[derekjackson]

Hi,
just getting my head around version 2.7.1 for the first time. Seems great so far, but having a weird issue in my first render:



I've created two columns of a simple plain white wall material (transmittance colour set to black, attenuation to 1nm). These are then clad in a laser cut steel material 2.5mm thick, created using two 'metal' layers, clip mapped against a transparent vacuum layer.

Not sure what's going on, but as you can see in the image above, the white columns within the metal cladding seem to have some transparency to them - you can see light and some of the chairs through them.

I'm stumped as to where the issue is.

Have I done something wrong, or is it a bug of some sort?

[eric nixon]

I dont really know why thats happened, but there a simple fix; use the weightmap for the lasercuts in the layer not the bsdfs - theres no need for a ghost bsdf.

[Fernando Tella]

Are you using two geometries then? One for the white wall inside and one for the metal?

For the metal it would be better to mask the layer (not the BSDF, the top layer) instead of using the transmittance. It's easier to handle and less prone to mistakes.

The odd thing is that the inside white wall material is going transparent. Are the faces of the two geometries coincident?

[derekjackson]

I'm using two geometries as I was hoping to get the subtle shadows cast by the edges of the metal.

The two geometry faces (the back of the metal and the front of the white column) are coincident - that could indeed be one thing that's incorrect.

Could that be it then?

[Fernando Tella]

I had funny problems before with coincident geometries that had transparency so I would try to fix that. Also you will get a better volume and shadow effect if they are not coplanar so that's worth trying. Win-win.

[derekjackson]

Thanks for that. Photoshop'll fix it this time, but worth knowing for next time.

Interesting though, as in real life the cut metal would be applied directly to the wall.

So what would I do for, say, a graphic film applied directly to some glass?

[Fernando Tella]

Well, not coplanar is just anything that is not exactly coplanar. A small distance like 0,001 cm away would be enough to fix it. A graphic film sticked to a wall is not coplanar to it, is on top.

[dariolanza]

Hello derekjackson,

Yes, having double geometry (two columns), make the metallic one a bit bigger (1mm gap will suffice) and there apply the metal material with the mask applied directly to the Layer (not to the bsdfs, as Eric says).

The diffuse column inside don't need any mask map.

You will get the results you want, the tiny shadows at the edges, and will skip any problem due coinciding planes.

Greetings

Dario

[derekjackson]

I've just re-rendered out of interest, with the columns shrunk by 0.5mm on each side.

It's worked perfectly (and rendered with less noise as a result).

Cheers for the tip!