Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[VBSmiles]

maybe... havent yet read all of the forum so it may have been noted.
If you need it I will send the file.



Seems that it doesn't like monti carlo

[RudySchneider]

Are you sure?  It looks like the sort of artifacts you get when using volumetric instead of ray-traced shadows.  Volumetric shadows have been known to be buggy forever.

[VBSmiles]

Yeah, I remember that bug a while back, but this is definately raytraced shadows. take note that it is only on the reflection itself.

[hihosilver]

This has happened to me as well.  I'm not exactly sure of the problem.  Sometimes changing the 'Samples Per Pixel' in the AA Setup section helps.
this issue has been posted before though.  Hopefully it will be fixed.

[Steve]

It looks like there's a problem with the way the light is sampled.  Can you send me the .an8 project for this?  I have been unable to find this with the other examples I have --- there's just too much data to dig through in them.  Maybe I'll get lucky with this one.

[VBSmiles]

Bah... didn't save the an8 sorry Steve

However I made another thread in relation to this one where I deturmined that global lighting works fine. I found that the problem was with Local lights. ( didn't check the spot lights though. )

Another thing I had messed with was pixel sampling but don't think that was it. I'll try to recreate the problem while I have time tonight, and send it to you.

[VBSmiles]

Sent ya a file...... pretty simplistic, to help you out.

  I keep refering infinate lighting as global.

[Steve]

Got it, and, yup, it's got stripes!  Something is definitely whacky in my pseudo random number tables.  Thanks for the project.

#097-017 - Reflections show horrizontal bands in ART local light using soft shadows.

[Steve]

Wa-hoo!  I found a subtle but serious problem with the fundamental way that Anim8or generates samples.  Single samples are fine but they can become correlated with subsequent samples of the same value.  For example, a reflected surface samples lights twice, once on the original hit and once on the surface hit by the reflection.  Correlated samples are bad because the loose some of their "randomness" and can appear as many kind of artifacts.  This happens in the Reflective Egg example when the number of samples is a multiple of 4 (like 4, 16, 36) but not nearly as much at other numbers (9, 25, 49) though it still can happen.

I've fixed the problem

[VBSmiles]

Good job Steven!