A natural situation

Get feedback from others on your works in progress
User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Tue Dec 31, 2013 6:12 am

Ono, about the medium air; is a universal definition? (Strange things, but in the case of extra atmospheric sun, travell the light from sun to air medium (a giant space of calculus) to athmosphere to earth?).
Last edited by yonosoy on Tue Dec 31, 2013 6:54 pm, edited 1 time in total.
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Re: A natural situation

Post by Zom-B » Tue Dec 31, 2013 8:36 am

OnoSendai wrote:You are using 0.353, I would suggest changing to 0.844.
Here a 3300s/px render of the same scene with adjusted HG value:
a natural situation_01.jpg
and here the IGI file: https://dl.dropboxusercontent.com/u/693 ... ion_01.igi
polygonmanufaktur.de

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: A natural situation

Post by CTZn » Tue Dec 31, 2013 10:31 am

OnoSendai wrote:* Indigo actually supports data from MiePlot. Currently I use the haze phase function for haze water vapour scattering in the atmosphere. I guess I could introduce support for the cloud droplet phase function, which would be more accurate than the Henyey Greenstein approximation.
I may have misunderstood some point regarding the phase function, if not it all. I thought that negative values in Indigo referred to backwards scattering, and positive values to forward scattering, while miePlot is providing strictly positive values based on the angle of ray incidence. But this doesn't sum up I'm realizing, specifically when zero is coming into play.

I'd gladly have more details on how Indigo is interpreting tabulated data VS. ie RGB data in PF, in a simplified way probably.

Now again if someone happens to stumble upon the tabulated data from miePlot I distributed earlier to this date, it is a truncated data set indeed.
obsolete asset


User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Tue Dec 31, 2013 7:17 pm

Update of the raw comparison.
Left: g=0.353 (a value) 3000 s/px
Right: g=0.844 (the physic value) 3300 s/px
a natural situation (two g values).jpg
Thanks Zom-B.

EDIT: The beauty of collect the exr raw format is that could be reevaluated forever. Hope don't to bore with this. A developt solution with adobe RGB.
Attachments
a natural situation (two g values)_b.jpg
Last edited by yonosoy on Sat Jan 04, 2014 3:02 am, edited 3 times in total.
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

User avatar
CTZn
Posts: 7240
Joined: Thu Nov 16, 2006 4:34 pm
Location: Paris, France

Re: A natural situation

Post by CTZn » Tue Dec 31, 2013 7:20 pm

I think that the clouds volume would deserve a much greater height; also keep in mind that the model is lacking high atmospheric layers and other phenomenon that would somewhat dim the Sun light.

I'm still amazed at this all thanks to you guys !
obsolete asset

User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Wed Jan 01, 2014 1:31 pm

