SSS material tests

[OnoSendai]

I plan to do some tests, messing round with parameters here.
Test scene is a modified one of radiances.
Will put up for download soon.
You will need to download the dragon from here:

http://www-graphics.stanford.edu/data/3Dscanrep/

The dragon is about 11cm tall in these renders.

You'll have to wait for 0.7 test 2 to use the henyey_greenstein phase function



EDIT: replaced with 2h45m render time version.

Material:

Code: Select all

	<medium> 
		<name>jade</name> 
       
		<ior>1.6</ior> 
		<cauchy_b_coeff>0.0</cauchy_b_coeff> 
		<precedence>20</precedence> 
			
		<absorption_coefficient_spectrum> 
			<rgb> 
				<rgb>2000 2 2000</rgb> 
			</rgb> 
		</absorption_coefficient_spectrum> 
       
		<subsurface_scattering> 
			<scattering_coefficient_spectrum> 
				<uniform>
					<value>1000</value>
				</uniform>
			</scattering_coefficient_spectrum> 
			
			<phase_function> 
				<henyey_greenstein>
					<g>0.8</g>
				</henyey_greenstein>
			</phase_function> 
		</subsurface_scattering> 
	</medium> 
    
    <material> 
		<name>default</name> 

      <!--specular> 
         <transparent>true</transparent> 
         <internal_medium_name>jade</internal_medium_name> 
      </specular--> 
      
		<glossy_transparent>
			<internal_medium_name>jade</internal_medium_name> 
			<exponent>600</exponent>
		</glossy_transparent>

	</material>

[fused]

That looks absolutely beautiful! Awesome render.

nice dof btw


when can we expect test 2?

[OnoSendai]

test 2 will be out in the next few days i think.

[Camox]

Play with sss. The results are not really adequate. I have to find big difficulties the right values.

The yellow material is from Bandar ! Testscene from Ono and Bandar, thx guys for sharing.

[sprocket]

Beautiful material!

Just curious, you say it is 11cm tall, is one unit in Indigo equal to one metre?

You'll have to wait for 0.7 test 2 to use the henyey_greenstein phase function Smile

Does thet mean there are other scattering functions like mie and rayleigh in the pieline too??

[OnoSendai]

Yes, Indigo units are metres.

I may get round to Rayleigh and Mie phase functions.
Would be pretty cool to actually simulate atmospheric scattering etc... with just a sunlight

[CTZn]

thx guys for sharing.

Yes, thx people for sharing ! I'm silently copying every posted code in individual xml files for later use

[Camox]

thx guys for sharing.

Yes, thx people for sharing ! I'm silently copying every posted code in individual xml files for later use

Jo silently and auricularly ! What you use ? Your self contrived xml code ?

You are the best.

[Kosmokrator]

i think Ono u are genius...anyway i found an calculator for
Mie phase function i think may help u in some way

u have done amazing work in a sort time!!

http://www.lightscattering.de/MieCalc/

[domparis]

Good result Nick, good to know a test2 is almost there
So i trying to get some decent result in a way to simulate skin, but like other already said, the specular effect is to much important at a ior between 1.3 and 1.5 and if i set it to a value near 1, i lost the real effect inside the volume.
Have u tracked the bug with skylight ?

Good work and thanks for your effort doing so many ehancement to your renderer !!

Dom

Ok, after seeing your code, using properly the glossy_transparent function help a lot

[Bandar]

Would be pretty cool to actually simulate atmospheric scattering etc... with just a sunlight

Oooooh, yes

[YaroslavL]

Can anybody share *.blend file with some SSS material?
Thanks.

[danyolgiax]

You will need to download the dragon from here:

http://www-graphics.stanford.edu/data/3Dscanrep/

What model i've to download?

Range data:
dragon_stand.tar.gz (6.1 MB compressed, 23 MB uncompressed)
dragon_side.tar.gz (4.2 MB compressed, 16 MB uncompressed)
dragon_up.tar.gz (5.7 MB compressed, 24 MB uncompressed)
dragon_fillers.tar.gz (6.7 MB compressed, 26 MB uncompressed)
dragon_backdrop.tar.gz (11 MB compressed, 44 MB uncompressed)

Vripped reconstruction:
dragon_recon.tar.gz (11 MB compressed, 43 MB uncompressed)

???

tnx

[OnoSendai]

You want
dragon_recon.tar.gz

[mrCarnivore]

I really have some problems with the dragon. I dl'ed the dragon and imported the .ply into blender. I tried res3 and res4 and with both I get some normals flipped and also non-manifold mesh. Re-calculating the normals made things even worse. I think it's because oof the non-manifold mesh that blender gets confused.
Can somebody pls post a blend with the dragon where those problems are not occuring?

Also I'm a little confused by the sss. I used blendigo to try to use this material:
http://www2.indigorenderer.com/joomla/f ... =8824#8824

My material code is this:

Code: Select all

	<medium>
		<name>DragonMaterial</name>
		<precedence>10</precedence>
		<basic>
			<ior>1.2</ior>
			<cauchy_b_coeff>0</cauchy_b_coeff>
			<absorption_coefficient_spectrum>
				<rgb>
					<rgb>0 4.5 5</rgb>
				</rgb>
			</absorption_coefficient_spectrum>
			<subsurface_scattering>
				<scattering_coefficient_spectrum>
					<rgb>
						<rgb>4 20 200</rgb>
					</rgb>
				</scattering_coefficient_spectrum>
				<phase_function>
					<uniform/>
				</phase_function>
			</subsurface_scattering>
		</basic>
	</medium>
	<material>
		<name>DragonMaterial</name>
		<specular>
			<transparent>true</transparent>
			<internal_medium_name>DragonMaterial</internal_medium_name>
		</specular>
	</material>

Seems almost the same to me, doesn't it? The result, however, looks nothing like the original post: