Page 1 of 2

Indigo 0.7 test 3

Posted: Sat Jan 27, 2007 10:37 pm
by OnoSendai
http://www.indigorenderer.com/montecarl ... _test3.zip

Changelog:

* added dermis and epidermis media types.
* added basic medium type
* better spectral sampling of SSS
* fixed constant on Planck's formula
* removed sky gain
* Added scene pathname to window title
* Messed round with sky/sun model a little
* printing out line + column number on most scene file parse errors
* took out PANIC msg.
* added optional efficiacy_scale for rectangle and mesh lights.



------ How to use ------------

Use the epidermis and dermis media types like this:

Code: Select all

<medium>
		<name>epidermis</name>
		<precedence>2</precedence>
	
		<epidermis>
			<melanin_fraction>0.01</melanin_fraction>
			<melanin_type_blend>0.25</melanin_type_blend>
		</epidermis>
	</medium>
	
	<medium>
		<name>dermis</name>
		<precedence>3</precedence>
	
		<dermis>
			<hemoglobin_fraction>0.05</hemoglobin_fraction>
		</dermis>
	</medium>

<!-- example material-->
 <material>
		<name>outerboundary</name>

      <specular>
         <transparent>true</transparent> 
         <internal_medium_name>epidermis</internal_medium_name> 
      </specular>
	</material>
See Jensen's and Donner's paper for more information ( http://graphics.ucsd.edu/papers/egsr200 ... 06skin.pdf )

Basic media are now defined like this (note the introduction of the <basic> tag:

Code: Select all

<medium> 
		<name>uniform_medium</name> 
		<precedence>20</precedence> 
		
		<basic>
			<ior>1.0</ior> 
			<cauchy_b_coeff>0.0</cauchy_b_coeff> 
			
				
			<absorption_coefficient_spectrum> 
				<rgb> 
					<rgb>0 0 0</rgb> 
				</rgb> 
			</absorption_coefficient_spectrum> 
	       
			<subsurface_scattering> 
				<scattering_coefficient_spectrum> 
					<uniform>
						<value>2000</value> 
					</uniform>
				</scattering_coefficient_spectrum> 
				
				<phase_function> 
					<uniform/>
				</phase_function> 
			</subsurface_scattering>
		</basic>
	</medium>


To use the efficacy scale element:
See http://en.wikipedia.org/wiki/Luminous_efficacy for some information and some example values.

Code: Select all

<rectanglelight>
		<pos>1.4 1 2.6</pos>
		<width>0.1</width>
		<height>0.1</height>

		<spectrum>
			<blackbody>
				<temperature>3500</temperature>
				<gain>1</gain>
			</blackbody>
		</spectrum>
		
		<efficacy_scale>
			<power_drawn>150</power_drawn>
			<overall_luminous_efficacy>17.5</overall_luminous_efficacy>
		</efficacy_scale>
	</rectanglelight>

Posted: Sat Jan 27, 2007 11:55 pm
by Kosmokrator
u are absolutely insane Ono...dermis And Epidermis?i think its too much for SSS.... :wink: :shock: :shock: nice work and thnx for release...

Posted: Sun Jan 28, 2007 12:23 am
by domparis
Coool 8)
Thanks for this new update nick !!!

Dom

Posted: Sun Jan 28, 2007 12:50 am
by Pinko
thanks for new up ONO....!!!!!
Pinko. :D

Posted: Sun Jan 28, 2007 12:53 am
by neepneep
Oh bless yee wee kiwi essence of being. :P :lol: :P :lol:

I am gonna give the new lights a whirl but I am little saddened by the loss of the panic message. :cry: :oops: :cry: :cry:

Thanx

Posted: Sun Jan 28, 2007 2:33 am
by tungee
This is Übercorrect man.

Thanx

Posted: Sun Jan 28, 2007 2:34 am
by tungee
This is Übercorrect man.

Posted: Sun Jan 28, 2007 2:52 am
by zsouthboy
Nick, you are now the sexiest man in the world!

What are you going to do now?

I"M GOING TO DISNEYLAND!






:D

You rock!

Posted: Sun Jan 28, 2007 12:14 pm
by CTZn
Thank youuu !

Posted: Sun Jan 28, 2007 5:24 pm
by Wedge
Good work! I am updating the exporter right now. Thanks for another great release!

Posted: Sun Jan 28, 2007 10:43 pm
by Zom-B
A Day when a new Indigo Test Version gets released, is a good Day :D

I put my Hands on the new epidermis/dermis stuff.

As far as I understood, its still necessary to have a model for the dermis, and one for the epidermis, so I tried giving my Buddha Test a new Skin.

its a 14m C4D model (reduced PLY to 655534 Tris) that gets scaled in the XML by multiplicator of 0.01 to 14cm...
I did an extrude to the dermis model to -0.07meters so the epidermis should be, after scaling down in the XML, about 0.07mm (thats what I find on the www for hairy skin).

I finally have no idea if scaling down in C4D using negative extrude is a good idea, or if the whole scale thing is done right, but here my result, rendered for about 15h (7k muts per pixel):

Image

looks good to me, but the reflections are way to high.... how do I reduce them?
Any possibility for me to add exponent or IOR settings to dermis/epidermis?

BTW:
I found this paper on the web... perhaps something interesting... perhaps not :wink:

Posted: Mon Jan 29, 2007 5:36 am
by Zom-B
Here a new dermis/epidermis test, with strange results:

I did a 500m big ring (c4d likes it big by default :roll: ) and exportet it using a scale factor of 0.000001.
again default dermis & epidermis settings like Nick posted above, BUT I used a Bump map on the Epidermis!
The ring is only illuminated from behind, so I expected some nice skin SSS effect...

OK here the result (3h rendering for 2,7k muts per pixel):

Image

you notice this strange halo effect, so I took a look at the untonemapped.exr,
and this spline like shadow gets visible ( a light mesh overlay from c4d)

Image

I zipped the whole Project, if somebody wants to take a look >>link<<

and here some stuff I found in test 3:

- the changelog isn't updated :wink:
- untonemapped.exr has resolution multiplied by super_sample_factor *cool*
- Black_body is now totally overpowered, I've got to reduce gain from 1.0 to 0.00000001 to gain a equal untonemapped.exr like in 0.7t2 :?

does anybody else played around a bit with this version???

Posted: Mon Jan 29, 2007 10:25 am
by Phil
Hi,

thanks for this new release Ono :D

@ ZomB:

I played around, tryed to improve my wine glass.
So I played with blackbody, and I combined them with a kitchen probe map.
I had to increase the gain of the map from 1 to about 1e5 to 1e6 :?

Regards, Phil

Posted: Tue Jan 30, 2007 11:25 pm
by OnoSendai
Phil wrote:Hi,

thanks for this new release Ono :D

@ ZomB:

I played around, tryed to improve my wine glass.
So I played with blackbody, and I combined them with a kitchen probe map.
I had to increase the gain of the map from 1 to about 1e5 to 1e6 :?

Regards, Phil
Yeah, i changed the Planck formula code a bit.
don't be afraid of the big numbers :)

Posted: Wed Jan 31, 2007 9:24 pm
by 8bstudio
Ono... absolutely wonderfull! No comments! Very good and perfect work!! :o :o :o :o :o :o :o :o :o