A Mie web calcutor http://omlc.ogi.edu/calc/mie_calc.html

Extract of the results for initial values and a IOR of 1.33

# Mie Scattering # Version 2-3-3 (15:30:19 01/28/12)
# Oregon Medical Laser Center # http://omlc.ogi.edu
# by Scott Prahl # prahls@ohsu.edu
#
# radius 0.50000 [microns] (sphere radius)
# n_medium 1.00000 [---] (refractive index of medium)
# n_real 1.33000 [---] (refractive index of sphere)
# n_imag 0.00000 [---] (absorption of sphere)
# lambda_vac 0.63280 [microns] (wavelength in vaccuum)
# density 0.10000 [#/micron^3] (spheres per cubic micron)
#
# lambda 0.63280 [microns] (wavelength in medium)
# X 4.96459 [---] (size parameter)
# g 0.84539 [---] (average cosine of phase function)
# Qsca 3.5786 [---] (scattering efficiency)
# Qext 3.5786 [---] (extinction efficiency)
# Qback 0.38038 [---] (backscattering efficiency)
# Csca 2.8106 [micron^2] (scattering cross section)
# Cext 2.8106 [micron^2] (extinction cross section)
# Cback 0.29875 [micron^2] (backscattering cross section)
# mu_s 281.06 [1/mm] (scattering coefficient)
# mu_s' 43.456 [1/mm] (reduced scattering coefficient)
# mu_t 281.06 [1/mm] (total attenuation coefficient)
#
# The second column is normalized so that the integral of it over
# 4*pi steradians will be unity. The average of the 3rd & 4th
# columns is the second. The next three columns are normalized
# to the value at 0 degrees.
#
# natural = (|S1|^2+|S2|^2)/(pi X^2 Qsca)
# perpen = |S1|^2/(pi X^2 Qsca)
# parallel = |S2|^2/(pi X^2 Qsca)
# polarization = (|S1|^2-|S2|^2)/(|S1|^2+|S2|^2)
# S33 = Real(S2 * S1^*)
# S34 = -Imag(S2 * S1^*)
#
##theta natural perpen parallel natural perpen parallel polarization S33 S34
-176.400 0.00826 0.00812 0.00841 0.00402 0.00395 0.00409 -0.01755 -0.99938 -0.03045
-172.800 0.00774 0.00722 0.00826 0.00377 0.00351 0.00402 -0.06769 -0.98990 -0.12460
-169.200 0.00706 0.00607 0.00805 0.00343 0.00295 0.00391 -0.13973 -0.94785 -0.28644

.....

Zom-B, the pigs with IOR=1.0 ,g=0.84539 and two changes in the parameters, sacttering=281.0 (with the doubt about if this parameters are right related) and absorption=0.
Some minimal explanations of how to incorporate the data, please... There are part of the data that actually could not be.
Believe that the IOR could be now 1 (maybe break...). Light has change dramatically over scene (what I do?).
Edit: light changes because g=0.84... (not viewed this process). The light is subtle less dim. Do another one since there are doubts if you wish.

Note: the real tabulated data by itself don' t gives any kind of fireflies!!!
Attachments
a natural situation_02.pigs
(17 KiB) Downloaded 272 times
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Wed Jan 01, 2014 11:42 pm

Hi. Happy New Year and sorry (it's holydays...).
A test over pure water and material test scene. At the same time condition.
Left: normal situation.
Right: base blended with precedence.
pure_water_test.jpg
The material has real tabulated absorption and uniform SSS with value=0 (experimental).
pure_water.pigm
experimental
(951 Bytes) Downloaded 237 times
base_p.pigm
A possible mix
(18.98 KiB) Downloaded 223 times
Try to detail more some over (if possible).
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Re: A natural situation

Post by Zom-B » Thu Jan 02, 2014 6:28 am

rendered the last version for 6000 s/px.
looks very similar to the last version, regarding the data its out of my league to understand it with my new years eve hangover :D

https://dl.dropboxusercontent.com/u/693 ... ion_02.igi
a natural situation_02.jpg
polygonmanufaktur.de

User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Thu Jan 02, 2014 9:26 am

<ior>1.00000003</ior> could be 1.0, good new (here a little bit faster version (and more stable sample count, the slow cam)).
If you test as is intentioned with the wireframe shader over 16 subdivisions (in this case, vary with cam position and maybe the code) artifacts appear.
A server could not express... thankfully for the suppot.
A real tabulated absorption for clouds don't give any kind of fireflies at all, the lost of data string (number of values are wrong, the extreme high numbers of values,... gives a lot).
CTZn for more over, seems that is better option a g=0.84... and then the chroma (tabulated data) will be related through absorption (Ono's test for example). Some test could be done in this direction with the actual (modified) rig from here. If success the pigs will be post.
Thanks to Ono too for the effort and the success (LOD, awesome; bump and displacement repercusions was very high).
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

User avatar
OnoSendai
Developer
Posts: 6241
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Re: A natural situation

Post by OnoSendai » Fri Jan 03, 2014 11:29 am

galinette wrote:Hi Ono,

Last images are interesting for my work, as I'm struggling getting perfect high definition HDRI envmaps with direct sun & clouds. The sun is so bright that it requires ND filters with very high optical thickness of about 3 (1/1000 factor), which are never neutral despite the name, and color correction is not perfect. So I'm thinking of baking envmaps.
Which atmospheric gas data are you using (paper?) ? Would it be possible to change it in table data?

