Christoph Hormann wrote:
> What seems very inefficient to me is
> 
>   scale .001 warp{turbulence Terrain_Turb} scale 1000
>   scale .01 warp{turbulence Terrain_Turb} scale 100
>   scale .1 warp{turbulence Terrain_Turb} scale 10
>   scale 10 warp{turbulence Terrain_Turb} scale .1
>   scale 100 warp{turbulence Terrain_Turb} scale .01
>   scale 1000 warp{turbulence Terrain_Turb} scale .001
> 
> since turbulence warps already introduce different scales of turbulence 
> (how many is influenced by 'octaves').  So by having 6 warps with a 
> scaling factor of 10 in between you overlay a lot of turbulences of 
> equal scale which is unnecessary.

   Hmmm... you got me: I didn't really understand how turbulence works. 
Do you mean that I only need 1 warp, but with higher octaves, so the 
turbulence will affect the pattern at higher scales?

--
Jaime

Post a reply to this message

In article <40df35da@news.povray.org>,
 Jaime Vives Piqueres <jai### [at] ignoranciaorg> wrote:

> Finally! Sorry all for the delay...
> 
> Note: it requires Lightsys: http://www.ignorancia.org/zips/lightsys4.zip
> 
> --
> Jaime
> ---------------------------------------------------------------------
> [Image]

  thanks sharing your code...
  just downloaded, just installed, trying to render and :

  >>
  >>File: i_land.inc  Line: 18
  >>
  >>// select pigments combination :
  >>
  >>#declare c-snow=ReflectiveSpectrum(RS_Fine_Snow) <---- ERROR
  >>
  >>Parse Error: Expected '( or [', ) found insted
  >>
 

  what i'm missing ?
  

  louis

Post a reply to this message

Kurts wrote:
>   >>#declare c-snow=ReflectiveSpectrum(RS_Fine_Snow) <---- ERROR

   I suppose you didn't paste directly, because c-snow must be c_snow ...

>   >>Parse Error: Expected '( or [', ) found insted

   That's strange... I checked on a clean POV-Ray install that it works 
"out of the box". IIRC, this error did happend to someone with the first 
3.6 betas or with MegaPOV, and I think it disappeared enclosing the 
spline identifier within a spline sentence, in this case:

   #declare c_snow=ReflectiveSpectrum(spline{RS_Fine_Snow})

   Are you using MegaPOV?

--
Jaime

Post a reply to this message

In article <cbovqn$5de$1@chho.imagico.de> , Christoph Hormann 
<chr### [at] gmxde>  wrote:

>>   I only ask you to not publically say what you are surely going to
>> think of me when reading my code... :)
>
> ROTFL
>
> I assume Thorstens experience is writing POV-SDL is not sufficient to
> actually judge much about your code (although what he thinks might still
> be not suited for public citation).

:-)  Actually, I have not even looked further than the main file...

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

In article <40dffe42$1@news.povray.org> , Jaime Vives Piqueres 
<jai### [at] ignoranciaorg>  wrote:

>> Still, in the end, I care more about optimising the isosurface code (also I
>> might as well tweak parsing speed first), so I do not "care" that much about
>> the image itself.
>
>    All the optimising I know of is "keep the max_gradient low" and "use
> simple functions"... :(

Well, there is more to be done about the rest in the POV-Ray source code for
me :-)

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

In article <40df4204@news.povray.org> , Jaime Vives Piqueres 
<jai### [at] ignoranciaorg>  wrote:

>> Great!  After rendering at a 2:1 aspect ratio I still do have one question:
>> Should the landscape and sky look like the one you posted April 29th?
>
>    Of course no! Any resemblance between images I post and the code I
> publish will be mere coincidence... :)

OK, maybe I should start picking on the "unrealistic" parts of the images
your scene generates: I think the general landscape looks really good and
sufficiently detailed (currently rendering one at 3600*2400 on my system at
work), but the clouds suffer from being colored to heavily when compared to
a "real" photograph.  Just compare the cloud colors with those used in your
scene with real clouds (taken both at about 14:00 a few days prior to summer
solstice at about 48 degrees latitude) to your scene's clouds* -
<http://mac.povray.org/new/clouds.jpg>.

