Jump to content
  • Advertisement
Sign in to follow this  
stephanh

Ice

This topic is 5472 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi, i am looking for ways rendering ice surfaces (the ones found on galciers). BRDFs might be an option, although a little bit too complex and hard to get a model or captured reflection data for ice. Simple cubemap reflections dont account the important anisotrop. features of ice. Are there any other ways to simulate/render fancy looking ice surfaces? regards, stephan

Share this post


Link to post
Share on other sites
Advertisement
Quote:
BRDFs might be an option


not BRDFs, but BSSRDFs would be, ice has a _huge_ amount of subsurface scattering going on.
you should look into ways of faking sss, there are a few papers around talking about sss rendering on graphics hardware (ATI might have some, they had sss demos for the x800). and I don't really think ice has any really high anisotropy, if any at all...

Share this post


Link to post
Share on other sites
Quote:
Original post by sBibi
Quote:
BRDFs might be an option


not BRDFs, but BSSRDFs would be, ice has a _huge_ amount of subsurface scattering going on.
you should look into ways of faking sss, there are a few papers around talking about sss rendering on graphics hardware (ATI might have some, they had sss demos for the x800). and I don't really think ice has any really high anisotropy, if any at all...


thanks for the suggestion (and corrections :)) seems i messed around with terms alot.

I found a chapter on subsuf.scattering in gpu gems but i think that is out of question due to the high gpu demands.

So pure faking would be also ok, i shouldnt have used the term simulating.

Share this post


Link to post
Share on other sites
if you want pure fake sss, you can use a color shift + ambient in your lighting calculation.
(the color shift basically consists of moving the shadow line back a bit when you compute the N.L coeff, and shift it towards some color that would depend on the surface (a blue-greenish color in your case, while the normal diffuse color would be white-blueish)

you also have better ways to render a translucent object (but costs more than the simple color shift above), you'll have to additively render front faces depths first, then subtract back faces depths. now for each fragment of your translucent object you've got a buffer containing the depth of the material along the line that goes from your eye through that fragment, and you'll just have to use that depth to lookup a texture that will give you the material color/transparency/whatever depends on the material depth.
but on older hardware, this last methods is really expensive (still much less than an sss approximation), you'll need quite a lot of passes depending on the card (you can get away with two passes on latest hardware)

Share this post


Link to post
Share on other sites
If you can scrounge up some articles on how Halo did it's Ice, you'd be in the right area. I don't know how it's done, but I'd assumed it's fake, and it looks really good.

I'll have a quick glance around when I get home(on the school PC at the moment).

Share this post


Link to post
Share on other sites
Try using a noisy gloss power map to simulate variable light scattering in real ice, caused (usually) by small air bubbles inside the frozen water. This, in effect, forms the "sparkliness" evident in most icy surfaces.

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!