University of Illinois uses a Cluster for Immersive VR

It seems the folks down at the University of Illinois at Urbana-Champaign have created a 6-sided CAVE like system called ALICE. But, instead of running it off of a SGI Onyx, they've developed a distributed environment for visualization called Syzygy. Slap a few computers together and make your own holodeck!

Comments

Great!

[darkov]

This should make for some stupendous pr0n! Come on all you vouyers, start stringing those polygons together.

Programs and graphics for this needed NOW

[Warvi]

See their HOWTO for building up your own setup at


http://www.isl.uiuc.edu/ClusteredVR/szg/doc/Para me terExamples.txt

That seems like pretty doable by any geeks with enough boxes.

That would say if the 3D immersion has any usability to it at all, it will be in common use in 10 yrs. It might become the next big thing in living rooms like TV became 40 years ago.

However, what makes a new technology break through is not what it enables, but what you get with it. TV would have had no use whatsoever without the television programs made to be watched with it. Computers only broke through when there were programs for it.

Umm, I sense a great opportunity for all people who can really do great 3D graphics.

Not to mention what I already said about getting 3D GUIs off the ground at


http://slashdot.org/comments.pl?sid=25035&cid=27 18 945

Re:Programs and graphics for this needed NOW

[Mike Connell]

That seems like pretty doable by any geeks with enough boxes.

#include "Wry smile.h"

6 linux boxes and the clustering software is only half of the problem. The harder half is 6 or 12 videoprojectors (or more!), the mirrors and the (back) projection surfaces for the CAVE. Add in the tracking hardware (cost and complexity (ie EM interference)), and you have a lot more work to do.

Don't get me wrong - it excellent that it's feasable for a department to install their own fully immersive VE without an Onyx and a team of engineers, but you'll still need a team of grad students, builders and time.

Also dont forget that you need a *big* room for one of these too - say you have 2m high surfaces, you basically need 6-8m of vertical space once you include the space for the projectors and mirrors for the top/bottom 'walls'

Re:Programs and graphics for this needed NOW

[Mike Connell]

You could - it just wouldn't be very immersive. With a 3D environment in a CAVE you dont just stand in the centre, you actually walk around a bit. If a virtual object is in the way, you might lean around it, or crouch down to look under it, just like in real life.

This means that the views have to be created based upon where your head is - which is why at least one person will have a head tracker, if you dont have this, it doesn't feel immersive, it just looks plain wrong.

Of course, you could keep your head in the centre, and not move it but it kind of ruins the point. Also keep in mind that I am taking 3D stereo display for granted here, but it's not so easy with real cameras, especially if you want 360 degree recording.

Often people stick video images onto 2D objects (like the front of a virtual TV) within the 3D environment. That's pretty cool :-)

0.02

Re:#include "Wry smile.h"

[Warvi]

It's easy to play the devil's advocate, but I think there's more poewr for enthusiasm than downplaying here.

The harder half is 6 or 12 videoprojectors (or more!), the mirrors and the (back) projection surfaces for the CAVE. Add in the tracking hardware (cost and complexity (ie EM interference))

Their source is GPL'd; it could be modified to use standard or flat monitors in a downsized setting.

you'll still need a team of grad students, builders and time.

Same goes for developing operating systems.

How many /. readers are grad students and builders, I leave as an exercise for the reader to wonder.

you need a *big* room for one of these

Unless you downsize it. How many of us have spare space for wondrous technical projects...

The electronic (wall)paper is being developed at least by IBM and, err, was it HP (my memory fails me here and I'm too lazy to check /.'s comprehensive archives on all the stories of this subject). Say it will hit market 5 yrs from now; you'd still have 5 years to make a product out of this to get it to market by my wildly suggested deadline of ten yrs from now.

All that is needed for this breakthrough is all that work to create the usable content, just like DVD's need movies to sell.

I am sure some of us are willing to put together the effort in small groups of friends interested in this achievement. Just for the fun of it!

Re:roll your own

[Anonymous Coward]

The Cornell Theory Center (www.tc.cornell.edu)
has a 3-wall cave setup that uses 3 off the shelf dell workstations running w2k and software from vrco.com. Seems to Just Work; I played quake on it a couple months ago.

Re:#include "Wry smile.h"

[Mike Connell]

Noting that this is clustering software for a *CAVE*.

1. You can't use "standard or flat monitors" in any form of "downsized setting", because unless you are 6 inches high, it's not going to be immersive. You can't stand in the cubic space formed by the faces of 6 monitors, flat or not, unless you are a person of very very small stature. Even if you can find a 6" high person to get in this contraption, they'll probably get fried by all the EM radition & heat after a few minutes, so you'll need a supply of these people...

2. Bringing up kernel.org looks like blatent karma-whoring - it's completely irrelevant. The problems you missed, and which I pointed out to you in my first post are the physical problems - you have to build an immersive space environment - a physical one - out of wood. You can't email an 12 foot high wooden structure around on the internet you know!

3. (You need a big room for this) "Unless you downsize it". Oh sure, if you can also "downsize" your users. You have to build something that you can get inside of. There's a limit to downsizing - at the limit is your body!

If you're only going to use a powerwall or something it's a different kettle of fish, and you can just hook up WireGL or something of that nature with what I expect will be a lot less effort.

So as I attempted to point out (and obviously failed) in my first post - there are real physical engineering problems that make the construction of your own CAVE environment much much harder than downloading some code and playing with the configuration files.

0.02

Iowa State's Immersive Environments

[john@iastate.edu]

ISU has probably the premier researcher in this area, Dr. Carolina Cruz-Neira. Here's some info on the environments at ISU: C6 and more...

How long until Nolan Bushnell sues?

[Junior J. Junior III]

He originally wanted to name his company "Syzygy"...