Slashdot Mirror
A Kinect Princess Leia Hologram In Realtime
mikejuk writes with this snippet from I, Programmer: "True 3D realtime holography is not only possible — it makes use of a Kinect as its input device. A team at MIT has recreated the famous 3D Princess Leia scene from the original Star Wars — but as a live video feed! It's a great stunt but don't miss the importance — this is realtime 3D holography and that means you can view it without any glasses or other gadgets and you can move around and see behind objects in the scene. This is more than the flat 3D you get in movies."
Nearly there, it is.
"Real holography" my ass. Unless I'm misinterpreting the video, what they're producing is a ~15 FPS red blob, with no 3D except what's captured by the Kinect. You're still going to see a flat image on the screen (and those on the left and right of the theater will get the same image).
That'll save Hollywood!
The first 2D electronic television displays had similar levels of performance. This is a tremendous achievement. If you want proper 3D - sans glasses - this is almost certainly how it will happen.
I know people will hate me for saying this, but in a way, it's better if everyone sees practically the same movie. If we're all seeing slightly different views, then we won't all have quite the same experience. I think there's something to be said for having a particular view of the scene intended by the director.
Why OpalCalc is the best Windows calc
I watched the video. The summary is very misleading - it's talking about where this may, someday, end up. Looking at the so-called real-time hologram, without foreknowledge you wouldn't be able to guess what was being reproduced, even if you were given 20 guesses. Someday this may end up as something cool - maybe.
This is only news because hacking the Kinect is currently a trendy topic in certain tech circles - so any Kinect-related story is getting airtime, no matter how immature (speaking tech-wise) and non-newsworthy.
#DeleteChrome
Oh My God, Yet Another Star Wars Release over the Horizon. Hide those news from Lucas immediately!
What's the point of wasting CPU and bandwidth on real time?
Perhaps their demonstration would be more impressive if they focused on actually generating a passable pre-rendered video first.
^^vv<><>BA
Came for Slave Girl Leia. Leaving disappointed.
Eh, even a slightly different angle will make you miss that face in the dark background that is just around the corner.
3 fps, 80 scanlines, in the wrong color, against a black background. Genius recreation guys.
You're sending slow-ass plain text from one computer to another and you call this thing ARPANET? Genius idea, guys.
I dont really know if it would really work or not, but I've had this idea for an interferometer based "holo-tank" for over a year now.
(I really don't care if somebody steals this idea.)
The phenomenon of self-interference is the life-blood of traditional holography-- basically, one beam is split in a beam splitter, one of the resulting beams scans an object, while the other then interferes with the refracted light from the scanning beam as it exposes a photographic plate.
traditional holography
This stores the interference pattern on the plate, so that when it gets illuminated by laser light of the same frequency, a virtual 3D image of the scanned object gets produced.
That's basic holography; The idea I have in mind is quite a bit different:
Since this is slashdot at least some of you guys will be familiar with the micro-mirror arrays found in some modern DLP projection television sets, (For those that are not, here is an obligatory wikipedia link.) and probably some of you already know about multi-mode lasers for use in frequency combs. (Another obligatory wikipedia link.)
Essentially, you take the beam from a multimode frequency comb laser that is calibrated to produce a series of discrete frequency spikes within the visible light spectrum, and run it through a beam splitter, just like traditional holography.
However, instead of sending one beam to interact with a real object as the scanning beam, you direct BOTH beams onto DLP chips. These DLP chips reflect and refract the laser light so that the beams will have a very subtle phase incongruity when they intersect within a transparent medium. This causes the beams to interfere with each other and scatter at the point of intersection. By carefully controlling the beam lengths to be highly specific to the individual frequency spikes of the laser comb's beam, you can modulate the apparent "color" of the glowing 'dot'. (Or, at least I think you should be able to anyway.)
Now, if you "Scan" the two lasers over the DLPs, you should be able to use them to produce a purely computer generated holographic image, in something that would approach real color. (Would not be true real color, because of the discrete nature of the laser comb you are using.)
Due to issues of blinding people with the laser light, you would need to project the image inside of a transparent block of material, like high clarity glass or crystal, with some kind of beam trap at the far end-- however, this "tank" doesnt need to be very thick to theoretically produce a nice 3D object. I would think a mere quarter inch thick would be more than sufficient.
Someone call Uwe Boll!
Homonyms are fun!
You're driving your car, but they're riding their bikes there.
Yes, but it's a proof of concept. Keep in mind that this is using off-the-shelf hardware. If someone picks this up and starts to work with it, then it's only going to get better with time. I'd imagine the first televisions would be similarly "not any good", and then think back when telegrams were the only way to communicate with others. Give it time.
Our culture doesn't get smarter, it just finds new ways of being retarded.
This reminds me of some of the early attempts at television... also of equally lousy resolution due to bandwidth issues.
As mentioned in the article, true "holographic" representation of an environment would take an insane amount of processing and bandwidth. There are some "tricks" that can sort of simplify this issue after a fashion and still not require stereoscopic glasses or anything fancy on the part of the viewer, but even those have their limitations.
Making a credible Volumetric display is the real trick... something several people have worked on to some degree or another. I can only hope that eventually something will actually happen with the technology but in the meantime it is still and experimental toy and not something for serious work... yet.
This attempt here is nothing more than the equivalent of Felix the Cat as used by Philo Farnsworth on some of the early broadcast television tests.