Slashdot Mirror
IBM Develops New 3D TV Technology
neutron_p writes "IBM has recently announced a new and affordable 3D video system that works with normal DLP (Digital Light Processing) televisions. IBM demonstrated the new system on a 50-inch, flat-screen Texas Instruments rear-projection digital television at the 22nd annual Flat Information Displays conference held in San Francisco this month. This "black box" device can be connected to any DLP projector or television via the common VESA 3 pin stereo connector. Exact details concerning the 3D technology - still unnamed - were not forthcoming, but the company spokesperson said it was compatible with OpenGL and Direct Draw, which is definitely aimed at software developers who make 3D games."
Yeah, and the sysadmins at IBM just noticed a proliferation of sex sites in the proxy logs,
I'm not it will work. People don't like wearing glasses, except maybe gamers. OTOH, Sharp3D seems to be a more promising way. What do you think?
Million Dollar Screenshot
Bang bang tish!
public class null extends java applet { System.out.print ("Tabula Rasa"); }
On the downside, you still need 3D glasses to correctly view the image and practically no video is shot in 3D as it requires more expensive cameras, but as price drops and general interest rises, this is sure to change.
Uh-huh. And we're sure to get virtual reality sometime soon as well.
It's great that IBM (of all people) have developed a system to allow 3D movies at home. But the problem is, there's no content. And I doubt just because there's a $1,000 US piece of equipment on the market that content will suddenly come spewing forth. ESPECIALLY when glasses are still needed for this to work.
This is a chicken and egg problem, but unlike DVDs and High Definition televisions, 3D has been promised for quite a while but has yet to come. The article makes it sound like IBM has made a giant break through and 3D movies are about to become common place. I doubt very much this is so (especially while we need glasses for these things). This is a positive step, but it's a small one. The adoption of 3D movies will come one day, but the road towards it will be paved with lots of small steps, and we aren't anywhere near the end.
Creating 3-D games won't be that hard. All the game software needs to do is render each frame twice with a slightly different POV corresponding to the right and left eye. The only downside is that frame rates for the game will probably almost half what they are for the mono version (assuming that rendering is a significant chunk of the total CPU/GPU processing budget).
Two wrongs don't make a right, but three lefts do.
This may not be such a thing as chicken and egg. Maybe it's jut that people like 2-D TV. As has been noted many times, esp in regard to the ipod, video+audio is not always superior to video only. Stereo graphic books have been around for 200 years and they did not take over regular books. 3-d movies have not taken over regular movies. Holograms have not take over photographs.
When I wear a pair of Crystal-eyes doing anything elese with my computer or in the room is prohibitive. Somethings just work better as 2-d experiences. Our brains understand that not everything in the 2-d picture should be in focus. But in 3-d everything shoul dbe in focus if we focus our eyes on it--but that won't happen here. only some of this will be in focus so it's going to be mighty strange for our brains.
Some drink at the fountain of knowledge. Others just gargle.
At my last job at SEGA, my lab was in charge of coming up with many different and pioneering ideas for new ways to play video games, many of which, for one reason or another, never made it to market.
One of those was HOLO-GENESIS. It was a 3-D laser holographic projection device for the MegaDrive/geneis. It could have displayed 3-D rendered images, in full-color, in real-time, using a system of 3 red/green/blue lasers, and a finely-meshed micro-faceted surface which gave a pseudo 3-D effect based on carefully utilized light diffraction effects, a la printed holograms.
It was slated to come out in mid-1995, but at the time, we couldn't get a acceptable frame rate (3-D graphics accelerator hardware was still very primitive and expensive, the province of SGI workstations and arcade machines), so we decided to not commercialize it at the time.
In any case, I must say, this is a very interesting announcement, and I must congratulate IBM for further and seemingly admirable work on bringing such technology to the market. Hopefully they can continue to lower the price point and make it adopted wider.
-- Samir Gupta, Ph. D. Head, New Technology Research Group, Nintendo Co. Ltd., Kyoto, Japan.
Your left and right eyes can't always see the same set of objects.
If you use occluders and occluder volumes, then you run the risk of missing objects that were occluded from the view-point of the primary eye, but are not from the view-point of the secondary eye. So, you have to perform occlusion tests per eye or not use occluders at all.
You also run the obvious problem of objects "popping" in from the sides of your vision. There will be a moment when an object approaches from the side, into your peripheral vision, that the object is only visible from one eye. If this is your secondary eye, then I guess you'll miss it. When it is finally visible from your primary eye, it will suddenly "pop" into view. This problem can be fixed by performing your visibility tests from neither eye, but from an in-between eye with the view frustum scaled up to encompass both eye's frustums.
I just wanted to point out that this wasn't as easy as your made it sound. You can't just draw the primary eye, remember what objects you drew, and then draw them again from the secondary eye. Atleast, you can't do that if you want accurate results.
Justin Dubs