Page 1 of 3
Clorophyl test, need help plz
Posted: Tue Mar 13, 2007 5:19 am
by Kram1032
This should be a testscene for leaves... It's not that good, now
I know, it needs bumpmap and stuff...
it's lit by a sun!
Atm reinhard 1/1/6...
More DoF is required, I guess, but my real problem atm is the mat.
Here is the code:
Code: Select all
<medium>
<name>Clorophyl</name>
<precedence>20</precedence>
<basic>
<ior>1.5</ior>
<cauchy_b_coeff>0</cauchy_b_coeff>
<absorption_coefficient_spectrum>
<rgb>
<rgb>1 0.1 1</rgb>
</rgb>
</absorption_coefficient_spectrum>
<subsurface_scattering>
<scattering_coefficient_spectrum>
<uniform>
<value>1</value>
</uniform>
</scattering_coefficient_spectrum>
<phase_function>
<uniform/>
</phase_function>
</subsurface_scattering>
</basic>
</medium>
<material>
<name>Clorophyl</name>
<glossy_transparent>
<exponent>100</exponent>
<internal_medium_name>Clorophyl</internal_medium_name>
</glossy_transparent>
</material>
any ideas?
Posted: Tue Mar 13, 2007 6:26 am
by Kram1032
21 have been here, but 0 answered...
Does this mean, that you think, the mat already is ok?
Posted: Tue Mar 13, 2007 8:11 am
by neepneep
I don't have any leaves on hand at the moment to compare but I think bump mapping and a thin waxy surface layer would help sell it better.
Posted: Tue Mar 13, 2007 8:11 am
by Stur
Sorry, as I'm new I have no idea for your mat. However, the shadows look strange, as if the leaves were generating caustics.
Posted: Tue Mar 13, 2007 8:48 am
by Kram1032
yes... bumpmap would be great...
I didn't find a good one, yet...
yes the caustics are looking strange too, but I guess, if I'd render it longer, they'd mostly disappear.
ATM I'm testing blend-mats:
one with green absorption, orange SSS and 500 exponent
and one with green absorption, greenblue SSS and 50 exponent.
both IOR 1.5 and precedence 20...
Maybe, I could solve the problem with the thinner edge, where light goes through more, if I had an alpha-map, that graduently changes from one mat to the other...
I have problems with DoF: it's suposed to be very high DoF, but I can't see any at all..
Maybe this is a matter of time too...
How to set f-stop?
Code: Select all
<medium>
<name>Clorophyl.002</name>
<precedence>20</precedence>
<basic>
<ior>1.5</ior>
<cauchy_b_coeff>0</cauchy_b_coeff>
<absorption_coefficient_spectrum>
<rgb>
<rgb>1 0 1</rgb>
</rgb>
</absorption_coefficient_spectrum>
<subsurface_scattering>
<scattering_coefficient_spectrum>
<rgb>
<rgb>0.75 0.5 0</rgb>
</rgb>
</scattering_coefficient_spectrum>
<phase_function>
<uniform/>
</phase_function>
</subsurface_scattering>
</basic>
</medium>
<material>
<name>Clorophyl.002</name>
<glossy_transparent>
<exponent>50</exponent>
<internal_medium_name>Clorophyl.002</internal_medium_name>
</glossy_transparent>
</material>
<medium>
<name>Clorophyl.001</name>
<precedence>20</precedence>
<basic>
<ior>1.5</ior>
<cauchy_b_coeff>0</cauchy_b_coeff>
<absorption_coefficient_spectrum>
<rgb>
<rgb>1 0 1</rgb>
</rgb>
</absorption_coefficient_spectrum>
<subsurface_scattering>
<scattering_coefficient_spectrum>
<rgb>
<rgb>0 0.5 0.75</rgb>
</rgb>
</scattering_coefficient_spectrum>
<phase_function>
<uniform/>
</phase_function>
</subsurface_scattering>
</basic>
</medium>
<material>
<name>Clorophyl.001</name>
<glossy_transparent>
<exponent>500</exponent>
<internal_medium_name>Clorophyl.001</internal_medium_name>
</glossy_transparent>
</material>
<material>
<name>Clorophyl</name>
<blend>
<a_name>Clorophyl.001</a_name>
<b_name>Clorophyl.002</b_name>
<blend_factor>0.66666666666</blend_factor>
</blend>
</material>
Posted: Tue Mar 13, 2007 10:04 am
by drBouvierLeduc
The material in itslef seems ok, but I think the shape of the leaves needs more work.
Add somes details, and make them look less "perfect".
Maybe a bump map will be enought, but I suggest you subdivide all that stuff and add details, like slight random noise and spots, make it more thin on the edges than in the middle etc...
Btw do they have thickness or are they flat ?
I noticed that differences in thickness make wonders with sss, when lite by behind.
This comes at a price : high polycount, but it could make the difference I think.
Keep going !
Posted: Tue Mar 13, 2007 10:13 am
by Kram1032
If I didn't have thickness, they'd appear black with anything aove IOR 1 wouldn't they
yes, that should be the main work on it... but it should have been a test...
Maybe, better clorophyl would require SSS-colmaps...
Posted: Tue Mar 13, 2007 10:42 am
by DaveC
I'm going to do some tests, Kram. I'll report back with my findings
Posted: Wed Mar 14, 2007 2:47 am
by Kram1032
Ok, thanks
I actually wanted to make this, as nobody seemed to do it until now...
Posted: Sun Mar 18, 2007 12:20 am
by Kram1032
Did you find something?
I found, that chlorophyll has two peaks:
Wikipedia wrote:The absorption peaks of Chlorophyll a are at 665 nm and 430 nm. Chlorophyll a fluoresces at 673 nm.
Is there any way to exactly translate these to rgb? (Or to absorbtion amount)
Normal plants are using a mix of Chlorophyll a and Chlorophyll b...
So I should blend these two for a nice mat...
Posted: Sun Mar 18, 2007 3:16 am
by Colezero
This looks good, but what about SSS.
Some Subsurface scattering would be great.
You need bumpmaps....every organic stuff, like leaves or skin or what ever, are rough, not that even.
Anyway looks really good, someday someone needs to great a database, something like the one from maxwell, where users are able to upload materials, you know(
http://mxmgallery.maxwellrender.com/)
Posted: Sun Mar 18, 2007 7:20 am
by Kram1032
Read the material code...
It actually IS SSS
That's why I've started this! I wanted to make leaves.
Yes, I know, bumps wont be bad...
I did a highpoly leave with a displacement modifier (blender's new tools are great
) but it was TOO high poly -> Indigo crashes
At least, I found out, what was wrong with DoF
I just forgot that the mainly common sight of DoF is wrong -> DoF /=\ Blur
Posted: Sun Mar 18, 2007 7:40 am
by suvakas
Wow..this is starting to turn into a science project.
It would be interesting to see where will you end up an what the results will be.
Suvakas
Posted: Sun Mar 18, 2007 7:47 am
by Kram1032
Poorly, there is no thing, that says, how the absorbance is calculated... Is it in meters?
If so, I'd need something, that can calculate nm to specific colours...
Posted: Mon Mar 19, 2007 1:32 am
by Colezero
Read the material code...
It actually IS SSS Wink
Whooops...
i didn't read the code^^, i just took a look on the picture...^^
Now i know what i don't like on this image: all the leaves have the same size, same bending, same look.
i know this is a material-test.
but here is what i mean:
