3
$\begingroup$

I want to achive this result for my column without using additional modeling (just shading and procedural textures, sorry for bad quality, I don't have better one):

result

For now I have this material (the top part for contour and the bottom part for spiral): material But I have a two problems:

  1. for now it looks too flat even with the bump node connected to normals;
  2. I can't find a math function to make the spiral to look more realistic: it should have not the straight angle, but something like this in every coil:

enter image description here Can I solve that problems with math, or I should use something else? Does exist a best way to achive the final result with variations, because I think it's too much for me to use math for shapes like this (without additional geometry):

enter image description here

enter image description here

$\endgroup$
4
  • $\begingroup$ What's the end-purpose of your models? e.g. Lo-poly for games.. expensive for a scenic still,,,, $\endgroup$
    – Robin Betts
    Aug 30, 2022 at 8:51
  • $\begingroup$ @RobinBetts do you mean math is expensive for scenic? I want to create scene in Blender and than recreate it in ue5. $\endgroup$ Aug 30, 2022 at 11:58
  • $\begingroup$ If for a game, you will probably want to bake to textures anyhow? So whether you make this with shader displacement or geometry is up to you, whatever you find easiest. $\endgroup$
    – Robin Betts
    Aug 30, 2022 at 12:00
  • $\begingroup$ @RobinBetts yes, you're right, but I didn't think so far ahead :) $\endgroup$ Aug 30, 2022 at 12:25

1 Answer 1

4
$\begingroup$

Here's one route to barley-sugar moulding:

enter image description here

.. here, including the edge-piping, although, for generality, that's probably better stripped out as a separate group. This starts out assuming your panel is UV mapped, 0-1 in U.

The strategy is distance-to-0 (a circle), but in sheared UV cells. (Giving tilted ellipses.) The distances in that masked section and the outer piping are then mapped using color-ramps to give rounded height-profiles.

This is the sort of result when used as displacement:

enter image description here

The other patterns you refer to are also possible, but really a separate question? Different answers may come up with different approaches.

$\endgroup$
15
  • $\begingroup$ Sorry, but I can't recreate that material. My displacement node output looks the same as your but with principled bsdf connected to material output I can't see anything. I checked it 3 times and the only difference I can find is in the material displacement settings. There is a "reroute" option I can't find. $\endgroup$ Aug 30, 2022 at 11:52
  • $\begingroup$ Maybe share a problematic sample on blend-exchange.com ? I'll have a look.. it could be so many things.. :) $\endgroup$
    – Robin Betts
    Aug 30, 2022 at 11:57
  • $\begingroup$ here, I upload file (blend-exchange.com/b/m7E64lLG). Strange thing, I tried to create test file and recreated material - everything works fine. But in my base file it doesn't work. $\endgroup$ Aug 30, 2022 at 12:24
  • $\begingroup$ @Сергей The differences are numerous: Your material is set to Bump Only, Robin's is set to Displacement Only. Your plane has no geometry to be displaced, it's only the four corner vertices. But that wouldn't matter if you would use Adaptive Subdivision, but your model has no Subdivision Surface modifier. Even then it wouldn't work, because your settings use the Feature Set "Supported" while Robin uses "Experimental". To be fair, that's too much for a beginner to realize from the answer. I would recommend searching for tutorials on "microdisplacement" for the basic understanding. $\endgroup$ Aug 30, 2022 at 13:38
  • $\begingroup$ @GordonBrinkmann thank you, I've tried to change "Bump Only" to "bump and Displacement" or "displacement only" without result anyway. I forgot about needed geometry for displacement. So, it's almost correct answer, but I asked about solution without additional geometry. But answer is still cool because we just need to replace displacement node with bump node and connect normals to principled bsdf. And change settings to "bump only". That's all. So that's still right answer with that note. $\endgroup$ Aug 30, 2022 at 15:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .