Hi, I have only started on POV RAY for a day or so, so please go gently!
 I am trying to produce a 3D set of point light sources which the camera can
view.
Imagine that they were 10 or so stars in a Night Sky constellation. I know how
to get the
star coordinates etc etc, but have so far only used SPHERE as each object star.
The spheres size is determined by distance from camera, but I don't want
that...I want a small fixed size point of light for each coordinate star, that I
can vary the brightness of (to mimic the abs mag of the star), and also the
colour of each.

Any ideas...even if you just refer me to a tutorial !

Many Thanks
Jezza

Post a reply to this message

hi,

"Jezza" <ast### [at] gmailcom> wrote:
> Hi, I have only started on POV RAY for a day or so, so please go gently!

:-)

>  I am trying to produce a 3D set of point light sources which the camera can
> view.
> Imagine that they were 10 or so stars in a Night Sky constellation. I know how
> to get the
> star coordinates etc etc, but have so far only used SPHERE as each object star.
> The spheres size is determined by distance from camera, but I don't want
> that...I want a small fixed size point of light for each coordinate star, that I
> can vary the brightness of (to mimic the abs mag of the star), and also the
> colour of each.
>
> Any ideas...even if you just refer me to a tutorial !

varying brightness and controlling colour, you could use emissive media instead
of light sources, perhaps.

<http://wiki.povray.org/content/Reference:Media_Types#Emission>


regards, jr.

Post a reply to this message

Le 2020-04-27 à 13:34, Jezza a écrit :
> Hi, I have only started on POV RAY for a day or so, so please go gently!
>   I am trying to produce a 3D set of point light sources which the camera can
> view.
> Imagine that they were 10 or so stars in a Night Sky constellation. I know how
> to get the
> star coordinates etc etc, but have so far only used SPHERE as each object star.
> The spheres size is determined by distance from camera, but I don't want
> that...I want a small fixed size point of light for each coordinate star, that I
> can vary the brightness of (to mimic the abs mag of the star), and also the
> colour of each.
> 
> Any ideas...even if you just refer me to a tutorial !
> 
> Many Thanks
> Jezza
> 
> 
> 

First, you can never see a light source, so, that's not an option.

Place all of your «stars» at the same distance. That way, they can all 
be of the same size.
Next, it's just a mater of setting the brightness and colour.

You can use the aoi pattern. It's similar to the slope pattern, but the 
direction is relative to the direction at witch it is viewed.
Povray wiki page on that subject :
  http://wiki.povray.org/content/Reference:Aoi_Pattern

Make the spheres large enough that they cover a few pixel. Make the 
pigment black when viewed tangentially and the desired colour when 
viewed perpendicularly.
Now, if the colour get large, it will looks like more of the surface is 
coloured.

You don't need any light, just use emission with a value ranging from 0 
to 100 or even more in the finish.
Now, the pigment only need to define the actual colour and the emission 
will determine the brightness.

The final result can look similar to :
texture{
  pigment{
   aoi pigment_map{
     [0.02 Star_Colour]
     [0.5  rgb 0]
     }
   }
  }
  finish{ emission Bright diffuse 0 ambient 0}
}

Where Star_colour and Bright are previously defined values.



Alain

Post a reply to this message

"Jezza" <ast### [at] gmailcom> wrote:

I want a small fixed size point of light for each coordinate star, that I
> can vary the brightness of (to mimic the abs mag of the star), and also the
> colour of each.


Well, unlike reality, where you can have things which are very VERY bright - any
render that you make will have a color value ranging from rgb <0,0,0> to rgb
<1,1,1> - and that's it.

So I'd say define a way to convert from abs magnitude or apparent brightness to
a grayscale value, and then just vary the r, g, and b values to give the color
that you want.

The angular size of something is what you want - so you don't need to model an
actual size star at the actual distance - you can divide all of that by a
billion light-years and [visually] get the same thing.

So you can just figure out where you want things and what color you want them to
be.   If you have an rgb value of <R, G, B>, you can just multiply it by
something to vary its intensity.  So <R, G, B>*M .

So I'd say just model your scene, don't worry about the fine details, and then
adjust the color and intensity after it's made.

Post a reply to this message

On 27/04/2020 19:34, Jezza wrote:
>
 >> Any ideas...even if you just refer me to a tutorial !
> 
> 

I've just posted in p.b.i an old image that may help you. The image 
represents The Pleiades with the "real" positions of the stars, the 
magnitude and spectrum.

Tell me if this is what you want, I will share the SDL.


-- 
Kurtz le pirate
Compagnie de la Banquise

Post a reply to this message

kurtz le pirate <kur### [at] gmailcom> wrote:

> I've just posted in p.b.i an old image that may help you. The image
> represents The Pleiades with the "real" positions of the stars, the
> magnitude and spectrum.
>
> Tell me if this is what you want, I will share the SDL.


Hi Leroy,

I'm certainly interested in the underlying method for establishing the proper
positions of the stars - especially if I can couple that with other astronomical
data to show where in the sky a given star is at a particular longitude/latitude
and given date and time.

- Bill

Post a reply to this message

On 30/04/2020 19:35, Bald Eagle wrote:
> 
> I'm certainly interested in the underlying method for establishing the proper
> positions of the stars - especially if I can couple that with other astronomical
> data to show where in the sky a given star is at a particular longitude/latitude
> and given date and time.

Not in my code... My SDL just "map" right ascension and declination from 
  SAO Catalog. Datas are stored in text file with comma-separated values.


-- 
Kurtz le pirate
Compagnie de la Banquise

Post a reply to this message