Atomontage Engine
category: general [glöplog]
http://www.atomontage.com/
Seems to look promising, or is it another "Unlimited Detail Technology" stuff?
Seems to look promising, or is it another "Unlimited Detail Technology" stuff?
hmm, at least the demo's are more interresting.
There was another thread about that.
Looks more promising than that other one, and it's by a scener it seems (there's a pack of 256b intros on the site, including this one: http://www.pouet.net/prod.php?which=5522). But then he's going on about simulating quantum mechanics and electricity in the voxel world, which sounds somewhat nuts :)
Who cares, I like watching videos of this new voxel stuff.
Btw, anyone remember an amiga demo with a "non-blocky" voxel engine? It was like the voxels had been smoothed or something.. probably mid-90s AGA.
Who cares, I like watching videos of this new voxel stuff.
Btw, anyone remember an amiga demo with a "non-blocky" voxel engine? It was like the voxels had been smoothed or something.. probably mid-90s AGA.
Perhaps the rescue on fractalus-ish landscape in Roots 2.0?
Ken Silverman, author of Duke3D's engine, wrote something with similar rendreing a while ago:
http://advsys.net/ken/voxlap.htm
Same voxel style, real little cubes...original and great, and hundred percent dynamic (you could destroys walls one woxel at a time).
However, I hardly see this working for real games if there's no way to make things smoother when the voxels are big enough to become big cubes on screen.
Besides, hardware on-the-fly tesselation somehow looks like the "final" solution to rasterizers limitations...not the end of history of course, there can and will always be completely new ideas, but quite seems to do the job...
Thoughts?
http://advsys.net/ken/voxlap.htm
Same voxel style, real little cubes...original and great, and hundred percent dynamic (you could destroys walls one woxel at a time).
However, I hardly see this working for real games if there's no way to make things smoother when the voxels are big enough to become big cubes on screen.
Besides, hardware on-the-fly tesselation somehow looks like the "final" solution to rasterizers limitations...not the end of history of course, there can and will always be completely new ideas, but quite seems to do the job...
Thoughts?
Quote:
he's going on about simulating quantum mechanics and electricity in the voxel world
I think he just means that he's going to apply some "montecarlo" method to approximate physics and save on computation time.
useless, we have MEGATEXTURE(TM) technology that does this way better without the voxel crap
discuss!
Hollowman: no, not like that.. it was (from what I remember at least) a fairly standard colourful voxel landscape, but it wasn't 'boxy' at all.. no cubes were visible. It wasn't polygons too, it was definitely voxel but somehow filtered to look smooth.
broderick: yeah probably just a bad choice of words.. trying to simulate a large style world based on a quantum mechanics would be hardcore to say the least.
Electricity simulation is more reasonable, as you could assign a conductivity value to voxels and then work out where the electricity flows.. the calculations would get nasty fast though as electricity flows pretty much instantly so you could end up with a LOT of voxels to test each frame. But he does at least say that heat conduction will be much easier for just that reason, I count that as a good indication he's not going down uncrashable API road :)
Electricity simulation is more reasonable, as you could assign a conductivity value to voxels and then work out where the electricity flows.. the calculations would get nasty fast though as electricity flows pretty much instantly so you could end up with a LOT of voxels to test each frame. But he does at least say that heat conduction will be much easier for just that reason, I count that as a good indication he's not going down uncrashable API road :)