Maxwell Render | Official public forum

Wondering if you guys can help me out with what is going on with this. every bottle I make, I run into this issue on the base where the refracted light seems to get trapped in the glass and turns out black. No matter how i light, i am not able to get rid of this blackness. Is it a maxwell bug, when dealing with this 4 surface light interaction?

Here's a couple samples.

I am familiar with that problem. Perhaps related to the issues I was having figuring out caustics for this kind of stuff:

http://www.maxwellrender.com/forum/view ... 97&t=36754

I never did get any input from NL on it and kinda gave up and moved on, but I'd like to know the answer for sure.

/b

It is reflected caustics. Maxwell is limited in that area. You may need to resort to trickery.

oh man that sucks really bad.

what exactly is a reflected caustic that is causing the issue? Is this double refraction from one surface that isn't working correctly? do I need to split up these surfaces so the base is fully a different piece so that it isn't trying to refract in and out of the same surface multiple times?

The problem is that the light you would see under there would be all caustic light since it first traveled through glass (twice!) before it reaches the camera. Here is the path that the ray travels:

Light from source travels through glass entry normal -> travels through glass medium -> exits through glass normal -> bounces off of floor surface -> travels through glass entry normal -> travels through glass medium -> travels through glass exit normal -> to camera. That's some heavy calculations going on there. Theoretically, Maxwell will eventually converge to a physically accurate result, but it will be a looong time. If you render at very small resolution for a while, you will start to see bright dots in that area. Those dots are the rays starting to make it back to the camera.

ok, that makes sense. do you have any thought about a workaround for this? nothing is coming to mind right now.

if you remove any plane below the bottle, shouldn't that theoretically brighten up that base? that way there are no caustics being reflected back through towards the camera...you only have light traveling one way through the base.

it doesn't fix it, but based on what your saying the issue is, it seems like it should?

For what it's worth, I did a low res render of the base of the glass bottle and it got to SL 27 and showed no sign of light coming through. not even any noise sparkles.

Does Maxwell just completely stop calculating after a certain amount of refractions?

You would have to wait much longer than that -- see this thread:
http://www.maxwellrender.com/forum/view ... 25#p340525

The thicker the glass (and more attenuation) the longer it will take to see much -- SL 31-32 you should see something looking like a caustic pattern beginning to break through.

Best,
Jason.

S.L. 33

itsallgoode9 wrote:Does Maxwell just completely stop calculating after a certain amount of refractions?

I'm not sure. That's a question for Tom or Juan.

Bubbaloo wrote:

itsallgoode9 wrote:Does Maxwell just completely stop calculating after a certain amount of refractions?

I'm not sure. That's a question for Tom or Juan.

BTW with something like this the best choice is to use render region to resolve that little bit at the bottom... but still even with a smaller image you are looking at a couple of days with a standard computer -- render farms for the render region would be the most practical solution (timewise).

Best,
Jason.

well this is all quite discouraging

It is, I just cant render realistic looking fluids (which is my main profession) because of this "limitation".

Water just doesn't look right not getting illuminated from the caustics it spreads on any surface or collision objects..

I really doubt NL gets around that though.. so far no one did.. it lies in the nature of unbiased engines, I'm afraid.

Maxwell is still the best renderer out there though. The optimized sun caustics have been a tricky workaround though. Maybe NL comes up with something similiar..

The need to render to some SL beyond 30 to start to see things clear up is no option at all..

I think I remember some thread with glass/water drops, that generated the caustic halo beneath the surface..

just hoping for the soon to be released update, since the end of June actually

cheers

Aniki