What is Digg?77 Comments
- digid, on 10/12/2007, -1/+31There's also a video of the iTunes visualizer projected onto his concoction.
http://www.youtube.com/watch?v=-T4wnHfHiqw - Ark7, on 10/12/2007, -5/+28This video goes on forever with absolutely nothing happening. Didn't need to be 9+ minutes.
- Artifez, on 10/12/2007, -4/+26It's called proof of concept, bud, not entertain the yokels.
- hohead, on 10/12/2007, -0/+21"Uhh.. whaa.. as soon as I interact with them, they sort of distort because.. umm...
because.. uhh.. because as soon as a human body sort of converges with digital space... uhh... they can't co-exist really... so..."
LOL
Gold. Worth it to watch for that explanation alone.
Picture this guy on a date! - pseudojd, on 10/12/2007, -3/+22This is Genius. movies in 3d. Porn in 3d. Awesome.
- spyrochaete, on 10/12/2007, -0/+17+digg for the awesome story description
-digg for video of a guy in a wifebeater
I'm conflicted, but my nerdiness takes priority. - BlackCow, on 10/12/2007, -2/+18"distortion because of a human interfering with "digital space"...they can't coexsist"
lol wtf? - SuckMyDigg, on 10/12/2007, -2/+14He's got more nuts there than a roswell convention
- molsen311, on 10/12/2007, -0/+12porn in 3d? that's called human to human sexual contact. maybe you'll get to experience that one day.
- Evoguy, on 10/12/2007, -2/+13OpenGL and Zbuffers would have absolutely no use here. Please stop throwing around basic 3D graphics terms in an effort to sound intelligent. Zbuffer information is built up from polygon rasterization, what could that possibly help with here? The data isn't polygon based, the depth information is implicit to the surface he's projecting on.
- Artifez, on 10/12/2007, -2/+11Excellent idea, it would take a serious rig to do 3D video but the basic nuts and bolts are there. Great job.
- sinembarg0, on 10/12/2007, -0/+8Dugg for:
On the bottom there's 256 nuts, all hanging just above the floor.
Wow I'm immature. - usbserial, on 10/12/2007, -1/+8Imagine this but on a larger scale with the strings taking up a 5x5 x 7 high space, and with a few thousand strings. I can really see how this could make a really stunning image. Brilliant concept.
- schroduggity, on 10/12/2007, -1/+7it's hard to imagine that working better than current stereo 3d displays like:
http://www.edimensional.com/
which work with any opengl game. it's awesome I have a pair and if your computer is fast enough and your monitor has a high refresh rate. it really improves the gameplay. for instance in quake4 you can literally gauge the vectors of rockets being fired by other players and it seriously improves your aim.
really neat trick for a cardboard box and some strings tho. he should wrote it in c instead of java to improve the speed. - Ramtech, on 10/12/2007, -1/+7Hmm... maybe make a giant wire map for a concert and have 3d images in background while the music is playing...that would be cool.
- brucebeh, on 10/12/2007, -2/+7wow, i think the itunes visualizer is pretty cool.
- samadam, on 10/12/2007, -0/+5@ Tenlow:
Did you watch anything past the two minute introduction in which he explains how it is set up? Yes, that part dwells on strings.
If he had skipped that part and just gotten to the 3d part, half the people on digg would have cried "fake!". - simpleid, on 10/12/2007, -0/+5I don't think this guy built this, just presented it, only because of that line.
- inactive, on 10/12/2007, -4/+83d movies would really just be annoying. You'd have to more around to get any benefits from it. The whole point of me watching TV or a movie is to sit on my lazy ass instead of being productive.
/Course I said the same thing about video games before I got the Wii. - hodrige, on 10/12/2007, -0/+4There is this show "Bones" on TV, and they have a 3D thing. I think it is strings and oil droplets that reflect a projector. Does anyone know what I am talking about.
- interiot, on 10/12/2007, -0/+4With only 256 strings, it's feasible to hand-record a rough gauge of how deep each string is. Though it's probably much more accurate to have the computer automatically do calibration beforehand, using a camera so it can tell (fairly accurately) how deep each line or point is.
I don't know if it's really possible to calibrate in real-time... maybe if all pixels were lit, and they were all lit with unique colors or modulation, but it would probably difficult, if you could even display a useful image with uniquely distinguishable colors. Given the alpha-quality of the project, I'm guessing he calibrated the position of each string at the beginning, either by hand or automatically via a camera. - felis, on 10/12/2007, -1/+5A little editing goes a long way, Mr. Video Man.
- interiot, on 10/12/2007, -1/+4The limiting factors are 1) how much resolution does your projector have, and 2) how precisely can you line up your strings with the projector's vertical lines (since projector resolution is a heckuva lot more expensive than string resolution).
There are 1920x1080 projectors out there, and they're pretty expensive, so that may be the practical limit, depending on budget.
