This is the 1st time I've tried a pure radiosity illuminated scene. I 
also wanted to demo the most recent round of changes to RC4 ... ie: 
subsurface + radiosity. The important settings are listed below ... Oh 
and the test object I found at: http://shapes.aim-at-shape.net/ it was 
in Stanford triangle format converted to mesh2.

First off I think I got darn close on the radiosity settings for a first 
attempt. I think the obvious question here is what's going on where the 
sitting figure meets the base. Is that the base pigment in it's 
material, showing through the subsurface effect? If so I'd expect it has 
more to do with my choices for the subsurface globals ... mm_per_unit? 
When dialing it in I noticed that starting at about 75% of the final 
setting (12.7) the material started to almost become see through, after 
about 250% greater the subsurface effect became non existent ... that's 
my range, correct? I used a conventional light source while doing the 
subsurface tuning. The settings I settled for left a slight over hint of 
the effect in the shadow of the right-sided arm and the same with the 
shadow the leg cast onto the base.

Lastly ... yikes a render time of 54 hours 31 minutes. I was doing other 
stuff (@#$%! tax time) so definitely not a fair performance indicator, 
but wow won't be doing this type of render any time soon. I'm wondering 
... is 2 pass radiosity still an option with this incantation of 
subsurface? What about the other radiosity time savers like 
Radiosity_Vain_Pretrace ... the last paragraph of it's doc entry reads 
like probably not:

"At the moment, turning vain pretrace off will affect only classic 
lighting computations (diffuse lighting, highlights and iridescence); 
other features expendable during pretrace may follow in future versions."

Subsurface globals:
mm_per_unit 12.7
subsurface {
   radiosity on
   samples 256,128
   }

Subsurface material: unglazed bone china, like some Wedgwood or Doulton 
figurines
material {
   texture {
     pigment { rgb <0.9600,0.8000,0.6900> }
     finish {
     subsurface { translucency rgb <0.4350,0.2580,0.2580> }
     diffuse 0.5
     ambient 0
     emission 0
     specular 0.1
     roughness 1e-2
     reflection {0.01}
     }
   }
   interior {ior 1.486}  // coral
	// chalk 1.510
}

Radiosity globals:
   subsurface on
   pretrace_start 0.08
   pretrace_end   0.008
   nearest_count 15
   count 150
   gray_threshold 0
   error_bound 0.5
   recursion_limit 3
   low_error_factor 0.5
   minimum_reuse 0.005
   maximum_reuse 0.05

Post a reply to this message

Attachments:
Download 'fertility.png' (598 KB)

Preview of image 'fertility.png'

>Jim Holsenback  on date 25/03/2011 03:01 wrote:
[...]
> Subsurface material: unglazed bone china, like some Wedgwood or Doulton
[...]
Very nice, could be interesting even in kenya soapstone material.
Paolo

Post a reply to this message

On 03/25/2011 06:23 AM, Paolo Gibellini wrote:
>  >Jim Holsenback on date 25/03/2011 03:01 wrote:
> [...]
>> Subsurface material: unglazed bone china, like some Wedgwood or Doulton
> [...]
> Very nice, could be interesting even in kenya soapstone material.
> Paolo

Good suggestion ... since this version was basically a test, I wanted 
wanted to stay pretty basic. Being that the object is African maybe 
ivory would be another option.

I like the object and hope I can come up with something that does it 
justice ... back to the drawing board ;-)

Post a reply to this message

On 03/24/2011 11:01 PM, Jim Holsenback wrote:
> Lastly ... yikes a render time of 54 hours 31 minutes.

OK ... the render time was really bugging me, so I did a little 
postmortem and I've uncovered an epic fail on my part (a real news flash 
... eh) ... what a goober :-(

Post a reply to this message

On 25/03/2011 2:01 AM, Jim Holsenback wrote:
> First off I think I got darn close on the radiosity settings for a first
> attempt.

I think you did. Very nice, I missed that model it reminds me of Barbara 
Hepworth's "Mother and Child.

54 hours 31 minutes O_O

-- 
Regards
     Stephen

Post a reply to this message

On 03/25/2011 06:23 AM, Paolo Gibellini wrote:
> Very nice, could be interesting even in kenya soapstone material.
> Paolo

I did some looking around ... I had no idea that there were so many 
varieties of soapstone. Different from region to region.

This version didn't turn out to bad ... thanks for the nudge!

Post a reply to this message

Attachments:
Download 'fertility.png' (901 KB)

Preview of image 'fertility.png'

>Jim Holsenback  on date 25/03/2011 23:07 wrote:
> On 03/25/2011 06:23 AM, Paolo Gibellini wrote:
>> Very nice, could be interesting even in kenya soapstone material.
>> Paolo
>
> I did some looking around ... I had no idea that there were so many
> varieties of soapstone. Different from region to region.
>
> This version didn't turn out to bad ... thanks for the nudge!
>
Thank you for the image!
Is very fine,
;-)
Paolo

Post a reply to this message