Did A Google Earth Mistake Cause A Couple's Brutal Murder?
Redditor Shamed After Girl Finds His Request For Her 'Porn…
Unlimited Detail claims to leave current 3D technologies in the dust
Unlimited Detail is definitely the most interesting technology demo I stumbled on today. In a nutshell: current 3D technology is based on polygons. Each 3D shape you see on the screen is made out of multiple straight facets (polygons). The more polygons (or facets), the rounder and more natural it seems. The current battle is all about polygon counts -- how many polygons can a certain graphics card render per unit of time. This metric has been rising 20% per year on average, for the past few years.
But polygons aren't the only way to display 3D information. Unlimited Detail works with dots, rather than polygons. The thinking is that if we use 3D "pixels" instead of flat shapes, objects can be far more realistic and lifelike because you don't have as many straight surfaces. Each object is composed of a mass of "dots" virtually positioned in 3D space.
This, in itself, is not a novel idea. A very similar principle exists in voxel technology. What is different about Unlimited Detail is the way they select what pixels to present. And this is where the "snake oil" bit comes in, because there is absolutely no hard proof for this one. They say they're applying "search technologies" to figure out exactly what pixels to show. In other words, while an entire 3D scene may contain billions of "points", you may only have 1024x768 (786,432) pixels on your monitor. So if there's a very fast way to figure out which of these points must be shown on each of the pixels, you get very fast 3D technology.
And this is what Unlimited Detail say they can do. There's an impressive (if quite smug) screencast after the jump, with a good explanation for novices (and potential investors, I'm guessing -- the site seems quite shoddy at the moment). Graphics experts in the audience, help me out here: Is this snake oil, or is it really the next big thing?
Comments
25
Subscribe to commentsArnieMar 10th 2010 5:31PM
Did not seem to give any details what so ever. I would like some known quantity to use this. How about some small game? Something playable. Also a lot of games now a days also use dynamic stuff. I wonder how it scales. How stuff changes around.
I wouldnt be quick to call them out as snake oil folk as those two engineers formerly of Apple have created hardware which makes Ray Tracing possible whereas most folks had said it wouldnt be possible. All I am saying is that I need more details proof etc. regarding this.
SugarDaddyMar 12th 2010 10:50AM
There's probably no known quantity using this tech because they haven't developed the SDK yet. It'll likely be a few years before this gets any kind of foothold in the market. After all, there are currently billions of dollars in the polygon market. This is sure to piss a lot of people off.
Holographic RealityMar 10th 2010 5:42PM
That's fractal geometry, yeah - I sense fractals there, especially after seeing that pyramid structure. The mathematical approach to generating 2D fractals is also quite simple as is the 3D technology described here. It's actually quite wierd that no one has thought of this before, applying the fractal algorithm that displays only the relevant pixels. Brilliant.
JonnieMar 10th 2010 5:56PM
Smells like a revolution to me! Hehehe, no wonder ATI and Nvidia are scared of it; this technology truly renders their hardware efforts obsolete! Talk about money "well" spent. Daayumn! :)
Oh boy!
JustinMar 10th 2010 6:07PM
Its only useful for static images because its supremely difficult to animate or use lighting effects on it.
CarlosMar 10th 2010 6:23PM
I wouldn't say so. They mention on their website that, quote: "Unlimited Detail can use a variety of different lighting techniques, in this case, we are using a shadow map.".
I think this technology could really be a game changer. Pun intended ;)
OskieeMar 11th 2010 8:23AM
Im with justin... i find it weird after watching it, that nothing was animated.
JustinMar 11th 2010 1:35PM
Dont look to this company to be a game changer. this tech was already discussed to be the post Rage engine for id tech.
Warner YoungMar 10th 2010 6:26PM
I think my questions are: What form is this data that is ultimately sent to the GPU? Are they essentially sending a 2D picture at this point, or do all their "unlimited" points have to be converted into a 3D format that the GPU can handle?
