Normal Map confusion

[aaouviz]

I'm just wanting a clarification:

When using a normal map (such as this one) what should I set the 'bump type' to? I get dramatically different results for each option. I'm looking for the most realistic, but expected result.

Cheers!

[aaouviz]

To elaborate further.

I've done a test.

1. "Bump"
2. "Normal map (tangent space)"
3. "Normal map (obect space)"
4. "Normal map (camera space)"
5. "Normal map (world space)"
6. "Bump map (local space)"
7. "Explicit normals"
X. No bump (diffuse only)

As you can see each option gives a different result. NONE of these seem to be what should be expected... at least not what I want. Why can't I get nice* bump without the diffuse getting dark? *Nice = #2, #3, #4, #5...

When I do get a bump without the diffuse colour change, the bump is too noisy, ie #6.

Why is the reflection also different between each option?

All of the surfaces have the same normals (pointing up).

Any help appreciated!!

[Peter.Chaushev]

Hello,

We understand the current Bump map UI setup could be confusing and have therefore already redesigned it. The newer simplistic UI will be available in the next upcoming version.
As for the current v3.40.03 setup - the correct option would be "Normal map (Tangent space)".

Kind regards,
Peter

[aaouviz]

Ok, understood. Thanks for the info.

However, I'm still worried/confused that adding the 'proper' normal map type makes my material a lot darker. Can I not avoid this, ie; have a normal map working correctly and the diffuse work as expected?

[Grigio18]

I recommend you to read this post from vlado...

https://forums.chaosgroup.com/forum/...s-gamma-or-not

I'll give you a brief summary:

In any case you should use: normal map (tangent space).

In most cases you should use the map in linear color space.(you can check it by looking at rgb values in photoshop)

check if the normal map has been created with the right channels: red channel has light coming from the left, green channel has light coming from above

Don't use values superior to 1.