Unlimited Detail Real-Time Rendering Technology revisit

[krele]

To me what it looks like is just polygons that zoomed way out. Take the rock for example, a normal polygon one would be 30 or what not, but they have theirs to be like 300, in order to do that they made it very big, like 100x the size of the original and then zoomed the camera  out so it looks like what they call "unlimited detail". Its just polygons and not that crap "atoms" This is a fake.

And what exactly would they gain by doing this? What makes you think they couldn't create a small rock with 300 polys, that would be so much better for both memory and performance. These are not polygons, these are voxels.

[T-Rok]

Once it can animate as well as polygons, I'll be impressed. Until then, voxels can go die in an art museum.

[ARTgames]

Once it can animate as well as polygons, I'll be impressed. Until then, voxels can go die in an art museum.

http://www.youtube.com/watch?v=tkn6ubbp1SE

[T-Rok]

That's old news. I'm talking animating an entire world and keeping a steady 60fps. A single animated object using voxels quite frequently just barely reaches 30fps.

[ARTgames]

Well you did not say that before. I don't know. Have not looked into it.

[T-Rok]

Yes, I realize I was very obscure about it, sorry haha. I myself have looked into it with my C++ buddies. They all agree that voxels are the future, every single one of them, with the condition of A: A future machine powerful enough to handle it available to the every day consumer. or B: Programmers figure out a way to animate better with voxels. I asked about the idea of a voxel world with top quality vectors doing the animations, but apparently it's not viable.

[RainbowDash]

I agree. I can see a time when something like this can be a viable way of doing things. They need something to push the power of new hardware.

[krele]

Yes, I realize I was very obscure about it, sorry haha. I myself have looked into it with my C++ buddies. They all agree that voxels are the future, every single one of them, with the condition of A: A future machine powerful enough to handle it available to the every day consumer. or B: Programmers figure out a way to animate better with voxels. I asked about the idea of a voxel world with top quality vectors doing the animations, but apparently it's not viable.

Still, this kind of technology is impossible to make solely because of memory problems. Even if it could be animated, it would eat memory hundred times more than polygon models. I doubt the machine power could skyrocket like that :S

[T-Rok]

Yes, I realize I was very obscure about it, sorry haha. I myself have looked into it with my C++ buddies. They all agree that voxels are the future, every single one of them, with the condition of A: A future machine powerful enough to handle it available to the every day consumer. or B: Programmers figure out a way to animate better with voxels. I asked about the idea of a voxel world with top quality vectors doing the animations, but apparently it's not viable.

Still, this kind of technology is impossible to make solely because of memory problems. Even if it could be animated, it would eat memory hundred times more than polygon models. I doubt the machine power could skyrocket like that :S

Agreed, I watched a man animating a single voxel character and it took the fps from 600 down to 29 just animating it. Scientists can only pack so much into such a small space. Eventually we'll reach the limit and have to find a new way to be faster. Which is why I wish voxel worlds and polygon animations could work together.

[Lingus]

Some day (probably soon) this conversation will be silly and we will laugh about it.

[T-Rok]

When that day comes, I'd like to see them do precise collision checking with voxels. trololol. Billions of atoms checking for collisions? -_-"

[Lingus]

I'm just saying, it is a common thing for people to talk about what is "impossible" to accomplish and then be absolutely and completely proven wrong within a short period of time. I find it interesting, given the fact that the rate of progress continually increases, that people still feel that way. What we have seen happen in the past 10-20 years with technology should really show anyone that if we can't do something today, we'll probably be doing it in 5-10 years.

[T-Rok]

Oh, not saying its impossible. I'm saying currently it is. No matter how much they tweak their "engine" and such, we don't have the processors capable of handling it yet.

[Freeforall]

I can't wait until graphics look truly 'realistic'

Games say they have realistic graphics, but they are nothing compared to the real world. When will they be truly realistic?

[Jake]

I can't wait until graphics look truly 'realistic'

Games say they have realistic graphics, but they are nothing compared to the real world. When will they be truly realistic?

Well, static rendered scenes without people have gotten to the point where they are more or less photo-realistic. Only problem is that they can't be rendered in real time yet, but we're getting there soon. I'd say another 10-15 years before we can run extremely complex, photo-realistic scenes in real time. Even at this point though, I fear we will still be able to tell that living, organic things are 3d models, simply because of the amazing complexity that goes into a facial animation, the way our muscles flex and move, etc. Obviously, we're making amazing strides with this already, but there's still so much farther to go until a 3d face and it's subtle facial cues are indistinguishable from a real life person.

Ever see a 3d scene, and you're like "wow, is this real or fake? And then you see a 3d modeled person, and instantly realize that it's fake.