Is the algorithm handled by the system's CPU or GPU? How much data can you realistically handle on a slow (e.g. Atom) CPU with low amounts of memory? How well does it handle dynamic light sources, or translucency?
SimonMar 10th 2010 7:27PM
Hmm, it appears that the algorithm creates a theoretical 3D plane, then creates a slice, or a sort of a cut-through, relevant to the viewing position, which then basically is just a simple 2D image, the rendering of which can be handled by even the ordinary CPU, utilizing negligible resources. Very interesting indeed.
I hope it catches on as it does seem to be a break-through of an approach.
Now where's that popcorn, this could be quite a show! :)
The CheeseMar 10th 2010 7:06PM
ohhhhh....NOOOOOOOO!
THIS WAS MYE SECRET PROJECT....PROJECT ALGO!
ViaTorciMar 10th 2010 7:36PM
Ahh, sparse voxel octrees. Carmack intends to use them in the id Tech 6 engine (post-Rage). Here's what the video didn't mention: you can't animate these. They are as utterly inflexible as sprites. The closest you can come is to define every frame, can still have amazing object detail, but can't be procedurally animated (e.g. by ragdoll physics) and will move at a fixed framerate with no obvious tweening method. You can have this voxel world and it will look awesome, but it will be almost completely static. By the sound of things this company is using a binary search like the PULS 256-byte raytracing demo, so at least bounding volumes will work and moving static objects around won't be a huge performance issue.
Five-sided lettuce, on the other hand, can be tessellated into thousand-sided lettuce and molded in accordance with its depth & normal maps, which can in turn use virtual texture quadtrees to have infinite two-dimensional detail (see id Tech 5). The underlying model of dozens or hundreds of polygons can be warped, skewed, or reshaped at an arbitrary framerate from sparse keyframes or procedural instructions.
I like the technology presented, but the presentation itself is vapid and disingenuous. This is not a cure-all. Even Carmack is only using it in conjunction with polygonal actors.
http://www.reddit.com/r/gaming/comments/bbg9c/unlimited_detail_the_end_of_poligon_based/c0lxc1e
Ian WoollardApr 10th 2010 9:44AM
I'm pretty sure this system would permit animation. The scene is already moving as the camera moves around, and animation is just a different sort of movement.
The normal way people use voxel oct-trees is to prerender the entire scene, but that takes lots of memory and is very slow. I'm pretty sure he's only rendering the voxels that are actually visible- essentially using the net of voxels as a cache to save recalculation costs for each frame; and creating new voxels only when he needs to because there's a visible gap.
DeoWulfMar 10th 2010 7:44PM
I don't know. Nothing this good ever works like it should.
MuskieMar 10th 2010 8:12PM
Look closely at 0:58. at the chat window. they're using a WOW EMULATOR.
Hellooo lawsuit!
AndersMar 10th 2010 8:19PM
Did you really have to be the Mr. Dickhead of Perception?
Geez.
KevinMar 10th 2010 8:26PM
You sharp-eyed tosser.
MuskieMar 10th 2010 8:48PM
I love how i'm being both adonished and applauded at the same time. the WoW Forums are starting to heat up about this too. :P
Pbutt77Mar 11th 2010 3:16AM
Well we should hope this is halfway true atleast. If we could integrate it with the current way graphics card do things...it could take a huge workload off gpu and cpu possibly. I agree animation, lighting an ect seem questionable. But what if we use this only for the stationary scenery and objects and let gpu/cpu handle the rest. The synergy between all of this technology could be exactly what we need to free us from visual limitations and push us pass the borders of our creativity. Developer cost would drop, everybody would still have their jobs, deadlines would be less of an issue hopefully and that means more free time to focus on quality and quanitity as the only bi-product! zzzz....oh sorry, just woke up!...thats was one hell of a dream!
Cile1977Mar 11th 2010 5:21AM
Isn't this similar to what Microsoft has done with 2D graphics with SeaDragon technology? With Seadragon one can put very big resolution pictures on web page - seadragon shows only pixels that can be seen at current zoom level - when you zoom in it again shows just pixels at that part of picture.