As you can see, the bottom colors of the clouds are far to heavy in
particular for the given amount of sunlight.  Also, the atmosphere is a bit
too dense (and too dark at the horizon) for a good weather mountain day.  I
think you you tweak that a bit, your images if rendered at appropriate
resolutions cannot be told apart from photographs!  You could even get
prints of them...

    Thorsten

* The bottom one.  Note that all three image slices had previously been
color corrected on their own using the Photoshop "auto color" function,
which actually made the colors of your scene much more realistic, too :-)

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

Jaime Vives Piqueres wrote:
> 
>   Hmmm... you got me: I didn't really understand how turbulence works. 
> Do you mean that I only need 1 warp, but with higher octaves, so the 
> turbulence will affect the pattern at higher scales?

Not necessarily only one warp (you only have limited control over the 
intensity of the different octaves with omega and lambda) but definitely 
6 are more than necessary.  If you really want to use many warps like 
this it would be a good idea to use lower octaves count than default for 
them.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 01 May. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Thorsten Froehlich wrote:
> Well, there is more to be done about the rest in the POV-Ray source code for
> me :-)

   I hope so, because I'm developing an addiction to isosurfaces...

--
Jaime

Post a reply to this message

In article <40e05ad4$1@news.povray.org>,
 Jaime Vives Piqueres <jai### [at] ignoranciaorg> wrote:

> Kurts wrote:
> >   >>#declare c-snow=ReflectiveSpectrum(RS_Fine_Snow) <---- ERROR
> 
>    I suppose you didn't paste directly, because c-snow must be c_snow ...

  yes... :-)

> >   >>Parse Error: Expected '( or [', ) found insted
> 
>    That's strange... I checked on a clean POV-Ray install that it works 
> "out of the box". IIRC, this error did happend to someone with the first 
> 3.6 betas or with MegaPOV, and I think it disappeared enclosing the 
> spline identifier within a spline sentence, in this case:
> 
>    #declare c_snow=ReflectiveSpectrum(spline{RS_Fine_Snow})
> 
>    Are you using MegaPOV?

  yes... :-))


> --
> Jaime

  thanks for all !
  louis

Post a reply to this message

Thorsten Froehlich wrote:
> OK, maybe I should start picking on the "unrealistic" parts of the images
> your scene generates: I think the general landscape looks really good and
> sufficiently detailed (currently rendering one at 3600*2400 on my system at
> work), but the clouds suffer from being colored to heavily when compared to
> a "real" photograph.  Just compare the cloud colors with those used in your
> scene with real clouds (taken both at about 14:00 a few days prior to summer
> solstice at about 48 degrees latitude) to your scene's clouds* -
> <http://mac.povray.org/new/clouds.jpg>.

   Yes, you are right about the clouds. I think it's due to my cheap sun 
model based on altitude only, wich gives too warm colors (I have to 
revise it). For the moment. you can correct it somehow by changing the 
white point of the color system to a lower temperature.

   I added also a parameter to control clouds brightness, as increassing 
it also helps a bit to get more white clouds.

> As you can see, the bottom colors of the clouds are far to heavy in
> particular for the given amount of sunlight.  Also, the atmosphere is a bit
> too dense (and too dark at the horizon) for a good weather mountain day.  

   The "atmosphere" density depends on the turbidity of the skilight 
model, but is also controlable with the Haze_Distance parameter.

> * The bottom one.  Note that all three image slices had previously been
> color corrected on their own using the Photoshop "auto color" function,
> which actually made the colors of your scene much more realistic, too :-)

   Yes, but it doesn't works always. Sometimes it gives wierd colors, 
other it darkens the image too much, at least with Gimp auto-levels.

   Thanks for the comments!

--
Jaime

Post a reply to this message