I have been trying for quite a while to produce some very realistic
procedural clouds. I can get the textures and colouring very close to
realistic but they never look quite right. I have tried mapping the textures
on to both sky spheres and planes with various sizes and distances but they
always seem too flat. Can anybody give me a little advice on how to give
them some depth.

Thx in advance

Dave

Post a reply to this message

Dave B wrote:
> 
> I have been trying for quite a while to produce some very realistic
> procedural clouds. I can get the textures and colouring very close to
> realistic but they never look quite right. I have tried mapping the textures
> on to both sky spheres and planes with various sizes and distances but they
> always seem too flat. Can anybody give me a little advice on how to give
> them some depth.
> 

For realistic volumetric clouds you could use media, but that is quite
slow.  A common method is to use a stack of planes of different height /
spheres of different radius with partly transparent textures.  Vahur
Krouverk has made some experiments with Povman shaders for this purpose
recently:

Subject: Procedural clouds
Date: Wed, 14 Mar 2001 20:57:22 +0200
From: Vahur Krouverk <vah### [at] aetecee>
Newsgroups: povray.binaries.images

Just look through the older postings in povray.binaries.images and the
scene-file groups and you will find quite a lot of possibilities.

Christoph

-- 
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other 
things on: http://www.schunter.etc.tu-bs.de/~chris/

Post a reply to this message

Christoph Hormann <chr### [at] gmxde> wrote:
: For realistic volumetric clouds you could use media, but that is quite
: slow.  A common method is to use a stack of planes of different height /
: spheres of different radius with partly transparent textures.

  My experiences show that media (method 3) is faster than using stacked
planes.
  Media (method 3) is optimized, so no too many samples are taken. With
a stack of planes povray always have to calculate rays through all the
planes, no matter if the resulting color is inexistent.

  Also media can be looked closely while the planes can't.

-- 
char*i="b[7FK@`3NB6>B:b3O6>:B:b3O6><`3:;8:6f733:>::b?7B>:>^B>C73;S1";
main(_,c,m){for(m=32;c=*i++-49;c&m?puts(""):m)for(_=(
c/4)&7;putchar(m),_--?m:(_=(1<<(c&3))-1,(m^=3)&3););}    /*- Warp -*/

Post a reply to this message

Hi Warp

Sorry to differ

>   My experiences show that media (method 3) is faster than using stacked
> planes.
>   Media (method 3) is optimized, so no too many samples are taken. With
> a stack of planes povray always have to calculate rays through all the
> planes, no matter if the resulting color is inexistent.

but in my experience method 3 (as with 1&2)leaves artifacts that sometimes
spoil the look of the clouds, to remove these it is necessary to bump up
intervals and samples which makes the rendering very slow.

Sorry but in my experience planes, especially if you use boxes instead of
planes and and scale them so they are non-infinate objects are faster.

Mick

Post a reply to this message

In article <3abcbec0$1@news.povray.org>, "Mick Hazelgrove" 
<mic### [at] mhazelgrovefsnetcouk> wrote:

> but in my experience method 3 (as with 1&2)leaves artifacts that sometimes
> spoil the look of the clouds, to remove these it is necessary to bump up
> intervals and samples which makes the rendering very slow.

The sample_spacing keyword will help fix this problem, and you could use 
transparent objects to force new intervals in the problem areas. Also, 
make sure you use a sphere, not an infinite plane.


> Sorry but in my experience planes, especially if you use boxes instead of
> planes and and scale them so they are non-infinate objects are faster.

I've always found them slower. And there isn't any kind of anti-aliasing 
that can be done to save samples in areas that don't need them, like 
media method 3 can do.

-- 
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/

<><

Post a reply to this message

"Chris Huff" <chr### [at] maccom> wrote in message
news:chrishuff-5EAB42.11131924032001@news.povray.org...
> In article <3abcbec0$1@news.povray.org>, "Mick Hazelgrove"
> <mic### [at] mhazelgrovefsnetcouk> wrote:
>
> > but in my experience method 3 (as with 1&2)leaves artifacts that
sometimes
> > spoil the look of the clouds, to remove these it is necessary to bump up
> > intervals and samples which makes the rendering very slow.
>
> The sample_spacing keyword will help fix this problem, and you could use
> transparent objects to force new intervals in the problem areas. Also,
> make sure you use a sphere, not an infinite plane.

I always use a sphere and the sample_spacing keyword!
but the idea of use transparent objects is intriguing, I shall have to
experiment with it!

Mick

Post a reply to this message

I've just run a quick set of tests

method 3 interval 1 samples 15  - 100ppm

sample spacing has no effect on the hard edges and paterns in the clouds

method 3 interval 2 samples 15 -  31ppm

hard edged patterns and artifacts dissapear

method 2

lousy colour 51 ppm

method 1 10 ppm

very spotty and broken texture


Mick

Post a reply to this message

I have tried using media but did not get very good results. I found it
difficult to stop the media touching the edge of the object it was contained
in thus producing artifacts that looked very unsightly. Any tips in this
direction would be greatly received.

Dave



"Warp" <war### [at] tagpovrayorg> wrote in message
news:3abca8b1@news.povray.org...
> Christoph Hormann <chr### [at] gmxde> wrote:
> : For realistic volumetric clouds you could use media, but that is quite
> : slow.  A common method is to use a stack of planes of different height /
> : spheres of different radius with partly transparent textures.
>
>   My experiences show that media (method 3) is faster than using stacked
> planes.
>   Media (method 3) is optimized, so no too many samples are taken. With
> a stack of planes povray always have to calculate rays through all the
> planes, no matter if the resulting color is inexistent.
>
>   Also media can be looked closely while the planes can't.
>
> --
> char*i="b[7FK@`3NB6>B:b3O6>:B:b3O6><`3:;8:6f733:>::b?7B>:>^B>C73;S1";
> main(_,c,m){for(m=32;c=*i++-49;c&m?puts(""):m)for(_=(
> c/4)&7;putchar(m),_--?m:(_=(1<<(c&3))-1,(m^=3)&3););}    /*- Warp -*/

Post a reply to this message

Hi

Try using a multiple density statement with planar if using a containing
box, i.e

        density {                  // The main cloud density statement
            normal cloud stuff
        }
        density {                  // Just to tidy it up this multiplies the
previuos density statement
          planar                   // by a slighly turbulated planar layer
to make the base and top of
          color_map {              // of the clouds less abrupt and more
wavy
            [0 rgb 0]
            [0.1 rgb 0]
            [0.1 rgb 0.9]
            [1 rgb 0.9]
          }
          scale <20,30,20>*5       // Just a bit smaller than the containing
box in y, with x and z components giving size of bumps
           turbulence 0.25
        }

See the picture "Doorway To Enlightenment" on my webpage for a better
example, the full code for the clouds is available for download in the
models section

Kev
--
http://perso.libertysurf.co.uk/kevin.ellis

Post a reply to this message

Can we have the code for this, Mick?

Post a reply to this message