I did two entirely seperate renders and i am ending up with these tiny square distortions. Is it a lighting issue, is something going odd with displacement? This is something new that i never encountered before until now.
I haven't seen that before. Is it affected by "GI surface details"?
Matt
GISD settings is set to Global Illumination.
Occlusion weight is 3, the top image is set to 1.
Bounce to the ounce is 1.
Radius is 24.
Falloff is on.
What I mean is, does it go away if you turn GISD off?
If does, I suspect it might be caused by small holes in the surface due to displacement errors. These might be fixable with "force all edges" in Render Subdiv Settings. But it's still odd that it looks like this.
Matt
I just did that. There are bright spots now.
Where are the Render Subdiv Settings? I looked inside the render node.
I had this also a while ago, when using quite extreme small displacements. And I've seen it before in threads. I guess the planet's surface is overlapping, and you're looking through the other side. I tried to stop it by adding a second less or not displaced dark planet/sphere inside the planet, but that didn't help.
Well, poopie. I will change the fractal sizes and see what happens there.
I looked over all of my fractal sizes and nothing is under 1 meter in size. Honestly i never thought i was pushing the displacement that far on the small scales.
If you go to the Internal Network of any render node that's rendered at least one image already, there should be a node called "Render subdiv settings 01". If you don't see it, it's probably covered by another node, so move them around a bit.
Matt
Correct. It was covered by the Render GI settings node. I will reply with the results soon.
The squares are gone but there still some minor abnormalities.
The remaining gaps in that last render could probably be fixed by increasing "displacement tolerance" on the planet node, because it looks like they are falling along bucket boundaries. You have some extreme vertical displacement, I think, but it looks like a pretty good imitation of fluvial erosion. How are you doing this, btw?
Matt
Alright, working on your suggestion.
It is a combination of strata and outcrops, fractal warps on a low octave power fractal shader.
I increased the displacement tolerance to 2 and there is still some artifacts but not as bad as before. It did increase the render time by double i think.
It is interesting that you mention an imitation of fluvial erosion because the fractal warp shader seems to have some significant power behind it. It seems to warp a fractal along multiple axis. I see potential of this node being improved somehow and expanded upon.
I use it all the time, it's a great asset. Reminds me of the temporary erosion fractal.
What temporary erosion fractal?
Ulco is talking about alpha features again which he shouldn't talk about ;)
Matt
Quote from: Chris on August 25, 2015, 03:14:28 AM
It is interesting that you mention an imitation of fluvial erosion because the fractal warp shader seems to have some significant power behind it. It seems to warp a fractal along multiple axis. I see potential of this node being improved somehow and expanded upon.
The Fractal Warp Shader looks most like erosion when on steep slopes like this, but on shallow slopes the features aren't stretched enough in the direction of the downhill slope to look like fluvial erosion. That's where we need a more sophisticated algorithm.
Matt
Quote from: Matt on August 25, 2015, 06:41:19 PM
Ulco is talking about alpha features again which he shouldn't talk about ;)
Matt
I'm telling Mom. ;D
Quote from: Matt on August 25, 2015, 06:45:38 PM
Quote from: Chris on August 25, 2015, 03:14:28 AM
It is interesting that you mention an imitation of fluvial erosion because the fractal warp shader seems to have some significant power behind it. It seems to warp a fractal along multiple axis. I see potential of this node being improved somehow and expanded upon.
The Fractal Warp Shader looks most like erosion when on steep slopes like this, but on shallow slopes the features aren't stretched enough in the direction of the downhill slope to look like fluvial erosion. That's where we need a more sophisticated algorithm.
Matt
Alight, the description certainly makes sense based on how i observed it working. Sophisticated algorithm indeed.
Don't tell Mom :-\
Sorry Matt.
OK, I won't. ;D