2
$\begingroup$

I've been using Blender for a few years now and never given this much thought, but I just started learning more about different PBR workflows, and I'm realizing I don't understand the Specular input in the Principled BSDF.

In the past I've tended to adjust it to whatever looked the best for the specific material and scene, but that doesn't seem like a very accurate workflow, especially when using texture sets with multiple materials (eg. from Substance Painter).

Often I've left it at the default value of 0.5, but this seems to make certain materials too bright and plastic-looking, for example with some types of rubber.

For some projects I've also connected it to the inverse of the roughness input, but I can't imagine that's very accurate either.

So I'm wondering what IS the "correct" or most accurate method? Is there a specific value that it should always be set to, or should I be creating Specular maps to use with it?

And how would this translate to other software, like Maya or Houdini?

$\endgroup$
2
  • 1
    $\begingroup$ The most accurate way would be to not use a value to begin with. Use a texture. You will have a hard time finding a real life material that is uniform. Specular and roughness work in tandem, specular is how much reflective a surface is, while roughness is how scattered the specular reflection is. The values depend on the material, and you should usually leave it to the textures to handle it, if they are made correctly to begin with. The only number theory is that you rarely ever have 0 specular. Most objects that appear non-reflective are actually well reflective and very rough. $\endgroup$
    – Lauloque
    Apr 13, 2023 at 18:50
  • $\begingroup$ So for example if I'm using Substance Painter, I would create a specular channel and paint the correct values for each different material? It makes sense but I've never heard of anyone doing it before! After reading about it I guess often people just leave it at 0.5, which seems to be the most common value for materials. $\endgroup$ Apr 13, 2023 at 19:59

3 Answers 3

4
$\begingroup$

After reading more about it, I've come to this conclusion which I'm posting here in case it's useful to others.

Simple explanation:

Leaving Specular at 0.5 is usually acceptable, as the specular values on most materials are close enough that number that it doesn't make a big difference. However, if more accuracy is needed a specular map can be used with more precise values.

More technical explanation:

The specular slider in Blender shows values from 0-1, but those actually coincide with reflectivity values going from 0-0.8% reflective. This is because non-metal materials fall within this range, usually with values close to 0.4% reflective, which is why the default Specular setting of 0.5 works well for most materials.

Metals, on the other hand, have reflectivity values closer to 70-100% reflective. With Metallic values of 1, the principled BSDF ignores the specular input and creates these higher reflectivity values automatically - but it is also possible to set the Specular input to higher values and create a metallic material even with the Metallic input set to 0.

Specular/reflectivity values for various materials

Source for this info and image:

https://substance3d.adobe.com/tutorials/courses/the-pbr-guide-part-1 https://substance3d.adobe.com/tutorials/courses/the-pbr-guide-part-2

I'm not going to mark this as solved quite yet in case someone finds issues with this explanation and wants to correct me!

$\endgroup$
3
$\begingroup$

This is just about all we have to go on, but they do give a formula that can be used to find a value.

https://docs.blender.org/manual/en/latest/render/shader_nodes/shader/principled.html

$\endgroup$
1
$\begingroup$

In the past I've tended to adjust it to whatever looked the best for the specific material and scene, but that doesn't seem like a very accurate workflow...

You can trust data about materials you find online very rarely - only when you are making pure, clean, ideal materials. Unfortunately that's very rarely the case.

Adjusting material parameters while judging the look of the material is an accurate workflow. To make it accurate, you have to have good reference and use it while making the material. You also probably want to have a scene with lighting matching the reference while you work on the material(if PBR and scene referred workflow is what you are after). Neutral lighting(D65) is a good choice to make it easy to setup and judge. That's white in sRGB, also that's approximately the light outside during the day. You also need to understand what your color management settings do to your colors to know what difference to expect in your result and reference.

It's good to know what ranges the values should be in based on what the material is, but judging the result visually and comparing it to your reference data is probably the best way to get realistic/believable results in most situations. Adjusting material properties to whatever looks the best is still the main idea, you just have to consider the lighting of your scene as well and also if the values are physically possible and likely. The idea behind PBR is that same material should look good/believable in different scenes under different lighting. It might be a good idea to test different lighting while you are making the material.

$\endgroup$
1
  • $\begingroup$ It might be a good idea to test different lighting while you are making the material and that's where Blender's material view really shines with all its options to change lighting and reflections in a snap. $\endgroup$
    – Lauloque
    Apr 14, 2023 at 13:37

You must log in to answer this question.

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