Page 4 of 5

Re: Milk (regular) optimized

Posted: Wed Feb 02, 2011 3:18 pm
by dag
Looks biased to me. Gotta test irl to compare.

Re: Milk (regular) optimized

Posted: Wed Feb 02, 2011 8:11 pm
by suvakas
Lemo wrote:30 days milk with a hundred percent unbiased physical renderer :lol:
Was it fresh when you started rendering? :)

Re: Milk (regular) optimized

Posted: Wed Feb 02, 2011 10:59 pm
by Lemo
Yea, but I guess a cockroach or two drowned in the milk in the meantime...
Can't avoid the bugs

Re: Milk (regular) optimized

Posted: Fri Mar 04, 2011 8:28 am
by dag
Well, that didn't work.
Lowfat milk, regular milk, lowfat choc milk, regular choc milk.

Re: Milk (regular) optimized

Posted: Fri Mar 04, 2011 9:22 am
by Meelis
dag wrote:Well, that didn't work.
Lowfat milk, regular milk, lowfat choc milk, regular choc milk.
Absorption is in balance with scattering.
If u lower the scattering then absorption color takes over.

Still the renders look interesting.

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 10:39 am
by dag
Drink your milk boys.

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 11:53 am
by StompinTom
Looks delicious! :D

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 6:08 pm
by suvakas
Nice !!

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 7:27 pm
by zeitmeister
Great results, dag!
How long did you render, and in which mode?

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 7:52 pm
by dag
Deliciously long...

Milk rendered to 20k spp in 131 hours and it's still a little noisy, and it doesn't seem to clear up anytime soon. Minimal absorption and high scattering is not a good combo.
Chocmilk 16k spp in 22 hours thanks to higher absorption.

Rendered with an i7 860 with MLT. Bidir is like drinking a gallon of syrup with that kinda scattering.

Re: Milk (regular) optimized

Posted: Tue Apr 19, 2011 10:19 pm
by PureSpider
Amazing milk! :shock:
Can you share your materials please? 8)

Re: Milk (regular) optimized

Posted: Wed Apr 20, 2011 12:19 am
by lycium
For SSS you probably want to use single directional PT with MLT on - as you say bidir really slows to a crawl when there are so many scattering events (as it tries to connect each such event to every other between the eye and light paths).

Re: Milk (regular) optimized

Posted: Wed Apr 20, 2011 12:34 am
by dcm
lycium wrote:For SSS you probably want to use single directional PT with MLT on - as you say bidir really slows to a crawl when there are so many scattering events (as it tries to connect each such event to every other between the eye and light paths).
+1

Re: Milk (regular) optimized

Posted: Wed Apr 20, 2011 2:25 am
by Zom-B
lycium wrote:For SSS you probably want to use single directional PT with MLT on - as you say bidir really slows to a crawl when there are so many scattering events (as it tries to connect each such event to every other between the eye and light paths).
Hmmm, quite interesting.. I thought BiDir does help here since it optimized the ray paths.

Is this BiDir thing a general SSS issue, or just tilt over on a certain point of SSS strength?

Re: Milk (regular) optimized

Posted: Wed Apr 20, 2011 2:49 am
by lycium
In general when you have highly scattering media, bidir will become quite slow.