jMonkeyEngine.org

Are you sure you are using the lightning material deffinition?
Have a try in Material Editor in jmpThe jMonkeyPlatform, a full-fledged IDE with plugins for editing jME3 content and assets. Integral part of the jME3 SDK.? You can set the material parameters easily there.

The last one.Last upgrade i`ve made was yesterday. And yes, i use the right material. It doesn’t even if i copy paste all the code from Hello Materials tutorial

[ ] jME3_2011-12-06.zip 06-Dec-2011 00:09 81M – for eclipse
and yesterday i have auto-update the JMonkey IDE

I`ve change the mat specular color to white and there is no difference

Yeah, so the fix for the fix for the fix that wasn’t a fix for the nvidia problem is the issue.

We should probably just fix the shader for real this time.

Yeah, this history goes like this if I have it right:
-default shininess used to be 0 = no shine
-0 caused problems on nVidia that made weird black areas and other problems
-a fix was attempted to detect 0 shininess and not do the code that broke nvidia… but it used an “early return” instead of more strict structured programming and so broke certain ATI cards (not just bad visuals but shader would crash).
-so a default shininess of 1 was introduced but now _everyone_ who had been ignoring shininess suddenly had ugly visuals because their non-shiny things were suddenly really shiny
-so the default specular was changed to black.

Personally, I think we should go back to the problem with the “early return” and fix the early return… it was the best overall fix, I think. Though I could be remembering it incorrectly and don’t have time to be fixing shaders myself at the moment.

I’m possibly wrong but… I suppose the problem can be in the vertex shader:

   #ifdef MATERIAL_COLORS
      AmbientSum  = (m_Ambient  * g_AmbientLightColor).rgb;
      DiffuseSum  =  m_Diffuse  * lightColor;
      SpecularSum = (m_Specular * lightColor).rgb;
    #else
      AmbientSum  = vec3(0.2, 0.2, 0.2) * g_AmbientLightColor.rgb; // Default: ambient color is dark gray
      DiffuseSum  = lightColor;
      SpecularSum = vec3(0.0);
    #endif

When SpecularSum = 0, so it becomes nothing. And in the fragment shader:

vec4 SpecularSum2 = vec4(SpecularSum, 1.0);
gl_FragColor.rgb =  AmbientSum       * diffuseColor.rgb  +
                           DiffuseSum.rgb   * diffuseColor.rgb  * vec3(light.x) +
                           SpecularSum2.rgb * specularColor.rgb * vec3(light.y);

So, specular effect becomes 0 in the fragment shader. I would suggest to have always SpecularSum = (m_Specular * lightColor).rgb; .