My curiosity was peaked after I asked about the possibility of using it as a
starfield.

Post a reply to this message

Attachments:
Download 'noisey2.JPG' (43 KB)

Preview of image 'noisey2.JPG'

V. interesting, especially since I was noodling around in the same area
looking at histograms of f_noise3d samples. I was actually concentrating
more on noise_generator(s) 2 and 3. The original noise (noise_generator 1)
is really significantly different in its distribution curve from the other
two!

I'd like to know more about how you chose and collected your sample data.

Abe

Greg M. Johnson wrote:
>My curiosity was peaked after I asked about the possibility of using it as a
>starfield.
>

Post a reply to this message

"Abe" <bul### [at] taconicnet> wrote in message
news:web.3d6fcf10c3cf9d4c6895cad0@news.povray.org...
>
> I'd like to know more about how you chose and collected your sample data.
>

The scene file uses standard "metallography"  theory and has a programmable
precision.  Those values were set to 0.5%.

#version 3.5;
#include "colors.inc"
#include "functions.inc"

#debug  "\n"

//-------------------------------------------
global_settings {
 assumed_gamma 1
 noise_generator 1



}

#declare CamLoc = < 0,16,0>;

camera {
 location CamLoc
 look_at < 0, 0, 0>
 angle 70
}

light_source {<50,335,0> color White}
light_source {<-50, 150,-75> color White*0.35}


#declare IsoFinish =
finish {
 ambient 0 diffuse 1
 specular 1 roughness 0.02
 brilliance 2
}

sphere{0,1 pigment{red 1}}


//fog {color rgb -.3*<1-0.129,1-0.373,1-0.518> distance 200}
//fog {color rgb <0.329,0.620,0.706> distance 800}

#declare seasc=.5;

#declare threshy=0;
#while(threshy<1.01)


#declare hills=
isosurface {
 function {f_noise3d (x/seasc, y/seasc, z/seasc)}
 threshold threshy
 max_gradient 15.0000
 contained_by{ box{<-1300,-0.060,-1300>,<1300,0.060,1300>}}
 //contained_by{ sphere{0,1600}}
        accuracy      0.1  //0.001 was default
 //open
 pigment{rgb <.6,.6,.2>}
        clipped_by { bounded_by }
}

#declare deviation=0.005;

#declare Oldvalue=2;
#declare Newvalue=4;
#declare powwer=0;

#declare tries=0;
#while(tries<100)

#declare nxmax= int(10*pow(1.3,powwer));
#declare nzmax= int(10*pow(1.3,powwer));

#declare Norm=<0,0,0>;
#declare counts=0;

#declare nx=0;
#while(nx<nxmax)

#declare nz=0;
#while(nz<nzmax)

#declare exxzee=<-10+20*nx/(nxmax-1),0,-10+20*nz/(nzmax-1)>;
#declare tracey=trace(hills,exxzee+10*y,-y,Norm);


#if (vlength(Norm)!=0)
        #declare  counts=counts+1;

#end


#declare Norm=<0,0,0>;

#declare nz=nz+1;
#end

#declare nx=nx+1;
#end


#declare areafraction=counts/nxmax/nzmax;



#declare varyance=abs(areafraction-Oldvalue)/(Oldvalue+0.0000001);
#if(varyance<deviation)
#declare tries=tries+200;
#end

#declare Oldvalue=areafraction;
#declare powwer=powwer+1;

#declare tries=tries+1;
#end


//#debug "Threshold "
#debug  str(threshy,4,3)
//#debug "    volume fraction "
#debug ","
#debug  str(areafraction,6,5)
#debug  "\n"

#declare threshy=threshy+0.05;
#end


object{hills}

// speedup steps :
//  place function inside function{}
//  smallmaxgradients <1

//open
//higher accuracy

//        clipped_by { bounded_by }  only works at end

Post a reply to this message

It's nice to see a graphical representation of those differences. I didn't
think 2 and 3 would be so close. Thanks

 - Slime
[ http://www.slimeland.com/ ]

Post a reply to this message