water

[Ros. Riha]

what is todays state of affairs?

I need to have nice water:

On one hand, I think of caustics, on the other I would like to use SSS - but it was said these two do not go along well - what approach to employ?

 

[akinlolu olugboji]

http://www.accurender.com/photo/equin00lite-1?context=latest

[Roy Hirshkowitz]

For most architectural water features-- fountains, swimming pools, reflecting pools, etc.-- I wouldn't use SSS.  I'm not crazy about the material that akinlolu just attached.  I'd recommend a physically based material-- 100% transparency-- IOR of 1.35.  You can give it a color and use caustics (especially with the Path Tracer + sunlight.)  For pools use a 3d-face on the surface and do not tag as Thin.

[Kevin Lockwood]

There are several good ones at AccuStudio.

Most of the pool water I see in renderings has got WAY too much motion in it. Unless some fat kid just did a cannonball in the deep-end, the water would be relativity smooth. In an empty pool, like the sample above, the only thing that would move the water would be to pool filter or wind. Neither would ripple it that much. 

[Ros. Riha]

what is the reason to use only one 3D face on pools?

[Marc Chaumier]

With 3D face or region, you avoid the coincident faces issue, and as you can only see the face from above, the calcul is less heavy. If you use solid, I suppose the engine have to work with a volume and more faces; and I tried this solution, and the result was bad, even when you let a little gap between the solid of water and the faces of the pool walls.

[akinlolu olugboji]

It is not the best of all but the best so far I have come up with so far and so serves as a template for me to launch up from when next the opportunity presents itself.

[Marc Chaumier]

Hi, here are the parameters of a water material that I made:

MAIN:Reflection & sharpness to 1.0, fresnel to 0.5 (keep transparency with little angle)

TRANSPARENCY:intensity to 1.0 and IOR to1.35 or 1.5 (I use 1.5, but Roy said 1.35)... nothing else

BUMP: Here is the most difficult. After many trials, I prefer use the "sandpaper" with x, y & z at 0.15-0.3. You can increase the z, but let the x & y at the same value. And now, you can play with the "strength" to obtain more or less waves, but stay under the 1.0 at the begining.

I use a region, tagged "caustic" but not "thin". The beautiful effect come after many passes. Good dive & good luck.

[Jorge Javier Lara Domínguez]

adjunto una Muestra de una alberca con los datos que proporcionaron Marc y Roy. Saludos

[Marc Chaumier]

I work in "meters", and I think the x, y & z values are better at 0.2. The strength is in my case, closer from 0.1 or less. I think 1.0 is too much. Did you tagged your face as caustic?

[Marc Chaumier]

I will post an image to show what I obtain...

[Ros. Riha]

I use rubble instead of sandpaper. I work in mm, so my settings would be x,y=2000, z=200 and then I play with strength

[Marc Chaumier]

I used rubble, before, but the caustics are, for me, not so good. But perhaps I don't find the good settings. I don't have explored the negative values of the strength for this bump parameter. What I want to reach is this sort of caustics... see attached.