[Eric Lagman]

Ok material masters I need your advice. I am trying to simulate water drops on a surface using this mxm http://mxmgallery.maxwellrender.com/sea ... earch=rain

My question is how would I set up the geometry to get this to look right?

I have it set up right now with the substrate the drops rest on as a single surface then the drop material on an enclosed volume that is offset from the substrate. I only put the displaced material of the water drop on the top surface of the enclosed volume so that the drops displace outward. On the bottom surface of the volume I just have a plain water material that is clipped with the drop alpha mask.

My concern is that the clipping mask opens up the side of the drops and causes the material not to refract properly. Is this a valid concern? If so how would one model this?

I first just but the displaced drop material on the entire offset enclosed volume, but then the drops would displace in both directions and the bottom surface would intersect with the substrate right?

I am stumped on how to get this done correctly. It kind of reminds me of the liquid in glass refraction issue that Thomas helped solve long ago. The attached image illustrates what I have now, but It does not look right when I render. I will post screen shots when I am at my home computer of the results. Also Im not sure if the image link is showing up. Our internet here at work is crippled right now.

[KurtS]

My concern is that the clipping mask opens up the side of the drops and causes the material not to refract properly. Is this a valid concern? If so how would one model this?

I don't think there is any problem if the sides of the drops have tiny openings - as long as these openings are not visually disturbing. The opening will not disturb the refraction as long as you have some geometry/a surface behind the water drop.

[Eric Lagman]

Thanks Kurt. The opening created from the alpha is so small that its not visible. I was just concerned that it would mess up the refraction. I tried upping the precsion on the material. It came with a displacement precision of 120 and I tried lowering it to 10. I bumped it back up to 100 and it looks more realistic now. I think its because the geometry it is on is a single plane. Solidworks has no way of subdividing a plane so it comes in as a single surface. I might try to make my backdrop in Cinema 4d so I can subdivide it and go with a lower precision on my displacement. My benchmarks plumeted when I raised the precision up to 100 on the single plane surface.

[simmsimaging]

You will find it works better with high precision for sure - the shapes get very blocky without that. I am not sure if the alpha affects the refractions much - it worked fine for me, but I was just after the look and not necessarily "reality" so it may be off a bit. Doubt anyone would be able to tell though

I haven't used this mat in Maxwell for a while so I'm trying to recall for sure- but I think that with Maxwell you want to avoid having a single sided plane or you'll end up with the transparent stuff casting weird dark shadows. Make it with some depth and you should be okay.

b

[Eric Lagman]

I haven't used this mat in Maxwell for a while so I'm trying to recall for sure- but I think that with Maxwell you want to avoid having a single sided plane or you'll end up with the transparent stuff casting weird dark shadows. Make it with some depth and you should be okay.

b

You mean like the enclosed thin volume shown in my sketch? The only single surface I have is the one the drops are slightly raised off of which is just a basic plastic material. When you used the material on a thin surface with some small amount of thickness did you just apply the whole material and have it displace in both directions or did you pick the back side of the water drops surface and just give it a water material with no displacement. The displacement will go in different directions on each side of a thin surface with thickness and intersect with the underlying geometry of what you want the drops to sit on right? Hope that makes sense. The test I am running now with the higer precision is looking ok. Its hard to tell because it takes so long to clear and tell if there is anything wrong. I will post a screen shot when the render gets further along. Thanks for creating such a great material.

[simmsimaging]

Basically your sketch is how I was doing it when making a window pane with water on it. In order to avoid having the map displace in all directions I just adjusted the UV's for all faces except the one I wanted wet to show only a black portion of the map. That was the simplest way - but only good if you don't need UV coordinates for something else. Alternatively you could do multiple materials or map channels for the other sides, which I think is what you are already doing.

You may not need that air gap between the substrate and the water, but you'll have to test and see. You could possibly just add the water drop layer to the substrate material and have it pop right off that object too. I think that is how I did one of my tests - but honestly I can't recall. I have redone that water drop thing in 3 different engines since I made that one and each one was different!

b

[KurtS]

You will find it works better with high precision for sure - the shapes get very blocky without that.

I agree - even though theoretically you can achieve the same result on a non-subdivided surface, but it will take forever to render...