PBR Lighting issue

[zzuegg]

To continue the discussion from:
https://hub.jmonkeyengine.org/t/zombiegirl-gltfloader-vs-monkeywrench/48007/15

Summary:

According to gltf specs, and on the reference viewers, the default value for metallic factor is 1. Jme's importers are setting the metallic factor to 0 to pass the test cases. The problematic line in my opinion is:

//223 of PBRLighting
vec4 diffuseColor = albedo - albedo * Metallic;

I have checked, a few other repositories, and i have not yet found that piece of math again. Additionally, the whole block:

float specular = 0.5;
float nonMetalSpec = 0.08 * specular;
vec4 specularColor = (nonMetalSpec - nonMetalSpec * Metallic) + albedo * Metallic;
vec4 diffuseColor = albedo - albedo * Metallic;
vec3 fZero = vec3(specular);

is quite unique. F0 is in all other shaders defined as:

vec3 F0 = vec3(0.04);
F0 = mix(F0, albedo, metallic);

[zzuegg]

When testing the MetalRoughSpheres jme shows also different results than the khronos sample viewer.

The reflexivity is lost completely in the second column in jme while in the khronos renderer it slowly fades

[stephengold]

Thanks for investigating.

When it comes time to solve this, instead of modifying the PBRLighting shaders (and potentially breaking many apps) I suggest defining a new PBR material with a new name.