simple water material

psor · Post by **psor** » Fri Sep 11, 2009 2:31 pm

@Borgleader

As far as I can recall, the issue was caused by intersection of two or more objects.

edit: I'd say your suzane is not that solid. What I mean is, on the suzane model
are some vertices that have to be welded. Take a look at the one out of glass ...
Especially at the eyes you can see what I mean.

take care
psor

Borgleader · Post by **Borgleader** » Fri Sep 11, 2009 2:39 pm

Ok by solid suzanne i meant opaque. The issue with the eyes is that theyre seperate from the head model and the vertices are probably closer to each other than ray nudge distance, but thats not my point.

He was talking about a beach with rocks, and my point was instead of modeling the water close to the rocks & sea bottom you could just shove it inside the water mesh.

psor · Post by **psor** » Fri Sep 11, 2009 2:52 pm

Hehe, ... and my point is, that you're not aloud to have another object that is intersecting
a dielectric. Water, glass, ... etc. So you see the issue. Imagine a floating boat on the
ocean. That's very much ... uhm - bollox!

I guess it's even an issue then to, lets say drop an olive into a martini for a nice
animation with blender fluids. What a mess!

take care
psor

Borgleader · Post by **Borgleader** » Fri Sep 11, 2009 11:09 pm

Did I break any rules?

In this scene, the water mesh goes through suzanne, into the glass mesh, into the metal mesh. The glass mesh goes through the metal mesh as well.

psor · Post by **psor** » Sat Sep 12, 2009 4:33 am

Good question. I asked Ono about it per PM yesterday. Because I would like to
know if the restriction is still there or not. It seems to be nobody really knows.

*LOOOOL*

nb: Me too started a render for testing yesterday.

take care
psor

Borgleader · Post by **Borgleader** » Sat Sep 12, 2009 9:36 am

To be honest I always did things this way because I thought the precendence value was specifically to decide which mesh (transparent that is) take "more importance" than the other if two different meshes occupy the same space.

EDIT: Also I believe that "air" (aka non-mesh enclosed space) has a precedence of 0 (lowest value allowed by exporter -1) so that it cant by material creating mistake take the place of a mesh.

psor · Post by **psor** » Mon Sep 14, 2009 12:08 pm

Ono?! Please give us a heads up about the restrictions with shaders that include a medium.

take care
psor

Borgleader · Post by **Borgleader** » Mon Sep 14, 2009 1:24 pm

Haha yeah it'd be nice to have a definitive answer from "God"

psor · Post by **psor** » Mon Sep 14, 2009 1:27 pm

Pssst, ... not to loud. He could hear you.

take care
psor

Borgleader · Post by **Borgleader** » Wed Sep 16, 2009 11:56 am

Still no news on the issue?

psor · Post by **psor** » Wed Sep 16, 2009 12:11 pm

Nope. And since I'm busy with other stuff I've not that much time for testing. On the other
hand I'm still investigating in another issue. Will post a bug report soon. I take an educated
guess that the issue of intersection is still there and it will cost rendertime.

I'm preparing a tutorial about "reflectance" where I show clearly that reflectance over 80%
will cost you rendertime. Especially when you hit 100% it will cost you so much rendertime
that you'll think of buying a new PC soon. Fortunately in nature there exists almost nothing
that goes over 88%. To make it clear, reflectance of 88% is a glossy white paper for a printer.
Like I said, I'm still busy but will keep you posted soon enough.

edit: I've added a graph just to give you a short preview.

take care
psor

Post by **OnoSendai** » Wed Sep 16, 2009 12:30 pm

Very interesting graph, thanks Psor.
Btw you could probably derive an analytical expression for the render speed given the reflectance, with a few assumptions about the scene.

psor · Post by **psor** » Wed Sep 16, 2009 12:36 pm

You're very welcome Nik. I'll need some time to prepare everything cause of lack of
time and PC power. I've to do everything on the old box of my girl cause my box is
gone caused by electricity. So mobo and stuff is history. Fortunately my drives are
all ok, hehe. I'll keep you updated.

take care
psor

Borgleader · Post by **Borgleader** » Thu Sep 24, 2009 1:25 pm

And....what's the final verdict on the issue? I don't recall every getting a "straight" answer from the higher ups

psor · Post by **psor** » Thu Sep 24, 2009 1:37 pm

@Borgleader

It's all a big secret and never has to be revealed till Maxwell Render 3.0 will show up.

Naaah, I had a small chat with Nik and he said that everything should be ok and the
only NoNo would be intersection of emitters and a medium. So a mediums boundary
should never be intersected from an emitter. Everything else should be cool.

Took some time to get the answer tho.

So let the intersections start!

take care
psor

simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Re: simple water material

Who is online