strange line in water

Started by Dune, December 14, 2019, 03:21:57 AM

Previous topic - Next topic

Dune

And another issue I can't explain. Does anyone recognize these strange lines in the water. Any idea how they could have gotten there? I will check it out myself again, but I'd like to hear any suggestions.

Oshyan

Pretty hard to say without knowing more about the scene and configuration of the water, etc. I'd just call it subsurface detail. ;)

- Oshyan

Dune

Yeah, I thought it would be difficult to pinpoint. It's not in the PT/RT preview, so I don't know where it came from. We'll just hope it disappears...

gasbutan

I had a similar effect recently:

test200k2.jpg

The interesting thing is: the shape is different in each new render (without changing any parameters). Sometimes it disappears completely.

Hannes

Ulco, if you hadn't mentioned the line, I would have thought, it's sort of a wet line or just a darker area of the ground. It doesn't really look incorrect to me.

Matt

#5
I think this is the edge of where the GI cache affects the riverbed. You can extend this range by reducing the GI cache detail or increasing the blur radius. Alternatively, use the path tracer. If none of these options are acceptable, you'll have to darken the riverbed before it reaches this line (and then you could fake the lost colour using luminosity if necessary).
Just because milk is white doesn't mean that clouds are made of milk.

Dune

That sounds logical. I hardly ever use GI cache, but Dorian did, I saw in the file. Thanks Matt!

WAS

Looks like same shadow artifacts I posted on in underwater surface discussion.

This was never a problem before. Underwater scenes rivers, oceans. Even when using a card surface with clouds. Would be nice to have a fix rather than fiddling with settings. Sooooo many bugs pouring out of these latest updates.

Matt

#8
It has always worked this way; it's not a new bug. Surfaces seen through transparency or reflection aren't sampled in the GI prepass, but they will use nearby GI samples that were taken from above the surface (in this case the nearby trees and dry land), and at some distance from these samples the contribution falls off to 0. The path tracer doesn't use the GI cache on surfaces so it doesn't have this problem.
Just because milk is white doesn't mean that clouds are made of milk.

Oshyan

Quote from: Dune on December 15, 2019, 09:45:53 AMThat sounds logical. I hardly ever use GI cache, but Dorian did, I saw in the file. Thanks Matt!
To be clear, if you use GI at all, you are using caching! The cache is generated at render time and kept in memory. This is different than GI *cache to disk*, of course, but both use a GI cache.

- Oshyan

Dune

Yes, of course, didn't realize that.

WAS

Quote from: Matt on December 15, 2019, 12:08:32 PMIt has always worked this way; it's not a new bug. Surfaces seen through transparency or reflection aren't sampled in the GI prepass, but they will use nearby GI samples that were taken from above the surface (in this case the nearby trees and dry land), and at some distance from these samples the contribution falls off to 0. The path tracer doesn't use the GI cache on surfaces so it doesn't have this problem.

Perhaps it's just gotten "dumber" because I don't remember this being a big issue before. Such as returning to my underwater scene to encounter them in brute force. There are so many scenes that this would be prevalent on in the past without much question or error. Hmm.