GI method and dielectric materials

[simmsimaging]

Just wondering what would be the best GI method for rendering the most realistic dielectrics - namely fluids like wine/juice etc and glasses. I am not sure if it makes a difference or not, but thought that that the light distribution could make a difference in the quality of the reflections and refractions - but maybe not. I'm trying to test it out, but it's very slow going rendering out these mats and I don't have that much time before my deadline!

I can't even start on caustics - it seems totally out of control and the testing takes way to long so I will have to live without that I think.

Any tips would be much appreciated.

Thanks!
b

[Paul Oblomov]

Well, caustics is really simple to control. Plz make some test scene in max9...
Don't forget to check reflect on backside, use some fog and that's it.

[simmsimaging]

Thanks for the tip Paul. I was not aware it needed fog, but the material I am using does have fog in any case. I can't say I find it very easy - I still find I get nothing or tiny splotches and playing with the various parameters isn't showing much in the way of differences - and it's taking so long to render tests that I just had to give up for now. I'm sure it's me and I may return to it later.

So far I am finding no real observable difference in the dielectrics with various GI modes.

b

[MoonDoggie]

Hey Paul, I've tried using caustics on simple scenes time and again and for whatever reason I can _never_ get them to work. Do you remember that simple scene daforce setup or something with the bathtub of water and the blue tile? I downloaded that and they worked great, I tried to recreate it and nothing. Any really quick step by step for getting caustics to render through a glass sphere on a plane?

-Colin

[vlado]

What kind of lights are you using?

Best regards,
Vlado

[simmsimaging]

What kind of lights are you using?

Best regards,
Vlado

Which question is this relevant to - caustics or GI? Both apply in my case, so I have been using Vray lights in plane mode. In the current scene I am using an HDR image of a softbox light as the texture as well.


b

[vlado]

Well, normally it's as easy as creating a VRayLight, the sphere, and turn on Caustics from the Indirect illumination tab of the Render Scene dialog.

Best regards,
Vlado

[simmsimaging]

Is it critical that it be a sphere light? Is that why it's not really showing up as more than speckles even with the subdivisions of the light at 15000? It also seems that the caustics are being cast *through* my liquid but not showing up *in* the liquid. I'll try and render out that direct caustics map (from that script you gave in an older thread) to show what I mean.

b

[simmsimaging]

Here is a quick test. The fluid is a high res mesh out of realflow (although it's a bit odd looking it's fine for my purpose )

First jpg is straight render, no caustics on. 2nd is the caustic map and the 3rd is the render with caustics using default settings. The light is a vray plane, and I turned up the caustic subdivisions to 5000 and the multiplier on the light to 5. Without turning that up you couldn't see any specks of caustic within the liquid at all - it was totally black).

What settings need to be changed to get this to look more like regular caustics on the background, but more important to me would be to get some nice light interaction happening inside the fluid?

Thanks for the help!
b

[MoonDoggie]

Looks excellent. Thanks.

[vlado]

To get smoother caustics from this you will need to increase the Search distance for the caustics.

Best regards,
Vlado

[simmsimaging]

To get smoother caustics from this you will need to increase the Search distance for the caustics.

Best regards,
Vlado

Thanks - I will try it.

b

[simmsimaging]

Vlado - I tried raising the search distance but now I'm getting a weird problem where only part of the image is changed (see attached shot, caustics are smoother at the bottom and right). Any idea why this is happening? I get this with and without DR turned on. Sorry the image is so small but generating the caustic map is very slow for iterations.

b

*EDIT*

Just did a couple more small renders with a different crop to show the problem better. One with caustics, one without. I tried several GI options to make sure that wasn't it - but the no caustics is IR/LC and the caustics one is IR only - but it was the same even with BF/LC

Thanks
b

[Paul Oblomov]

Khem where's da scene ?

[simmsimaging]

Hi Paul - just PM'd you a link to the scene file.

Thanks!

b