absorption and scattering coefficients for liquids

General questions about Indigo, the scene format, rendering etc...
User avatar
CoolColJ
Posts: 1734
Joined: Mon Jun 25, 2007 1:47 pm

Post by CoolColJ » Thu Aug 16, 2007 12:09 pm

yes milk renders black, but it is among the few there that have such high scattering values, when I divide em by 10 or 100 they start to render non-black, but still funny looking, kinda bluish/grey

so who knows

can't test the scene with v0.8 stable as the scene format is different.

User avatar
manitwo
Posts: 1029
Joined: Wed Jul 05, 2006 4:50 am
Location: Tirol - Austria

Post by manitwo » Thu Aug 16, 2007 5:22 pm

big thanks for that whaat! :)

User avatar
Whaat
Developer
Posts: 1827
Joined: Fri Dec 22, 2006 6:15 am
Location: Canada
Contact:

Post by Whaat » Sat Aug 18, 2007 2:42 am

Rendered with Indigo 0.8. The orange liquids are actually chocolate milk and regular milk. I think they are orange just because the convergence is so slow for this type of scene (10 liquids all with SSS, milk has VERY high SSS). I have tried the same milk values in another simpler scene and the milk seemed to render correctly (using Indigo 0.8).

Back row left to right: apple juice, chocolate milk, cranberry juice, grape juice, chardonnay
Front row left to right: coke, merlot, milk, budweiser, gatorade

All liquids using values for 100% concentration from the PDF I recently posted. Diameter across the top is 7.5cm for all liquids.
Attachments
liquids.jpg
approx 9 hour render
liquids.jpg (46.91 KiB) Viewed 2046 times

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Sat Aug 18, 2007 2:46 am

very strange looking liquids

User avatar
OnoSendai
Developer
Posts: 6227
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Sat Aug 18, 2007 2:48 am

Hi Whaat,
looks pretty good, apart from the Milk. That is just wierd :)

User avatar
Whaat
Developer
Posts: 1827
Joined: Fri Dec 22, 2006 6:15 am
Location: Canada
Contact:

Post by Whaat » Sat Aug 18, 2007 2:50 am

OnoSendai wrote:Hi Whaat,
looks pretty good, apart from the Milk. That is just wierd :)
I will render another simpler scene with just the milk and show you that it works (or someone else could do it.... :wink: )

User avatar
DaveC
Posts: 596
Joined: Mon Sep 11, 2006 12:20 am
Location: The Tottenham, London, UK
Contact:

Post by DaveC » Sat Aug 18, 2007 2:52 am

when i had this problem with milk (yes, the same problem), I changed it to glossy transparent, and it rendered fine. Try it in Indigo 0.9 test 6 and see if it comes out different.
The hardest part of BEING yourself is FINDING yourself in the first place...
http://thebigdavec.googlepages.com

User avatar
OnoSendai
Developer
Posts: 6227
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Sat Aug 18, 2007 2:52 am

Well I'm just setting up my own SSS test scene (wine glass), so I'll test it soon. But it definitely shouldn't be orange.
Tell me, what are your render parameters?
Can you paste the inifile or renderer_settings?

User avatar
OnoSendai
Developer
Posts: 6227
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Sat Aug 18, 2007 3:11 am

ZomB wrote:Great Work Whaat :!: :!: :!:

Such a PDF should be distributed directly with Indigo,
means in the manual!
Yup, good idea.
I'll add this as an appendix to the next version of the manual.

User avatar
Whaat
Developer
Posts: 1827
Joined: Fri Dec 22, 2006 6:15 am
Location: Canada
Contact:

Post by Whaat » Sat Aug 18, 2007 3:25 am

OnoSendai wrote:Can you paste the inifile or renderer_settings?
I don't have the render settings right now but I know I had Bi-dir MLT=on and I think MNCR set to 1000 (I never use lower than 500). I never change the inifile so the settings are the default from Indigo 0.8 stable.

I am now thinking that the orange color of the milk is due to a bi-directional bug. I rendered the milk again with with the same settings and it was orange. Now, I am rendering with bi-dir off and the milk is coming out white. Will post soon.
Remember, these tests are being done with Indigo 0.8 stable...

