Announcement

Collapse
No announcement yet.

Static Mesh Specular Highlight Problems, But Normals Seem Fine

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Static Mesh Specular Highlight Problems, But Normals Seem Fine

    [EDIT: This problem has been solved by importing using binary .FBX rather than .ASE.]

    I made a fairly large and complex static model in 3ds Max with two materials and imported it into UDK. (I used a lot of Booleans, and used mirroring to symmetrize the model.) I don't see any problem with the model in 3ds Max, but when I look at it in UDK, it has some places where the specular highlights don't look right. It is as though the normals have a problem, but whenever I look at the normals in either 3ds Max or UDK, I don't recognize any problems. The model has no normal map applied. I have checked the smoothing groups in 3ds Max; I'm pretty sure all polys have them, and all of them appear to be assigned appropriately. I have checked for double faces, double edges, double points and micro-faces. I have checked the polygon normals in 3ds Max, and none seem to be tweaked or inverted. Is it possible that this could be a UV mapping issue? If anyone could give me any clues to what's wrong, I would much appreciate it.

    Here's one area with the problem. I am mainly concerned with the issue on the dark rim, but the facets on the tube aren't too hot either:



    Here are the UDK normals (green), binormals (blue) and tangents (red) for that area.



    Here is the geometry in 3ds Max (realistic lighting):



    Thank you in advance for any help.

    #2
    That could more be due to the UVMapping, are there seams there?

    Comment


      #3
      The dark circular rims are a separate island from the grey parts and the long, dark, straight supports. But there are no seams between the improperly highlighted polys and the normally highlighted ones on the circular rim.

      Comment


        #4
        I finally found the problem: I was importing the model into UDK using the ASCII (.ASE) format as a certain YouTube tutorial recommended. When, instead, I imported using binary FBX, the problem disappeared.



        So I guess the lesson for me here is: Don't import models using .ASE.

        Comment

        Working...
        X