Cheers,

Etienne
Hi Etienne,
I'm using data from a combination of papers.
The problem is every paper gives a different value for key constants :)

See 'Display of the Earth taking into account Atmospheric Scattering' Nishita 1993
for altitude density profiles.

and

http://augerlal.lal.in2p3.fr/pmwiki/upl ... choltz.pdf

for scattering coefficients.

User avatar
OnoSendai
Developer
Posts: 6241
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Re: A natural situation

Post by OnoSendai » Sun Jan 05, 2014 1:10 pm

Made a blog post about your guys' work among other things.
http://www.indigorenderer.com/content/a ... provements
Cheers!

User avatar
yonosoy
Posts: 3432
Joined: Sat Dec 01, 2012 9:45 pm
Location: Heaven

Re: A natural situation

Post by yonosoy » Fri Jan 10, 2014 10:23 am

Hi. Only for a question.

It's an "old school" aproximation of a thin film through precedence concept (with ingenuity again !!).
The sequence of mediums are: atmosphere, diamond (material database), gold, vacuum.
What are we seen here?
Bronzes (natural flowers)_probe_test.jpg
atmosphere-diamond-gold-vacuum.pigs
(35.15 KiB) Downloaded 250 times
There is a official request of Zom-B (unanswered) about a trick in order to a high level of compatibility for nk files. Seems that actually only the "solid" part of the nk definition is usable.
So, are the materials damaged in this mixtures? This is because the pure water test; this one is a complete example.

Several goles for this old one:
An approach to the inorganic material and the first try to a finished photographic serie (for this year).
Will be composed of a probe test (it's like a multiply table: this is the first regular prism with the first simply intention (no more than three operations)), some referencial objects with a meaning of scale and some kind of small simulations.
Amongs others, the complet periodic table, some primary materials, some gases mixtures,and some more metals and minerals will be expressed as nk files and real tabulated spectrums (and posted periodically after some test...).

Atmosphere seems to work well for bidireccional with all kind of lights except extra atmospheric sun.
For this case, bidirectional metro with an environment fill ligth.
120 hours of process (i3) !!!! If is a plausible physic behaviour the computer do it at expenses of time... If not, ...
Is expectable a thin film material in some point (quasi rhetoric question)?
Thanks in advance.
"... nor 0.2 galaxys, nor 0.8 little chikens ..."

neo0.
Posts: 1784
Joined: Thu Feb 28, 2008 7:11 am
Location: the US of A

Re: A natural situation

Post by neo0. » Fri Jan 10, 2014 8:03 pm

Wow, very impressive stuff guys. Any chance a simple ui could be added for those of us who arent very adept with coding? :)

User avatar
Zom-B
1st Place 100
Posts: 4700
Joined: Tue Jul 04, 2006 4:18 pm
Location: ´'`\_(ò_Ó)_/´'`
Contact:

Re: A natural situation

Post by Zom-B » Fri Jan 10, 2014 11:35 pm

neo0. wrote:Wow, very impressive stuff guys. Any chance a simple ui could be added for those of us who aren't very adept with coding? :)
There already is since quite a long time some code parameterization!
parameters.jpg
Here you see the bump shader for for the milled_aluminium of CTZn from the mat DB. Its up to the shader coder to setup such parameters, and name them understandable :)

Back in the day I requested also some description text for parameters, so you would get some mouse-over tooltips, maybe I should ask for that again...
*** EDIT*** I just heard that this is already possible, just maybe not documented probably...


The scene above is a real beauty but also eats samples for breakfast and takes ages to clear up..... seems like you rendered it for days -.-'
polygonmanufaktur.de

Post Reply
362 posts

Who is online

Users browsing this forum: No registered users and 30 guests