User avatar
OnoSendai
Developer
Posts: 6227
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Sat Aug 18, 2007 3:37 am

It could be a bug with bidir, yeah.
It's interesting to note that the absorption colour of milk is more or less orange (i.e. the absorption coefficient is lowest for red light).
What we are seeing in your renders looks something like the milk liquid being rendered without scattering being taken into account. Why this would be the case is a different matter entirely. :?

User avatar
Whaat
Developer
Posts: 1827
Joined: Fri Dec 22, 2006 6:15 am
Location: Canada
Contact:

Post by Whaat » Sat Aug 18, 2007 4:15 am

Exact same scene rendered with both Indigo 0.8 and 0.9test 6. Only setting that was changed was bidirectional. Yes, gamma is set to 1.0 in Indigo 0.9test6 :wink: Hope these screens help, Ono.

Milk on the left, chocolate milk on the right. Properties from Jensen's paper converted to 100% concentration Indigo values.
Attachments
Milk MLT BIDIR_08.jpg
Milk MLT BIDIR_08.jpg (239.28 KiB) Viewed 2010 times
Milk MLT_08.jpg
Milk MLT_08.jpg (220.35 KiB) Viewed 2010 times
Milk MLT BIDIR_09t6.jpg
Milk MLT BIDIR_09t6.jpg (245.97 KiB) Viewed 2010 times
Milk MLT_09t6.jpg
Milk MLT_09t6.jpg (250.76 KiB) Viewed 2009 times

User avatar
OnoSendai
Developer
Posts: 6227
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Sat Aug 18, 2007 4:19 am

Thanks for the tests Whaat.
There are two issues in the 0.9t6 renderings. The first is that the non-bidir milk is too dark. I think i have fixed this issue since test6; it was caused by a rather subtle interplay between nudge distances and extremely short mean path lengths in highly scattering media.

No clue about the bidir orange issue tho, will check that out soon :)

User avatar
Kram1032
Posts: 6649
Joined: Tue Jan 23, 2007 3:55 am
Location: Austria near Vienna

Post by Kram1032 » Mon Aug 20, 2007 10:13 am

those formulas, where they use "cup" or "tbsp" can be updated :D
100% is at 23 litres, right?

a "cup" is:

Cup

1 Imp.cup = 2 Imp.gi. = 10 Imp.fl.oz. = 80 Imp.fl.dr. = 4800 Imp.min = 17,338715625 cubic inch = 284,130642624675 Millilitre

1 Imp.gal. = 4 Imp.qt. = 8 Imp.pt. = 16 Imp.cup = 32 Imp.gi. = 160 Imp.fl.oz.

1 Imp.cup = 1,20095 US.cup
1 Imp.fl.oz. = 0,96076 US.fl.oz.

1 US.cup = 2 US.gi. = 8 US.fl.oz. = 64 US.fl.dr. = 3840 Imp.min = 14,4375 cubic inch = 236,5882365 Millilitre

1 US.gal. = 4 US.fl.qt. = 8 US.fl.pt. = 16 US.cup = 32 Imp.fl.gi. = 128 Imp.fl.oz.

Tea cup

1 pint = 2 cup = 3 tea cup

1 Imp.tea cup = 3200 Imp.min = 11,55914375 cubic inch = 189,42042841645 ml
1 US.tea cup = 2560 US.min. = 9,625 cubic inch = 157,725491 ml

1 Imp.tea cup = 1,20095 US.tea cup

or, in metric values:


1 metric cup = 250 Millilitre
1 metric gallon = 8 metric pint = 16 metric cup = 4 litre


1 cup = 200 Millilitre (in Japan)

traditional sake amount:
1 Gō = 2401/13310 litre ≈ 180,39 Millilitre


and the tbsp:
1 US tablespoon = 14.7867648 millilitres

and the one value with tsp:
1 US teaspoon = 4.92892161 millilitres


I found all those in Wiki and Google :D

I guess, as the tbsp and tsp values are US (99,9999999... % non metric), the used cup also will be the non metric one...

User avatar
Whaat
Developer
Posts: 1827
Joined: Fri Dec 22, 2006 6:15 am
Location: Canada
Contact:

Post by Whaat » Tue Aug 21, 2007 2:37 am

Updated render of same liquids using Indigo 0.8 (bidir=off). I think they all look good except for chocolate milk. It just doesn't look chocolatey enough.
Attachments
liquids.png
liquids.png (255.32 KiB) Viewed 2292 times

Post Reply
109 posts

Who is online

Users browsing this forum: No registered users and 4 guests