Please forgive the noob nature of this post, but I'd really like to
understand what is going on with HDRI-lit renderings.  In playing around
toward that end I took the standard MLPOV hdri.pov scene and put a disc in
like this:

disc {
  <0, 0, 0> y, 8.0, 0
  pigment { color rgb 1 }
  finish { ambient 0 diffuse 1 }
}

the resulting rendering is attached below as HDRI_disc.jpg  I wanted to have
more defined shadows, and so asked Alan Holding about the settings he used
in his very fine skeleton.jpg image ( located here:
http://news.povray.org/povray.binaries.images/29668/ )

I lifted his globals and stuck them into the HDRI_disc scene; here they are:

global_settings {
  radiosity {
    pretrace_start 0.16
    pretrace_end   0.01
    count 300
    nearest_count 3
    error_bound 0.1
    recursion_limit 1
    low_error_factor 1
    gray_threshold 0.0
    minimum_reuse 0.015
    brightness 1  // was 1.2 in Alan's scene

    adc_bailout 0.01/2
    normal on
  }
}

and the result is attached below as HDRI_disc_lichens.jpg  Strange dark
spots...photons?

Finally, by changing the "recursion limit" from 1 to 3 bright spots that
look an awful lot like refractive caustics appear.  This one is attached as
HDRI_lichens_caustics.jpg.

The Questions:
1) Based on the orignal HDRI_disc.pov, what I settings do I need to change
just to get smoother, more believable shadows?  Setting low_error_factor
lower smooths the shadows, but doesn't touch the spots...
2) What *are* those dark spots?  How do I make them either go away or
resolve into something more attractive?
3) Are those really caustics?  How can I eliminate them?  How can I refine
them?
4) The contrast in the upper right of the middle sphere is exceptionally low
in the two later images--any idea why?

Thanks for your patience--RTFM responses completely welcome, especially if
they come with section numbers!

-Chad

Post a reply to this message

Attachments:
Download 'HDRI_disc_lichens_caustics.jpg' (36 KB) Download 'HDRI_disc_lichens.jpg' (32 KB) Download 'HDRI_disc.jpg' (24 KB)

Preview of image 'HDRI_disc_lichens_caustics.jpg'

Preview of image 'HDRI_disc_lichens.jpg'

Preview of image 'HDRI_disc.jpg'

"Chad" <ceb### [at] druryedu> wrote in message news:3e259c29@news.povray.org...
> 1) Based on the orignal HDRI_disc.pov, what I settings do I need to change
> just to get smoother, more believable shadows?

count 300 could be higher probably. If you raise error_bound toward 1.0
things will blend out too but then details are lost. At least that's all I
have to say on this.

--
Farewell,
Bob

Post a reply to this message

>4) The contrast in the upper right of the middle sphere is exceptionally low
>in the two later images--any idea why?

Try changing the sphere's diffuse value to 1, or make the pigment brighter
than 1.  It seems that sometimes radiosity will not brighten a surface past
diffuse*pigment, no matter how bright the surroundings.  Increasing
max_sample doesn't seem to help either.  Is this a bug?

Post a reply to this message

Chad wrote:
> 2) What *are* those dark spots?  How do I make them either go away or
> resolve into something more attractive?

This is a max_trace_level problem. If you put "max_trace_level 50" in 
the global_settings block they should go away.

I'd also increase the counts somewhat and the nearest_count to at least 5.

Bye for now,
	Mike Andrews.

Post a reply to this message