
    • Glad someone else mentioned on this... Lots of good stuff in that museam

      The first sphere you come across is definitely projected from the inside... there isn't an external projector in sight. If I had to guess they're using 3 for that one though...
  • This is also very useful for locating the vulnerabilities of and planning attacks on spherical battlestations. I even seem to recall a movie having a display system like this...

    P.S. Sorry, had to be said
  • The idea has come a long way since it came to Dr. MacDonald in 1995.

    "I was driving down a road and the thought came to me: Why don't people display things on spheres?" he said. "When I got home, I painted a beach ball white and projected pictures on the ball.

    So, I wonder if he took that beach ball to his presentations when he was trying to get research / grant money for the project?
  • OK, what do I need to build one besides a lot of money? I can get the money off my neighbor.
    • Re:I want one! (Score:1, Informative)

      by Anonymous Coward
      A sphere, a string, 4 projectors positioned: one below, three above (and to the side), the angles between any two projectors should be the same.

      Each projector projects a 2D image as per normal - the centre of the image will be an image that is non-overlapping with the other images projected. You could imagine the image that is projected as being generated by your bog standard 3D application, texturemapping a sphere and rendering it from the point of view of each of the projectors, then applying a mask to en
  • by JayDot ( 920899 ) on Tuesday June 13, 2006 @03:49PM (#15527039) Journal
    Birthday Trip to the planetarium...

    Who wants to open the pinata?
  • by CODiNE ( 27417 )
    Hasn't the Haunted House in Disneyland had this for ohhh... 50 years or so? In one room there is a crystal ball with a fortune tellers head projected on the inside somehow. As you ride all the way around it the face somehow matches every angle even as a dozen people are looking at different spots. I always figured it was some kind of multiple projectors, but how they got the overlapping and made it so smooth I never could guess.
    • Re:Ehhh? (Score:1, Informative)

      by Anonymous Coward
      Firstly, Disney fanboy nitpicking... the Haunted Mansion has been open since 1969, not 1955.

      Anyway... it's not the same effect. [] has an explanation of it is done. She doesn't follow you around the room, though. That's something else [].
    • In one room there is a crystal ball with a fortune tellers head projected on the inside somehow. As you ride all the way around it the face somehow matches every angle even as a dozen people are looking at different spots.

      I've never seen the thing, but you just described an effect normally created with a hologram. You don't see it often because there isn't normally much point. Presumably they thought it would look spooky because it's so uncommon.
    • In one room there is a crystal ball with a fortune tellers head projected on the inside somehow. As you ride all the way around it the face somehow matches every angle even as a dozen people are looking at different spots. I always figured it was some kind of multiple projectors, but how they got the overlapping and made it so smooth I never could guess.

      The illusion has to do with reversing of the depth of field. There's several good illusions like this, this Einstein one is truly awesome: http://www.grand- []
    • I think you're mixing up two different parts of the attraction. The woman's head in the globe is a simple bust. Her features are then projected onto the bust, allowing her to "speak." A similar technique is used for Buzz Lightyear in the Emperor Zorg attraction (I forget the name) in Toon Town.

      The following faces are done by employing concave faces (that is, caved inward, toward the wall) lit from below, rather than the more traditional convex faces lit from above. Because both concavity and lighti

  • by Anonymous Coward
    Don't Adam Sessler and Morgan Web on Xplay already have spheres on to which the project various things, such as the Xplay logo or Adam's face?
  • I guess it doesn't hurt to have contacts; A similar technology is mentioned in Douglas Coupland's recent JPod: A Novel. A great read for techies.

  • by llZENll ( 545605 ) on Tuesday June 13, 2006 @04:00PM (#15527124)
    "suspended by thin wires"

    Come on now, can't they use some kind of magnetic levitation system? Or even an air cushion, that would be pretty cheap and cool.
    • by MustardMan ( 52102 ) on Tuesday June 13, 2006 @04:04PM (#15527161)
      Either of those options is gonna be some sort of elastic force - the sphere will bounce. Even a slight bounce in a normal projector screen is quite noticable. I can only imagine it would be a lot worse when you're worrying about the alignment of three projectors.
    • well an air cushion would be noisy as you could hear the air moving.

