Maxwell Render | Official public forum

Because Maxwell treats each normal as an entry/exit surface regardless of the direction of the normal. Hence, no invisible faces of opaque objects' back faces.

Using the technique from the diagram that polynurb posted, here's my rendering of an ice cube inside the liquid that is inside a glass:



I modelled each element as a solid which means I had no choice in the surface normal direction. All my surface normals face away from the inner volume of the object.


Andrew.

The problem with this method comes when glass or liquid have different roughness, doesn't it?; frosty glass with water, for example.

Rogurt wrote:Hi all
It´s a pitty that the images of this topic are not accessible any more since the liquid in glass question is a pretty common one and quite important.
Unfortunately pictures cannot be uploaded here in the forum, right?

Indeed, most of my staff is gone. The Comcast account where all these images and files were hosted got deleted after moving to a different city with no Comcast coverage.
I put back a couple of the links.

The problem with this method comes when glass or liquid have different roughness, doesn't it?; frosty glass with water, for example

This isnt a problem if you use 2 mxm's for the glass applied to polyselections, mxm's will have to be derived from each other, i.e. a cloned mxm with different roughness.

@ Andrew, render looks fine, I guess you already know that .

This is an area where I really think Maxwell needs some strategic improvement. I'm sure it is non-trivial but having interpenetrating transparent materials is really the only good way to go in the long run. This method works fine when the shapes are uber simple like Andrew's example, but good luck with really complex shapes.

simmsimaging wrote:This is an area where I really think Maxwell needs some strategic improvement. I'm sure it is non-trivial but having interpenetrating transparent materials is really the only good way to go in the long run. This method works fine when the shapes are uber simple like Andrew's example, but good luck with really complex shapes.

Agreed ! It is too much work right now.

the "layered" refraction on the side of the first glass looks a bit wrong to me. The others are good.

eric nixon wrote:

The problem with this method comes when glass or liquid have different roughness, doesn't it?; frosty glass with water, for example

This isnt a problem if you use 2 mxm's for the glass applied to polyselections, mxm's will have to be derived from each other, i.e. a cloned mxm with different roughness.

True, and it's even easier in a plugin like SW which has "Multi-materials"

Still is a bit irritating to have to deal with poly selections, in C4D it's particularly annoying, though JD has been working on it diligently.

Really it's not a geometrically sound solution, too bad it's the best approach at the moment, but oh well.

I'd say it also depends how familiar one is with uv unwrapping, as a roughness map could do the same.
Unfortuneatly from within rhino that is the only choice.(no multimats)

Fernando Tella wrote:The problem with this method comes when glass or liquid have different roughness, doesn't it?; frosty glass with water, for example.

Not really a problem. If I remember correctly the surface of the glass over the liquid should have the same roughness as the liquid itself (thus not frosted) but keep the IOR of the frosted glass.

http://www.youtube.com/watch?v=OZxtwIc3ieM from 3:20 .
don't know is it usefull but still looks amazing

zdeno wrote:http://www.youtube.com/watch?v=OZxtwIc3ieM from 3:20 .
don't know is it usefull but still looks amazing

Nice one zdeno!
Seems easier to do in real life than with Maxwell Render...
Cheers!