|
|
I am really just guessing here but when we antialias, we are trying to
smooth edges and remove pixelation. Normally this is done by taking a
desired pixel and the pixels that immediately surround it and then
averaging.
If a star is a single point (pixel) in space surrounded by blackness, then
if we antialias, we will average the single white point by many more black
points reducing the overall value (brightness) of the white point.
One possible (but untested) solution would be to not use "Starfield1" but
instead copy in the texture definition that the Starfield1 symbol represents
and then multiply each of the color values by some multiplying constant.
For example:
#declare My_Starfield1 =
texture {
pigment {
granite
color_map {
[ 0.000 0.270 color rgb < 0, 0, 0>*10 color rgb < 0, 0, 0>*10 ]
[ 0.270 0.280 color rgb <.5,.5,.4>*10 color rgb <.8,.8,.4>*10 ]
[ 0.280 0.470 color rgb < 0, 0, 0>*10 color rgb < 0, 0, 0>*10 ]
[ 0.470 0.480 color rgb <.4,.4,.5>*10 color rgb <.4,.4,.8>*10 ]
[ 0.480 0.680 color rgb < 0, 0, 0>*10 color rgb < 0, 0, 0>*10 ]
[ 0.680 0.690 color rgb <.5,.4,.4>*10 color rgb <.8,.4,.4>*10 ]
[ 0.690 0.880 color rgb < 0, 0, 0>*10 color rgb < 0, 0, 0>*10 ]
[ 0.880 0.890 color rgb <.5,.5,.5>*10 color rgb < 1, 1, 1>*10 ]
[ 0.890 1.000 color rgb < 0, 0, 0>*10 color rgb < 0, 0, 0>*10 ]
}
turbulence 1
sine_wave
scale .5
}
finish { diffuse 0 ambient 1 }
}
Again, not tested ... just a theory thought ...
Neil
"Barehunter" <nomail@nomail> wrote in message
news:web.41db32f66a5cceb5123f95160@news.povray.org...
> I did a picture using a sphere with starfield1. when I made the
antialiased
> version the stars disappeard and the sky was solid black. The non aliased
> pictures showed the stars just fine.What could have gone wrong? Thank you.
>
> I used this
> sphere{<0,0,0>,1000
> hollow
> texture{Starfield1}
> }
>
>
>
Post a reply to this message
|
|
|
|
In article <41db47bc$1@news.povray.org>,
"Neil Kolban" <kol### [at] kolbancom> wrote:
> If a star is a single point (pixel) in space surrounded by blackness, then
> if we antialias, we will average the single white point by many more black
> points reducing the overall value (brightness) of the white point.
That's essentially the problem. Those textures rely on aliasing to
create the stars...each pixel is a point sample of a mostly black
pigment. Occasionally one hits a light area, producing a star.
Antialiasing just renders the image more correctly, showing that there
really aren't any stars. It gets worse when you clip colors before
averaging, since even increasing the brightness won't increase the
effect on a pixel. (However, not clipping produces jagged edges on very
bright objects. A real solution will be more complex...)
Simply brightening the stars is unlikely to give good results, it will
probably just reveal the pattern. Pretty much the only way I know of to
get decent stars with AA is to use a big shell of many randomly placed
spheres sized to cover approximately one pixel.
--
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: <chr### [at] tagpovrayorg>
http://tag.povray.org/
Post a reply to this message
|
|