so im having an issue where for some reason the render output is chopping off a peek of one of my displacements, no matter what i do.
is this a bug or im doing something wrong?
You need to raise the displacement tolerance in the Planet node. This will effect render time though.
oh my... why there are so many options scattered all around in this software... who would ever thought of that.
thanks Was.
is there any other way? im confused why this is happening in the firt place since the first pillar has way more displacement and its rendered fine.
i added other 2 pillars and they basically are all chopped off. I have to increase the displacement tolerance to 4 and the render time is simply absurd for a simple shape like this..
Yeah I'm not sure. That's what I had to do for my geonosis landscape and it was so bad render wise, just to get a few peaks to render.
Maybe patch size also influences where sampling is occuring? Probably not.
Would be cool if this was more dynamic. Like, a setting of an individual displacement node, so it can have better tolerance, and not affect other displacement nodes. Seems with doing it via the [planet] node it is applied to all displacement on the object. Instead of just the displacement driving what needs it.
would that work if it was a Compute Terrain instead of a Compute Normal?
They both have the same patch size setting and produce the same patching. Compute Normal just doesn't have texture coordinates. More meant just for displacement I gather.
No need to change the Displacement Tolerance at all here.
Simply set both the Simple Shape Shaders to Final Position.
That`s all.
CHeers, Klaus
Quote from: KlausK on May 13, 2021, 06:11:51 PMNo need to change the Displacement Tolerance at all here.
Simply set both the Simple Shape Shaders to Final Position.
That`s all.
CHeers, Klaus
That doesn't work for Power Fractal based terrains. Not even sure if it would work for simple shapes either, but if it does that's good.
Yeah, no. It doesn't work for Final Position for colour/disp of simple shapes, or for world space/final position for a PF
Well, I did only use his scene file. And in his setup it brings back the tops.
CHeers, Klaus
It probably changes where the position peaks end in world space, is why. The issue is actually related to too much displacement, where less would yield a fully rendered spike but look more or less the same.
hm.. :-\
That (noise space and disp) and whatever is going on with the underlying geometry. Which is why I wondered if patch size may influence it, but doesn't seem to be from my tests. I don't know if exported polys represent what the planet uses during a render, but they're not very uniform.
tried with final position, it doesnt work either. im actually shocked you can not simply duplicate a simple shape displaced with a mixer and you get this issue.. i tried in this example a surface layer, but even with mix shader when i increase the displacement of the second pillar to match the first pillar height, it chops it off.
so i painted a mask in photoshop and tried the displacement from that image, at the same height of the previous file displacement which was (60) it didnt chop anything off.
I raised the displacement from the photoshop mask to 1400 and the cutting started to appear in a random way, this seems more like a bug.
If it was a displacement tolerance issue it should have done it for every peek? why does it randomly on some peek yes and some no?
Pretty confusing
i also tried to remove the duplicated shape, and do one single pillar, increasing the displacement to over 1700 and it works just fine.
So once you add more shapes it bugs out. This is why i think its a bug, because it doesnt have anything to do with the displacement height, i raised it up to 10000 in height and it was rendering fine.
yeah, this doesnt make any sense. I duplicated 4 of them and it chops off just one.. some werid thing going on.
Displacement Tolerance seems like the tolerance of displacement before wonky things happen. So 2 mau be good for some spires but not others.
Notice in your test more disp causes the issue. I think this has to do with a combination noise space, displacement exceeding it, and where it falls in polys of the planet.
While I don't understand why this is happening, I found that turning off the Microvertex Jittering eliminates the sharp cutoff on the second displacement. See attached image.
The documentation for this parameter which is located under the Renderers Subdiv Settings states "This setting is important to reduce the appearance of straight lines in some displaced surfaces. For still images we recommend you keep this turned ON. For animations, OFF seems to produce a more stable animation, but ON sometimes reduces render times by a small amount."