      Now a mag lev system that could rotate and spin the ball would be cool. a Linear accelerator combined with brushless motor style controler surround the outside of the maglev system in the center with the ball just above it. A circlar linear accelerator(yes I said that and I mean it) with a center section for the actual levitation. come to think of it you would need 16 or 32 linear accelerators in a circular pattern. Computer controled an
  • by Anonymous Coward
    Great, just when I thought I had satisfied everyone's flatscreen envy I will now have to satisfy their sphere screen envy!
  • Google Earth (Score:2, Interesting)

    by Wormholio ( 729552 )
    wow, this would be a cool output device for Google Earth.
    • I have to disagree. The fun part of Google Earth is zooming in and "flying" over the surface, zooming in would just make whatever you're looking at cover the whole world. I suppose you could project the shadow of US pop culture over the rest of the globe.
    • That was also my first thought.
      I'm tempted to build a small scale one just to play with google earth.
  • Use this technique with a showing of "Snakes on a Plane".
    You'd get ... "Snakes on a Plane on a Sphere"
  • is an ellipsoid.
    is that too much to ask?
    • The variation is only a fraction of a percent, hardly perceptible to the unaided eye. Perhaps if you wait long enough, the force of gravity will compress the globe into the right shape for you. Besides, the Earth isn't quite an oblate spheroid, either (by ellipsoid is usually meant a prolate spheroid, I think); the southern hemisphere is ever so slightly bulgier than the northern hemisphere.

      Furthermore, this thing is supposed to be able to maps of other worlds, such as Mercury, whose oblateness is pr

  • I was just at Disney World and they had this cool globe that could show weather, maps, moons, etc. I just visited their page today to read about it (no price shown). NOAA was one of the companies that used it. On NOAA's page (the link in the article), they show it as being patent granted. Here is the link to the Global Imagination Company where the gallery shows NOAA with one. Looks alot like the one they designed. The Global Imagination model only requires one computer and screen. "..said Alexander E. Mac
  • ...if you need to display something on the surface of a cube then you're screwed. But have no fear, exp(pi*sqrt(163)) is here. Later today I'll file my patent application for a cubical screen and in a few years time you'll be able to purchase one of my cubical screen projectors for a reasonable price.
  • Brilliant (Score:3, Funny)

    by GimmeFive ( 978590 ) on Tuesday June 13, 2006 @04:56PM (#15527621)
    There are some arses here at the office that could now be used for IMAX.
  • It's for advertising...on the moon!
  • Sounds like it would be cool. The best application I can think of is a really fun shaddow puppet stage.
  • So, basically, you can display ANYTHING you want on it, as long as that thing is perfectly spherical. And you only want to show the surface.

    Hope that they got a patent on that. Man, the uses... This will be SO in demand.
  • This science is used by Goddard but was developed by NOAA, and its still under development by NOAA. NOAA owns the patent. More information here []
  • by Retired Replicant ( 668463 ) on Tuesday June 13, 2006 @05:33PM (#15527856)
    The cool think about something like Google Earth is how you can zoom way in to see tiny details. If the display is an actual sphere, wouldn't you be kind of limited in showing anything at a scale other than that which can be represented on a physical sphere the size of the projector-screen?
  • The hardest part was grinding a lens into the shape of pi
  • The Tech recently launched a Science on a Sphere [] exhibit on their lower level too. It's pretty cool stuff.
  • by zefram cochrane ( 761180 ) on Tuesday June 13, 2006 @06:31PM (#15528226)

    All jokes about the Deathstar attack aside, I actually had the privilege of seeing this display firsthand this past year. I was attending the Annual Meeting of the American Meteorological Society in Atlanta, and NOAA was showing this display with information about the 2005 hurricane season. It is quite a large display, but it has the capability of showing large amounts of data in an entertaining and easy to understand way.

    Here are a few pictures of the actual display in operation... E/ []

  • I'd like to know if anyone has any ideas as to how to do the following:

    I want to create planetary textures for Celestia
    (see [])
    but creating a rectangular image for the texture
    and then having this projected onto a sphere
    makes the process tedious trial and error --
    the poles are really hard to get right.

    I'd like to use some sort of modelling software
    to 'draw' the texture on a spherical body
    and then unwrap this into a rectangular image file
    suitable for importing into Celestia as a
  • The approach will not be easy. You are required to maneuver straight down this trench and skim the surface to this point. The target area is only two meters wide. It's a small thermal exhaust port, right below the main port. The shaft leads directly to the reactor system. A precise hit will start a chain reaction which should destroy the station.

  • "Honey, does this projection make me look fat?"

APL hackers do it in the quad.