• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Deformable Objects Video

I think it might be getting to a point where we're just trying to brute force these effects with beefy hardware, why use every ounce of power to render and calculate a zillion polygons when it can be much more efficiently used on a less power-hungry method. Why keep trying to roll a boulder over your tin can to extract the soup when you can use a small can-opener and little effort. The can opener affords you more soup and energy with which to eat and enjoy it.
 
"when it can be much more efficiently used on a less power-hungry method."
And this Method is?
 
What I mentioned before.

Voxel based deformation.

Sure it's going to be a little ugly at first but we have to start somewhere.
 
I don't think Voxel method is a better solution. It's more a different solution.
Both have there Pro's and con's. So it has strong point. But it isnt all better.
If the market would switch. somebody would say there is a better soultion metioning the method used now. Because off the Con's of voxel rendering.

Novologic dropped pure Voxel, besides the reason it needs hardware acceleration to compete to, but a big con rendering upclose was very bad. High resolution voxels needed.

But I was very impressed wenn I flew my Comanche up high and see the for then more realistic enviorment landscape. Commanche OverKill.

So they used after that a hybrid solution. Voxel for landscape, structures and object in normal method. Espacialy for there ground FPS titles Delta force series.
Till they dropped voxel completly
For a Pure Voxel soltion to be able to compete with the current Dx9/10 hardware.
It need a 700mil VPU to . More or less.

The problem with Voxel for deformation you must wrap a High detaild Voxel surface in a Barrel structure. To make it look good staticly. For the deformation, Physics get harry because each Voxel is a linked Particle linkt to the paticles/Voxels around it. Plus colision needed for each particle Voxel that is.
So while for a normal method a Deformable Barrel leed to a lot more High fidelity geometry. The Voxel solution would go way beyond that. Its very fine detail.
Wich would make a High detaild deformation but to heavy for a PPU to do.
Each voxel act like a Vertice and can be displaced.
Plus the Terrain solution of novalogic uses a Voxel surface wich is generated off a hight map. A Deformale Voxel object need a other voxel solution. The deformal solution for voxels?

So the way I see it.
Voxel and retrace solution are in some departments beter and correcter rendering solution. But because of this beter and more correct way. It is a heavier solution wich need more PU power and resources.
The Current solution is a aproximation that serves it purpouse. Worlds look believable.
And thus maybe the most efficient solution to get the best result in all circomstances.

Plus another Con? Merging Voxel solution with deformal Physics.
We know that the current solution use extra vertices wich can get manipuated by Physics computation and colisions to deform the object. Well how does a voxel surface would do that. Hight maps sound not so flexible for that.
 
well, the game developers can only make what can run on systems that are in the present time, so it is up to them to decide whether or not to add such physics to a game (or how much). they are obviously prepared for future hardware. it is kind of a tease showing us what we may or "may not" be able to run on current hardware. but i am satisfied with the way things are looking.
 
I think if more RnD went into voxels. It would beat out polygons altogether. Right now out GPUs cant accelerate voxels. If if we had a voxel acceleration method of some type they would be much more formidable. As for physics being applied to them I'm sure the brilliant minds we have today can figure out a solution.
Have you heard of Ken Silverman's work on voxlap? Look into it I haven't been ableto run the demos on this POS laptop but I hear it pretty good.
 
You asume that Voxel method is a lot of Pro's and no con's.
There are two levels of Voxels. Volume rendering 3D Textures.
the way medical hardware solution work with Cat scans.
Or surface voxels. Wat games use. That makes me more think of bumpmapping with interprolation.
Wich make me think that a tad different method of Pixel pushing.
What have GPU a lot off and in the futere unified shaders wich works like a General purpouse Co PU and G80 has 128 off them double puped.
that can be used for Voxel prosces in a way I might think.
Sm3.0 to BTW to.
GPU renderpipe of today and nextgen aren't so much Fixed anymore.

CPU are general purpouse a lot more then those new shaders but those are getting with each generation more general purpouse.
 
You must log in or register to reply here.
Back
Top