Hi there!

I am hanging around the newsgroups since a while. It is time for me to make
my  RSCOP contribution. Here is an attempt. It could be named "H + 1
microsecond".

   Bruno.

Post a reply to this message

Attachments:
Download 'rsocp_2.png' (551 KB)

Preview of image 'rsocp_2.png'

"Bruno Cabasson" <bru### [at] alcatelaleniaspacefr> wrote:
> Hi there!
>
> I am hanging around the newsgroups since a while. It is time for me to make
> my  RSCOP contribution. Here is an attempt. It could be named "H + 1
> microsecond".
>
>    Bruno.

Here is another render. "H + 2 microseconds"

   Bruno

Post a reply to this message

Attachments:
Download 'rsocp_2bis.png' (229 KB)

Preview of image 'rsocp_2bis.png'

Bruno Cabasson wrote:
> Hi there!
> 
> I am hanging around the newsgroups since a while. It is time for me to make
> my  RSCOP contribution. Here is an attempt. It could be named "H + 1
> microsecond".

Pretty. What are the spikes ? media or translucent cylinders?

Two points....
File size - if you'd jpeg compressed the image it could have been 30KB 
not 700+KB and you'd not have lost any noticable image quality. Some 
people feel quite strongly about this especially if they're not on 
broadband. My connections are 1Mb/sec at home and *far* faster at work 
so I'm not ppersonally bothered by download speed but the newsgroup 
server only has limited space.

Secondly my newsreader (Thunderbird) has a bug where it confuses threads 
with the same name so I'd appreciate it if people came up with more 
descriptive (and hopefully unique) titles than "RSOCP" or "IRTC WIP". I 
know it's a bug in my newsgroup client but its probably generally a nice 
idea to give more descriptive names regardless.

Anyway welcome to the group, and don't let the above two comments put 
you off (just try to keep file sizes down a tad.)

Happy POVing

Verm

Post a reply to this message

Verm <pov### [at] thirteeendynucom> wrote:
> Bruno Cabasson wrote:
> > Hi there!
> >
> > I am hanging around the newsgroups since a while. It is time for me to make
> > my  RSCOP contribution. Here is an attempt. It could be named "H + 1
> > microsecond".
>
> Pretty. What are the spikes ? media or translucent cylinders?
>
> Two points....
> File size - if you'd jpeg compressed the image it could have been 30KB
> not 700+KB and you'd not have lost any noticable image quality. Some
> people feel quite strongly about this especially if they're not on
> broadband. My connections are 1Mb/sec at home and *far* faster at work
> so I'm not ppersonally bothered by download speed but the newsgroup
> server only has limited space.
>
> Secondly my newsreader (Thunderbird) has a bug where it confuses threads
> with the same name so I'd appreciate it if people came up with more
> descriptive (and hopefully unique) titles than "RSOCP" or "IRTC WIP". I
> know it's a bug in my newsgroup client but its probably generally a nice
> idea to give more descriptive names regardless.
>
> Anyway welcome to the group, and don't let the above two comments put
> you off (just try to keep file sizes down a tad.)
>
> Happy POVing
>
> Verm

OK. Next posts will be jpeg.

The spikes are true media effect: the sphere has an internal light_source
and the light comes throught the holes of the cracked sphere, which is
sourrounded by scattering media. BUT, in order to have well-defined
'spikes' I had to increase media samples up to 200 (yes, two hundred), and
only one interval.

Here is the code for the sphere:

// ------------------------------------------------------------------
#declare THICKNESS = 0.01; // Thickness of the sphere
#declare THRESHOLD = 0.03; // Threshold for the crackle pattern

#declare CM_Incandescent = color_map
{
    [0 White]
    [0.1 Yellow]
    [0.2 Coral]
    [0.3 Orange]
    [0.4 Red]
    [0.5 VioletRed]
    [0.6 Light_Purple]
}