One way to enhance 3D resolution though may be to make the opaque bits be points rather than lines. Use transparent string (fishing line perhaps), and put opaque bits on them (small white balls, liquid paper, ...). If it's acceptable to have small gaps in the horizontal dimension, then it's probably similarly acceptable to have small gaps in the vertical dimension, if it improves the image. - imants, on 10/12/2007, -8/+11@PatrickFisher
I seriously doubt that he is at all concerned with impressing you. - Ark7, on 10/12/2007, -0/+3No, not at all. All I wanted to see was a 3D image. Instead I got nearly ten minutes of self-congratulatory verbal masturbation.
- d3dm, on 10/12/2007, -0/+3Hahah, that's exactly what I was thinking when he makes the comment at 00:42:
"... and then underneath the whole thing is 256 nuts that are just hanging around."
Sounds like a good description for the construction crews on a local highway project here in Seattle. - lolwtfhaha, on 10/12/2007, -0/+3@hohead
I was going to paste the same quote. When he said that, I thought "wow, this guy is an artist." Why the hell can't he just say "When you move the strings, the light misses them so they disappear." WTF is with this digital interaction *****.. ahhhh made my head hurt. - Terc, on 10/12/2007, -1/+3"as soon as I interact with them [the strings] they kind of distort because... erm.......... as soon as a human body... sort of converges with digital space... uuuuhhh... they can't coexist. Really."
Was I the only one that got the impression that he was NOT the person that created this? It's ingenious, but if he doesn't understand that because the wires were moved from the row that the projector was projecting at, they were no longer illuminated, there's no way he came up with the idea or wrote software to do this. - flamingmb, on 10/12/2007, -1/+3yeah skip to 3:00 and you wont miss anything.
- mbrindam, on 10/12/2007, -1/+3dugg for J5
- Adenosine, on 10/12/2007, -0/+2@schroduggity - one big advantage I could see for this, if it could be vastly improved, is that is projects an object in to space: you can walk around it, multiple people can see it and it can be looked at from any angle. But this doesn't compete against stereo displays at all.
- kefler, on 10/12/2007, -0/+2What I dont fully understand is the haphazard layout of the strings.. it seems that they were randomly placed.. makes calibration much harder.. lets say your display was
AxB pixels.. and you wanted to have N levels of depth.. you should be able to take every N strings from left to right and arange them from front to back, for an actual resolution of (A/N)xBxN pixels (because of how he did it, it would have much higher vertical resolution than horizontal resolution. - goggwah, on 10/12/2007, -0/+2@Curufir
Your thinking of stereolithography.
http://en.wikipedia.org/wiki/Stereolithography - DelMonte, on 10/12/2007, -0/+2Nice concept, it would be nice as an art installation, but unfortunately, this couldn't lead to anything serious, even when increasing the number of strings.
His installation, with 256 strings, have an effective volumetric resolution of 16x16x768 (768 being the presumed vertical resolution of his projector).
You have to understand that with volumetric displays, the complexity and precision needed increases exponentially as you want to increase resolution.
With 1024 strings, which would probably max out the horizontal resolution of his projector, he still could only get 32x32x768. To get a 768x768x768 resolution, his projector would need to display an horizontal resolution of 589824 (!) and would require incredible precision in putting the strings, which would have to be very thin.
There's also a major problem, which is a fundamental problem to all volumetric displays (well all those with non moving parts), which is in that case that the strings are opaque and will hide other strings behind them. Random placement of string may help a little, but the problem is still there, increasing the resolution could potentially make this problem even worse.
The perfect material needed for a volumetric display would be able to switch from being either completely transparent, emit/diffuse light or absorb light, and all the states in-between.
Such material doesn't exist as far as I know (if it does, just tell me!). LCDs can switch between being transparent and absorbing light, but they don't emit or diffuse light by themselves. You cannot make a true volumetric display just by stacking LCDs and putting a backlight behind them all.
LEDs can switch from being somewhat almost transparent to a state where they emit light, but they cannot absorb light and switch from transparent to black for example. Turning off lights in the room doesn't get around this problem, as an unlit LED may look black in a dark room, but it'll be transparent and show the lit LEDs behind it.
And then you have things like opaque strings that can diffuse light projected on them using a projector. The problem is (as described earlier) that they are opaque, and also they require you to turn of all other lights to get black and dark colors.
Anyway, just don't get too excited about this concept, it has several basic limitations, and scaling it to get a useful resolution would be unpractical, because of the exponential factor. - fatbastid, on 10/12/2007, -3/+5he's like an asain macgyver. amazing. all you haters are just jealous.
- interiot, on 10/12/2007, -1/+3He didn't explain how he did the calibration, did he? I don't think calibration would be that difficult really. Any automated calibration scheme would probably require a camera for feedback, and probably the best place to put the camera is 90 degrees to the side, so it's trivial to calculate the depth of a specific pixel or vertical line. Then you just scan across all lines (or pixels), one by one, calculating how deep each is. Voila, it's calibrated.
- blahtastic, on 10/12/2007, -0/+1Someone help me here...What the hell does the projector do? The image is a globe or cube from all directions, its 3d, i got it..how does the projector/computer know where the strings are? Did he program the string locations? or does it sense them or something...i mean, it does know where they are, it colored them different colors...i'm confused..
