Translucent Concrete Material

[Torsten]

Hello Corona members,

I was wondering if anyone has an idea how i should setup a material to resemble translucent concrete? Check the attachement. The trick is that you can't see through the material, but the material lets light through in one direction ( perpendicular). I tried using different settings, refraction, absorption. But the key thing is that you still need to see a strong diffuse color.

[maru]

Challenge accepted! :)

[Torsten]

Challenge accepted! :)

Great news! :-)

[maru]

Ok, here are my experiments. First one is with glossy refraction, the other one with volumetric media. First one should perform well on thinner walls, the 2nd should work well on thicker ones too. It would probably require more "tuning" to look better, but I think this is the right direction. The material is applied to front and back faces of the box only.

I read that this kind of concrete is made by drowning optical fibers in it, so I tried to achieve a simplified version of what might be happening inside.

[Torsten]

Holy sh*t :-) That looks absolutely amazing Maru. Yes this effect is achieved in real life by laying nets of optical fibers and than poor concrete in between. That's why it only has light transmitting in one direction, and the material is not transparent. Your examples look a lot like what i was looking for.

Could you please save the max files to version 2014? Than i will try to advance the materials to the concrete material.

Thanks a lot!

[maru]

No problem, here they are:

[Torsten]

Hi Maru, i tested your first setup, without sss, seems the simplest to start with for me. I attached a render with the result. I like the result very much, only there are some unrealistic things happening.

The problem with this setup is that the material becomes transparent because of the Refraction value. The bright dots on the walls are there because you see a bright sky behind the wall.
But actually the whole wall should have bright dots because of the light that falls onto the walls.
So in the render you should not be able to see the building in the distance. Only if you would put an object onto the wall  or close to the wall, thus blocking the light, you would see a dark spot (check the reference i attached). Also the floor behind it should not become a dark spot on the wall, because it doesnt block the light.

Any thoughts?



Thanks for the help Maru!

[Torsten]

I attached a concept diagram on how this material works in essence. The fibers let light travel from on side to the other. Only objects that block the light, and cast shadows, become visible as darker spots on the wall.

[maru]

Yes, I understand all of this. In fact, fibers will work in Corona, but you would have to model each of them. :)

I think it might be possible to create this behavior by changing refraction IOR, refraction glossiness, or using the 2nd setup with SSS. I will try recreating similar situation. Generally interactive rendering is extremely helpful in such cases, when you tweak some of the parameters to see how they affect the object/material.

[Torsten]

Great. Yes modelling the fibers I have stopped trying some time ago, way too much fibers (1 per cm2), and way too small diameters (1mm) ;-)

[maru]

Ok, last idea, it still doesn't work like real fibers, but I can't think of anything better. :) Basically I added additional "layers" inside of the wall with the same shape of holes to block some of the light coming from different directions. The front and back surface of the wall have the same shape of holes too so when you look at it directly, you see what's behind it. This would require a lot of work to look correct, I think the easiest way would be to use bitmaps instead of procedural maps and apply box uvw mapping. The wall is refractive inside with a little bit of glossiness. Changing any parameter slightly, like glossiness or noise map settings changes A LOT. Render times are pretty horrible (no wonder - lots of small holes).

[agentdark45]

Had a quick go at this, I found adding a Coronafrontback map in the refraction slot (jpeg in front face side, white in back face side) gave a fairly close result, at least when compared to just sticking a map in the refraction slot.

[Torsten]

Thanks for your help guys. The technique i use at the moment is usig a cellulor map to control the opacity of the wall. This noisemap comes very close to how and where the actual light dots would be seen. I still need to make sure the material is not see through. When i have some renders i will update the post.

[Nekrobul]

Omg it works.

Maru you are practising some black woodoo arent you?

[maru]

Maru you are practising some black woodoo arent you?

You have unmasked the true meaning of "rendering is magic" line.