[KurtS]

have a look at Jekmaxwells tests. He is "cheating", but it looks good:

http://www.maxwellrender.it/smf/index.php?topic=4419.0

[big K]

tok,

However, my question did not refer to faking techniques but to the lack of transparency the wizard water shows. Wizard water is inky, real water is transparent - even from a low viewing angle you can see shadows on the pool ground and the legs of the people (extremely refracted, of course).

this is what they are all talking about. the light that shines through the water lightens the objects within the water (or behind when talking about glass) as caustics ! this means as there is a problem with caustics from the sun (or other tiny distant lightsources) being visible through dialectrics, that everything viewed in this way will appear very dark.
have a search for caustics behind dialectrics in this forum for this very old and often sicussed topic.

michael

[Mihai]

If you check out Tims method here:

http://www.maxwellrender.com/forum/view ... 989#243989

wouldn't that give you what you want? The reason you're not seeing any light when using only glass is because all light that passes through glass becomes caustics, and it will take a lot of time for that caustic light to be visible when you want to see it behind the refractive surface.

[tok]

… this is what they are all talking about. the light that shines through the water lightens the objects within the water (or behind when talking about glass) as caustics !

AAAAAAAHHHHHH! Of course! Now I see! Got it at last. Thanks to you all!

[KurtS]

I've been running some tests, to find out if it is possible to create a pool water material that makes it possible to render without having to chose between refraction or caustics, and without applying fake caustic patterns. I have not succeeded, but I still think it could be done....

from wikipedia:
"Refraction can be seen when looking into a bowl of water. Air has a refractive index of about 1.0003, and water has a refractive index of about 1.33. If a person looks at a straight object, such as a pencil or straw, which is placed at a slant, partially in the water, the object appears to bend at the water's surface. This is due to the bending of light rays as they move from the water to the air. Once the rays reach the eye, the eye traces them back as straight lines (lines of sight). The lines of sight (shown as dashed lines) intersect at a higher position than where the actual rays originated. This causes the pencil to appear higher and the water to appear shallower than it really is."

All test renders have reached SL 15 in about 46 - 50 minutes, so there are no big differences in render time.

Normal water appears to dark, and there are no visible caustic patterns in the pool:


50%AGS and 50% water is a well known method that works fine to display very clear caustic patterns. The problem is that the missing refractions of the objects in the water shows that this is fake. The red stick is not bending like it's supposed to:


Jekmaxwells method, mixing water and AGS, 100+10 is interesting, because it is now possible to see the refraction. The problem is that it is still not very visible, and the "non-refractive" elements in the water are much more visible:


I've made a series of water tests trying make visible caustics, visible refraction and avoiding that the water becomes to dark and non-transparent. This is where I am at the moment, by no means perfect...:

[Fernando Tella]

You can get some quite good looking images with that method though you will always get the phantom shape of the non refracted objects of the ghost layer.

[Bubbaloo]

That's the closest I've seen so far. Although the two poles seen in the water is unacceptable I think. How did you achieve this method?

[KurtS]

the ghost layer is only weighted about 3%, and there are several water layers in additive mode to avoid the darkness of the "natural" water.

The water mesh/surface is modeled with displacement, and there is no material displacement in any of the tests above. Only bump.

[simmsimaging]

You are definitely within striking range of a Photoshop fix/finish at least - that's not too bad IMO.



b

[Bubbaloo]

That's what I was thinking.

[KurtS]

Maxwell Render is Physically Correct, so I never use photoshop.

(joke)

[simmsimaging]

lol

b

[Richard]

KurtS

Mate in your last test which I must say is the most realistic outcome I've seen so far, is the water a volume or just a surface???

I'm keen for a little more insight to your method to date. I'm starting a design job now for 16 individually designed dwellings on the most gorgeous slopping site (nice and large project) each dwelling will incorporate a plunge pool and I need to do some nice renders of each and the results of your last test are suffice for my thinking!

Nice work!!

[KurtS]

thanks Richard,

the water is just a surface, a modified mesh. The material works fine on any rectangle/mesh with or without displacement.
I have not tested the material under various lighting conditions or other scenes, but it works fine under normal sunlight conditions in my tests.

I'll send you the material if you will have a look at it/ use it, just pm me.