Maxwell Render | Official public forum

I made some tests on bump mapping waves on water. The map came as 3ds export from Vue5Infinite, but I should have used a higher resolution.
On the water I used a dielectic (abbe 120, nd 1,33) and a pale green colour. All caustics layers on.
How does this look?


I was of course curios to see beneath the waves. I first had to make the water have a thickness so I could se the surface from below. When you are looking from a denser material into light you should probably use a different Nd number for correct refraction? This is using the same 1,33. I don’t get a realistic refraction on the bottom of the pool, but I guess that can’t be expected of bump maps? On the right end of the pool you can see some “refraction”, but that is probably shadows caused by the bump map? Are we expecting displacement maps in Maxwell anytime soon?


I tested to see how the surface reflection changes, and added some trees and a background image from Vue. Different waves here also.


I used SketchUp, Vue, Daz Studio and Viz2005+Maxwell.

Regards,
Sheik

first one rocks

Your first rendering with the calm relatively flat water looks good, really good! i like that one. But the 2nd and 3rd renderings just don't look right. Especially the weird refractions on the 2nd image, that just don't seem right. But over all those are some nice tests.
Also interesting that way you used Vue for the water bump, i never thought of that.

Good job!

/ Max

Yes the firs one is better, image 3 is actually setup for testing how the landscape would look from under the surface, but that just became a green blur.
I became curios of the “refraction” effect I got from my first test of importing from Vue. I didn’t get this effect in the second test as you can see above. I wonder why? Could it be because the scale in the imported model was really small, (as Vue doesn’t handle real units), so the bump on the surface probably created very high waves, and the distance to the bottom of the pool is small. I think the refraction is actually all shadows from the surface(?).


Sheik

I realised I can use the Vue terrain editor to turn my bump map back into a mesh, witch I can export to Viz. So I now have the caustics sort of working. If I just have the mesh with faces pointing up I get nice caustics, but the refraction from water to air isn’t right.


I then made a copy of the mesh just below the first one and flipped the faces. The refraction is closer to reality (still used Nd 1,33), but now I get some less light & caustics. Am I doing something wrong, or is this the way it works in the real world? I also wonder why the material in the pool looks pink, as I assigned a pale green colour to it.


Any suggestions are welcome. Next I should try to fill the pool, with fog?

Regards,
Sheik

Ok, the pool is now the colour I assigned it. I deleted the white bitmap I had in the reflection and glossy slots of my plastic material, witch made the colour change, but I now cant make if less reflective. This looks like a bug to me.
I made the surface calmer and added some fog/scattering to the scene, still doesnt look like real water. I think I am done playing with this one…


Sheik

Looking great

Sheik

Im just curious. Shouldn't there be fog under that water?
My point is, are there plans for photorealistic and physically accurate water volumes as opposed to just materials. Or both. You create a box and apply the shader and the volume of the box is calculated according to real world properties.

Tell me this is worth exploring!

Beers.

Jep, that is the same question I was thinking of Beers. I did use some fog on the last one of the underwater scenes, but not so much that it would look like water. The problem with using fog here is you will have fog on the surface to, and that will change the light. If it would be physically accurate the colour of the light should start to become more bluish as it travels deeper in water.
The surface is just two meshes on top of each other (one with faces up and one with faces down), so the pool is actually filled with air. I don’t know how voluminous materials are handled here with surface mesh objects. The volume/thickness does have an effect on the transparency as you can see from this test with four cubes with the same glass material.

The thicker ones (A,B,C) are less transparent than the thin one (D). I guess the absorbance value affects this (here 0,59). One cube is just a fronface, a backface and a top (A), so it would be filled with air. The next one has the front and back in glass, and the other sides have a diffuse material (B), so I don’t now what the inside would be. The next two are of “solid” glass (C,D). A,B and C seem to have the same transparency, or is A less transparent? From this test I would conclude Maxwell assumes that when light travels thru two faces with the same material, that the space in between the faces is made of the same material. Right? So when being inside the material Maxwell doesn’t know it is inside. Maybe some of our more enlightened users could shed some light on this.
Maybe voluminous materials could also bee used for smoke, fire or rain. NextLimit should have no problem building a “water editor” as they have also produced nice liquid simulation software.

Sheik

I agree Sheik. Next Limit have made quite a reputation for themselves. All of their software, new and old, is as robust as if it had been beta tested for a decade. As far as I know, Mac OS X is the only other software that gives me the same warm fuzzy feeling indise when I use it.

BTW Sheik, the names Subspark (Sparx) for short!

Beers is like saying Cheers.
And Im Australian so you can expect a few beers here and there

Cheerz,
Sparx

Dennis,
I can’t find the map I used, but this should be very similar. This map doesn’t tile, so you may have to do some post work for using it on a lake. Let me know how it turned out.

Sheik