- tcissell, on 10/12/2007, -0/+1I was thinking the same thing about Bones as I watched this (of course the Bones version is fake, but just maybe they are basing it on potential technology like CSI does).
This could be a good technique for medical imaging. I had a bone scan some years back (early 90s), and while I don't remember exactly how they did it, I do remember the result of an image of my skeleton that the tech was able to rotate using a trackball. I thought it was extremely cool at the time (and still do).
This method would have to be improved of course, but perhaps it could be useful. - Topher06, on 10/12/2007, -8/+9Skip it period, you will live a long happy life for knowing nothing of this.
- ebs16, on 10/12/2007, -0/+1This isn't an action film, it's a rough introduction to his testing rig. The video basically fits the form of a scientific paper (abstract, methods, etc...).
The demonstration would have no value without the explanation at the start. Wouldn't you be curious as to how the effect was created if all you saw were the last 3min of the video? - samadam, on 10/12/2007, -1/+2Yeah, that show uses a totally fake display, afaik.
- Curufir, on 10/12/2007, -0/+1@tcissell
Probably an MRI scanner.
The way that works is you take a whole bunch of separate images, each of which is a "slice" of the thing you're building an image of. Then all those slices get thrown at some funky software which uses them to build a 3d model of whatever they were imaging.
Things actually get more fun when they send that data to another machine (Can't remember the name) which builds a 3d model out of polymer (Not a lathe, this machine has lasers operating on some kind of liquid polymer). That gives surgeons something physical they can use to plan out their moves when resetting bones (The programme I saw it on had surgeons reconstructing someone's face). - Topher06, on 10/12/2007, -7/+8Not to burst anybody's bubble, but this is a projector shining light onto a bunch of strings. If in the future I have to pay $10,000 for a TV that is a bunch of strings and a $1000 projector, I will be pissed off.
- d3dm, on 10/12/2007, -1/+2Right. You'd render the 3D primitives with Z buffering enabled, then use the value in the Z buffer as a lookup to determine which projector line would be lit. Doing this would allow him to display anything rendered with OpenGL or Direct3D as long as he could get at the Z-buffer data on a per-frame basis. Locking a depth buffer isn't cheap though.
I'm not clear on what he's talking about when he's says that the amount of data that he's having to process is slowing things down unless he's written his own rasterizer. Accessing the depth buffer and allowing 3D hardware to perform the rasterization should eliminate that problem. - Andrej73, on 10/12/2007, -2/+3>This is Genius. movies in 3d. Porn in 3d. Awesome.
What about porn with REAL 3d girl?
;) - DelMonte, on 10/12/2007, -0/+1I thought that the random placement of strings was unnecessary at first, but after thinking about it, I understand why it was done like this.
The problem is that if you put the strings in a nice grid like you described, when you walk around the display, some angles will make unlit strings mask the others behind them, and since they are aligned, whole parts of the image will disappear in a rythmic fashion as you walk (and potentially the whole image). This could be described as "Moiré" artifacts.
Random distribution of strings enable some "dithering" of these artifacts, so that only a few strings will mask others at any given angle, instead of the whole thing flashing as you walk by. - lifewithout, on 10/12/2007, -0/+1Movies in 3d is a bad concept. Here's why: film making is based on 2 dimensional composition (even 3d gets rendered to something our brains understand). Good filmmakers frame every shot so that you the viewer see what is necessary at that moment in time. If you could see *everything* in the environment from any angle then no one thing or relationship between things is relevant.
- umber970, on 10/12/2007, -1/+2That's seriously sweet, I can't wait for the code...
- Evoguy, on 10/12/2007, -1/+2No.. once again a Zbuffer would fail to be of use here. And yes I *do* know how they work, because graphics is what I do for a living. You're making this problem more difficult than it needs to be. Zbuffers are a brute force way to draw multiple overlapping objects projected to a 2D viewport, however the depth information is much more easily obtained from the plane equations representing the polygons themselves.
Zbuffers aid in *sorting* not *evaluating* depth information. In this case you already know the depth of your strings, and you already know the depth of each pixel from evaluating the plane equation in x,y,z space. You're not trying to depth sort a hundred polygons for 2D projected display in this application, you're simply trying to show a simple model (like a rotating cube) in 3D space. Actually rasterizing in a GPU and then looking up the Z value would be pointless, and just going in a circle of work. The very nature of displaying on a 3D surface is moving the depth-sorting operation to your eye and brain, that's the whole point. - zediker, on 10/12/2007, -1/+2@Evoguy
And this would be bad how? It still cant draw what is behind another pixel since the light is being projected from one direction. Its a 2D immage being placed into a pseudo 3d enviroment. It still cannot display occluded polys. The point is, utilizing the z-buffer would save precious computation time for drawing more complex scenes than intrepreting the poly coords via costly branches in a software based process. -
Show 51 - 77 of 77 discussions
Browsing Digg on your phone just got easier with our enhancements to the 
© Digg Inc. 2009
— Content created and posted by Digg users is