#declare P_Crackle = pigment {crackle scale 0.5 turbulence 0.15 omega 0.5
color_map {[0 rgb 0][THRESHOLD rgb 0][1.1*THRESHOLD rgb 1]}}
#declare fn_Crackle = function {pigment{P_Crackle}}
#declare FOUND_MAX_GRADIENT = 20;
#declare MIN_FACTOR = 0.6;
#declare O_BaseSphere = isosurface
{
    function {f_sphere (x,y,z, (1-THICKNESS)) -
THICKNESS*fn_Crackle(x,y,z).red}
    max_gradient 20
    evaluate FOUND_MAX_GRADIENT*MIN_FACTOR,
sqrt(FOUND_MAX_GRADIENT/(FOUND_MAX_GRADIENT*MIN_FACTOR)), 0.7
    contained_by {sphere {0, 0.9999}}
}


#declare O_CrackedSphere = difference
{
    object {O_BaseSphere}
    sphere {0, 1.001*(1-THICKNESS)}
}


#declare O_Media = sphere
{
    0, 1
    pigment {rgbt 1}
    hollow
    interior
    {
        media
        {
            intervals 1 samples 200
            scattering {1, White}
            density {spherical color_map {CM_Incandescent}}
            density {spherical turbulence 0.2}
            density {spherical}
         }
    }
}

#declare O_ExplodingSphere = union
{
    object {O_CrackedSphere}
    object {O_Media scale 2}
}
// ------------------------------------------------------------------

I am curently rendering another try with radiosity and partially reflective
ground, and I'll post it as jpg.

   Cheers.

Post a reply to this message

> The spikes are true media effect: the sphere has an internal light_source
> and the light comes throught the holes of the cracked sphere, which is
> sourrounded by scattering media. BUT, in order to have well-defined
> 'spikes' I had to increase media samples up to 200 (yes, two hundred), and
> only one interval.

Owch rendertimes? It took 12 hrs to render 800*600 on a 3.2 ghz machine 
then I saw I'd messed up the crackle pattern and got a seam.
Roll on the multiprocessor release of pov - although that won't help you 
with radiosity.

I'll be expecting the radiosity version of this scene in May then.

Post a reply to this message

Verm <pov### [at] thirteeendynucom> wrote:
> > The spikes are true media effect: the sphere has an internal light_source
> > and the light comes throught the holes of the cracked sphere, which is
> > sourrounded by scattering media. BUT, in order to have well-defined
> > 'spikes' I had to increase media samples up to 200 (yes, two hundred), and
> > only one interval.
>
> Owch rendertimes? It took 12 hrs to render 800*600 on a 3.2 ghz machine
> then I saw I'd messed up the crackle pattern and got a seam.
> Roll on the multiprocessor release of pov - although that won't help you
> with radiosity.
>
> I'll be expecting the radiosity version of this scene in May then.

I rendered a pure-radiosity version (only the light_source inside the
sphere) last night, and it took about 2:30 hours for 800*600 on a laptop
equipped with AMD Athlon XP3200+ (clocked 2.2 Ghz). I join here the result,
but the settings are not satisfactory yet (splotchiness), and I'll post a
new try asap.

     Bruno.

Post a reply to this message

Attachments:
Download 'rsocp_5.jpg' (88 KB)

Preview of image 'rsocp_5.jpg'

I'm impressed with your light rays effect, and how simple your code for it
is. :-) Nice.

William

Post a reply to this message

"Afishionado" <afi### [at] gmailcom> wrote:
> I'm impressed with your light rays effect, and how simple your code for it
> is. :-) Nice.
>
> William

Thanks :-) But the counter part is the render time. I rencently rendered a
800*600 radiosity version which took 2:30 hours on a laptop AMD Athlon
3200+. I'll post asap (still tuning). Depending on the lighting, it may
take half a day ...

Whatever can be my effots, I am million lightyears away from the POV-gurus
and what they can do. Thanks to them for their kindness and contributions,
and admiration for their art and skill.

    Cheers

         Bruno

Post a reply to this message

Wow! If only I had the source code, I would do some intergalactic "thing" of 
it!

Bruno, can I have the source code?

Thanks,

Sven



"Bruno Cabasson" <bru### [at] alcatelaleniaspacefr> schrieb im 
Newsbeitrag news:web.43eafcf196b04c3582fc96790@news.povray.org...
> Hi there!
>
> I am hanging around the newsgroups since a while. It is time for me to 
> make
> my  RSCOP contribution. Here is an attempt. It could be named "H + 1
> microsecond".
>
>   Bruno.
>

Post a reply to this message