[LivVis]

I had this before, found a solution somehow (back then the problem were tiled textures in maxwell material).

However, now I can't get rid of this squared looking surface while rendering.

Any ideas? Thanks!

[pipcleo]

looks like shadow error caused by bump map

[LivVis]

What do you mean? What would be the reason for it? A broken bump map?

[pipcleo]

a known issue.
just confirm by switching off the bump in the material

[LivVis]

Sorry, I don't get it. I need to use the bump.

So switching it off can't be the solution. What do you mean by "confirm"?

[Max]

subdivide more your geometry before applying bump

[Andreas Hopf]

Maybe a similar issue to this viewtopic.php?f=167&p=400120&sid=e27940 ... 5d#p400120 issue (bottom of thread, triangles showing with certain bump/normal maps). In my case, welded vertices and higher subdivisions solved the problem.

[Forum Moderator]

That's the typical "shadow terminator problem" in path tracer engines. Some things can be done from both the developers' and the users' sides to hide this issue but, if I'm not wrong, it cannot be completely removed because of the nature of the engine (Luis will know more than me about it).

From your side you can try this:

• Usually, a strong bump strengthens this effect, the softer the bump, the less noticeable. Displacement or a normal map can be an alternative.
• A strong and distant light like the sun makes it more visible too; usually, this is not seen with bigger and closer lights.
• Increasing the density of the mesh makes the effect less noticeable, like mentioned before.

[Nasok]

That looks like a low bpc (bits per channel) normal map.

Happens pretty often when you export you bump / normal map from software like Substance Painter / Substance Designer - in 8 bit (jpeg, png, etc)
- the same issue will happen in reflections (it just goes square patches)

The fix for that is to export your normal / bump map in 16 bits (at least). check out this thread - dude there had kind of a similar issue - viewtopic.php?f=146&t=45383&p=398153&hi ... it#p398153

Cheers!