|
 |
Hi(gh)!
Apart from mesh2s at close view generally looking angular and thus
non-realistic, I probably also won't get rid of the notorious gaps
between the triangles if I stick to mesh2s.
So I started to look up Christoph Hormann's earth_simple_free.pov
(http://www.imagico.de/pov/earth_scene.php), well aware that this script
is intended for global renderings at a small scale rather than
high-scale close ups like my Tristan da Cunha scenes.
In earth_simple_free.pov, Christoph turns an image_map mapped around a
sphere into a pigment function, which then is modified further:
#local fn_DEM=
function {
pigment {
image_map {
ImageFile_Auto(IMG_Topo)
map_type 0
interpolate 3
once
}
}
}
#local fn_DEM_Height=function { (fn_DEM(x, y, 0).red - 0.5)*2 * 32.767 }
#local fn_Shape=function { f_sphere(x, y, z, Earth_Radius) }
#local fn_Pattern=
function {
fn_DEM_Height(1-(f_th(x,z,y)+pi)/(2*pi), f_ph(x,-z,y)/pi, 0)
}
#local fn_Iso=
function {
fn_Shape(x,y,z)-
fn_Pattern(x,y,z)*Height_Exaggerate
}
isosurface{
function { fn_Iso(x,y,z) }
max_gradient 1.2
accuracy 0.001
contained_by{ sphere{ <0,0,0>
Earth_Radius+Max_Mountain*Height_Exaggerate } }
texture {
pigment {
Pig_Relief
}
finish {
ambient Terrain_Ambient
diffuse 0.5*Terrain_Brightness
}
}
hollow on
}
// end of code
But as ASTER DEMs are 3601 by 3601 data points per square degree and
probably noone of us is able to handle one single giant DEM of the
entire Earth at this resolution, I now would like to know whether it is
possible to constrain the DEM (with my script: heightfield) pigment
function to a certain range of geographical coordinate (with the Tristan
See you in Khyberspace!
Yadgar
Post a reply to this message
|
|
|
|
> But as ASTER DEMs are 3601 by 3601 data points per square degree and
> probably noone of us is able to handle one single giant DEM of the
> entire Earth at this resolution, I now would like to know whether it is
> possible to constrain the DEM (with my script: heightfield) pigment
> function to a certain range of geographical coordinate (with the Tristan
Just a guess, and I don't know how to do it, but the contained_by{} code
of the isosurface seems indicated. Probably a box with the geographical
coordinates transposed into xz coordinates and y determined by your
intended height?
Thomas
Post a reply to this message
|
|
|
|
Hi(gh)!
On 04.07.2013 10:38, Thomas de Groot wrote:
> Just a guess, and I don't know how to do it, but the contained_by{} code
> of the isosurface seems indicated. Probably a box with the geographical
> coordinates transposed into xz coordinates and y determined by your
> intended height?
Probably not, as the heightfield image is projected onto an entire
sphere already in the very first function... this would only work if I
used a GIANT heightfield (and by "giant", I mean 3600 x 360 by 3600 x
180 pixels, which would add up to about 839 GB...)!
See you in Khyberspace!
Yadgar
Post a reply to this message
|
|
|
|
Hi(gh)!
> which would add up to about 839 GB...)!
...if it was 8-bit, but it is 24-bit, so it would be even about 2.5 TB!
What PC can handle several terabytes of RAM? Perhaps around 2030...
See you in Khyberspace!
Yadgar
Now playing: Psychopoly (Zara-Thustra)
Post a reply to this message
|
|
