[Ernesto]

Finnally I could get the expected result, but avoiding bump mapping!
If you create a model of the water surface (Too much polygons) and assign smooth property several times and just in case press recalculate normals, the reflections and refractions will be continuous as in real life.

Unfortunately Maxwell cannot handle bump maps as other softwares do.
I have tested all the water bump maps used in this tests, on other software, and I have got great results, but Maxwell seems to read each pixel in the bump map as a separated polygon which is a perfectly horizontal plane.
In this way the whole thing will work as a bunch of steps, instead of a curved surface.

This problem explains the weird behaviour that are seen in the other Bump topic in this forum.
There it is seen that a blured bump map gets better results.
http://www.maxwellrender.com/forum/view ... hp?t=17632
This happens because Maxwell do not guarantee the continuity of the relief, taken from a bump map. Instead it reads it as independent pixels.
In the blured bump map, the map itself forces Maxwell to show a better continuity, but it is just an illusion produced by a reduced height between steps. biut they are still steps, between pixels.

I do not know if this behaviour has better results in other uses, but I hope NL will understand this as a bug, to be fixed, so that we could use Bump mapping for water again.

Ernesto

[Ernesto]

Here you have an image that should be similar to the result of using a bump map. The diference between this image and the rest using bump maps in maxwell 1.1 is that here you can see curve refractions and reflections.


The refracted joint lines goes up and down, without a break.

To get this effect I had to model the water surface, which is a waste in case you need to render heavy models. Bump effect should let us to render the same effect, with just two polygons.

In case of Bump Maps this is impossible in current Maxwell render version.
It only broke the lines into short segments, that are perfectly paralel to the bottom joints, which is not real.
the only case in which the refractions of the pool bottom could be paralel to the real joints, is when the water surface is perfectly horizontal without any wave. In case you have doubts about this BUG, download the posted model, and try to render an image which can show unbroken refractions and reflections lines. It will be impossible.

Another problem at least in Maxwell Studio is that the smoothness angle for meshes, has a limit of 90 degrees. It would be necesary to increase this limit at least up to 135 degrees to get full advantage of the feature, specially when reflections and refractions are involved.

Ernesto