Thomas de Groot <tho### [at] degrootorg> wrote:
> Some changes:
>
> 1) To the rain macro: replacing the *rand() of the rain angle by
> +RRand() with a total variation of 2 degrees max.
>
> 2) To the rain macro: adding a file-saving utility for rain drop
> parameters. The reading of such a file is much faster than complete
> processing each time.
>
> 3) To the rain macro: replaced the csg raindrop cylinder by a mesh2
> tapering cylinder.
>
> 4) 800k raindrops and slight down-scaling. The problem with down-scaling
> is that the drops become invisible, so increasing the number of drops
> does not change the final view much. A possible solution could be to
> slightly increase the size of the drops with distance from the camera
> but that might become weird. I shall investigate.
>
> 5) To the (ground) mist faking visibility loss has been added turbulence.
>
> 6) Specular and reflection added to the rock surfaces and road. /Not/ to
> the grass btw.
>
> Thomas

Getting better! If I look at the rooftop at

http://en.wikipedia.org/wiki/Rain

I get the impression, that it may suffice to have a less dense rain in the
foreground and a curtain of heavier, dense rain somewhere over your stream. Any
rain behind this curtain cannot be seen (first the curtain would hide enough of
the background and second due to Christian's argumentation) and you can spare
the effort of scaling with distance. (Also an interesting figure about the shape
and size of rain drops there ;)

Best regards,
Michael

Post a reply to this message