[numerobis]

what would be the best solution for an indoor pool with pool lights, down lights and windows? (quality+speed)

- water volume hidden to GI + fake caustic plane (hidden to gi+camera) + ags plane?!?

@tom:
could you tell me which setup you used in this scene? : http://www.maxwellrender.com/forum/view ... &start=555

[numerobis]

[zdeno]

[numerobis]

thanks...

[Bubbaloo]

That's a tough situation for Maxwell. Personally, I would experiment with rendering out some caustics patterns and using that as a map applied to the bottom of the pool. Haven't had to do it, though so I have no experience with it.

[Fernando Tella]

I use a plane with real water hidden to GI + a plane hidden to camera+reflections+refractions masked with a caustics-like pattern, so it lets all the light through the lines of the pattern and blocks a bit of light in the rest of the map; if you make this last material a bit reflective you also get real reflected caustics. As seen in the last image of this post: http://www.maxwellrender.com/forum/view ... 25&t=33878
So with this solution you get real reflected caustics and faked refracted caustics above and under the water, the under water part gets refraction and all cleans very fast.

Edit: I'm not sure now if reflected caustics show under water (from lights inside the pool)

[numerobis]

thank you bubba and fernando!
i've used the "two plane method" you described fernando. it's working ok, but i think it could be improved for the pool lights.
i'll do some more research next time...

i thought there where another solution posted by tom in the other thread - but this one seems to be secret...

[numerobis]

indoor pool once again and still no real solution on my side...

any chance to get an insight how you did this Tom?
Or was it some special secret test version of maxwell 4...?!?

You've meant this? (a 25 min ML render)

No fakes here and all are real caustics, 1 pass, no post.

Physical Sky


Room Lighting


Room + Pool Lighting


Pool-only Lighting

( from this thread: http://www.maxwellrender.com/forum/view ... &start=555 )

.

[Bubbaloo]

I think his secret was low resolution + high S.L...

[eric nixon]

This works, no tricks, but if you use sss for the water then you cant embed the emitters anymore.




For underwater shots you need to surround the camera with an 'air-bubble' which is a child of the water mesh. If you dont use a bubble you get the effect of how things look underwater when you have no goggles on I also use a flat polygon for the camera to look out of, this prevents internal reflections, just like real goggles..

My internet is not uploading at the moment, I'll try and post the render later.

[numerobis]

I think his secret was low resolution + high S.L...

This works, no tricks

really?!?
so you mean this is a normal water material on a water volume and 25min for 600x300 px for real caustics?

[eric nixon]

Putting an emitter inside a dielectric with roughness 0 is certainly fast, as is lambert glass, other roughnesses may be a bit slower.

Basically as long as the emitter rays are only passing through one layer of transparent geo theres not much slow-down.

[Bubbaloo]

I think this would work the same, no?

37pool.jpg

[eric nixon]

Yes, why not? or even make the ground shape solid, personally I'd make everything solid, always safest.

[numerobis]

not working for me... no caustics here - not after 25min and not after 2 hours

i think i'll have to use vray...