Author Topic: Bump texture rendering too smooth  (Read 3404 times)

2022-06-29, 02:12:54

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
Take a look at this example, why is this procedural texture rendering too smooth on the bump channel? That's not working as expected. If you connect it into the displacement channel, it is rendering properly.
We should  have a pretty close representation of this texture on both channels.
Is there some parameter I'm missing in order to make it work properly?

2022-06-29, 02:17:12
Reply #1

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
Also, I've noticed some weird projection going on. Take a closer look.

2022-06-29, 10:10:00
Reply #2

romullus

  • Global Moderator
  • Active Users
  • ****
  • Posts: 8779
  • Let's move this topic, shall we?
    • View Profile
    • My Models
I think the issue is that your noise texture has too much contrast in it. Both, displacement and bump loves smooth transitions between high and low values, but the former can still produce usable, albeit sub-optimal results with super contrasty textures, while the latter can't do that. It's the nature of bump mapping.

As for the projection issue, i think it's your mapping at fault, not Corona. It looks that currently your object has simple planar mapping. Give the cookie proper UV unwrap and it should render just fine.
I'm not Corona Team member. Everything i say, is my personal opinion only.
My Models | My Videos | My Pictures

2022-06-29, 10:20:42
Reply #3

ficdogg

  • Active Users
  • **
  • Posts: 112
    • View Profile
You can also try reducing the Delta value in the noise. For high-frequency noise bumps, a Delta value at the lower end sometimes makes a huge difference.

2022-06-29, 12:47:37
Reply #4

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
I think the issue is that your noise texture has too much contrast in it. Both, displacement and bump loves smooth transitions between high and low values, but the former can still produce usable, albeit sub-optimal results with super contrasty textures, while the latter can't do that. It's the nature of bump mapping.

As for the projection issue, i think it's your mapping at fault, not Corona. It looks that currently your object has simple planar mapping. Give the cookie proper UV unwrap and it should render just fine.

Hey thanks for you time. It seems to be none of them, sorry. The Delta parameter is responsible for the bump details, I didn't know about it. Credits to ficdogg.

For the projection issue, it doesn't make any sense to have UVs for procedural textures, as far as I know, that's the advantage of them. Material tags should be ignored when using texture projection in the space parameter. At least that's what the docs say. Also, the projection is working perfectly for the displacement. Some weird  behavior happens to the bump channel once we are using the same projection for both.

Cheers

2022-06-30, 01:39:09
Reply #5

burnin

  • Active Users
  • **
  • Posts: 1532
    • View Profile
For perfectionist, just bump is just a bitter tasting fake ;) so maybe corona/Normal shader can help you a bit more, since it takes light direction into account.
btwdo you have any reference you need to comply to or just cruising?

here's continuation w/o displacement and reference as I imagine it to be good enough, idk...

2022-06-30, 19:26:10
Reply #6

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
Yes I agree, I'm going to experiment with the normal shader, I didn't know I could connect procedural textures to it. These examples that I provided were something random and simple with the purpose of learnig the basic principles, now I'm going to try to create the material for real.
I really appreciate your help.

2022-06-30, 23:56:18
Reply #7

Cinemike

  • Active Users
  • **
  • Posts: 1000
    • View Profile
Remember that the Normalizer shader (which you need, there is no point in plugging the Noise directly into the Normal shader) needs a Noise in 2D (UV) space, so you lose your 3D shader.

2022-07-03, 18:27:53
Reply #8

prince_jr

  • Active Users
  • **
  • Posts: 83
    • View Profile
there is another point i wanna bring in...
at the moment i'm struggeling with bump and displacement in case of "uvw mapping" and "box mapping". when using uvw, the bump or displacement seems to be fine. changing the same material to "box mapping" seems to make the bump or displacement softer or even inexistent.
> please see the 2 pictures attached
> i'm using corona 8 hotfix 1, c4d r23.110

are you aware of this? is it a bug or just a user related problem?
thanks for all your inputs.

2022-07-03, 20:29:40
Reply #9

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
Would be very nice to have a file to look into. Try to reproduce it with a simple object.

2022-07-04, 18:37:17
Reply #10

prince_jr

  • Active Users
  • **
  • Posts: 83
    • View Profile
here it is...

2022-07-04, 23:31:59
Reply #11

burnin

  • Active Users
  • **
  • Posts: 1532
    • View Profile
Interesting "test scene"...
Open Project Asset Inspector, select (missing) files, (RC), "Globalize file paths" instant crash & freeze...
wt?

2022-07-06, 16:02:46
Reply #12

Mac3DX

  • Active Users
  • **
  • Posts: 147
    • View Profile
I'm pretty frustrated with using bump in C4D, what should be easy is turning into hell. When we have UVs everything works perfectly, but in projection modes it is very problematic or should be simpler if I'm not setting it up properly.

2022-07-07, 17:23:59
Reply #13

prince_jr

  • Active Users
  • **
  • Posts: 83
    • View Profile
@Mac3DX me, too.

i'm working on a grip with leather material > see attached images.
can't set up the material properly with bump/displacement, reflection etc. it looks like plastic :o(
> projection is boxmapping. and triplanar is doing a strange job here, too!? also tried to unwrap the mesh, bring it in photoshop etc. but it was a mess.

maybe someone can help me creating a realisitic leather material, which is circulating the grip naturally/seamlessly (like the example shows)
thanks in advance!

2022-07-07, 17:31:09
Reply #14

TomG

  • Administrator
  • Active Users
  • *****
  • Posts: 5434
    • View Profile
Not sure I understand here - box mapping would be wrong for this and give the results you show, cylindrical mapping would be the choice surely?
Tom Grimes | chaos-corona.com
Product Manager | contact us