I was finishing up the last piece of this school uniform set I've been working on, put it into game to see how it looks and... it didn't look anything like the CTU rendering. I tried fixing the texture linkages in MilkShape, I tried even extracting all the resources and moving it to a new clone package... still no dice.
The jacket should be nice and matte (fuzzy even, if the sim detail is set to very high), but it's coming out shiny with odd metal-stain looking patterns particularly visible on the sides of the jacket. This is occurring on the pants as well... and I didn't modify the pants at all from the vanilla mesh other than "sewing" them to the jacket at the waist!!
The female version of this outfit looks quite nice, so it's not (just) my computer sucking at graphics...
The problem looks like some kind of light reflection issue - it LOOKS like I'm getting artifacts of one of the models I derived my model from (Trace of a pocket texture???), but I have no idea how that's possible. I replaced all of the textures, I made sure they were linked correctly in the mesh comments; even CTU seems to think it's all dandy. It could potentially have something to do with normals, but I should think that would render in MilkShape and be quite obvious...
Fact of the matter is, beyond the specular texture, I haven't the faintest clue how lighting pattern is determined in game.
Of course, it could be a texture artifact that I missed. If I had any idea, I would not be asking.
Now, this is not my first CAS part - I have made a few somewhat elaborate ones before. Pretty much all of them were Frakensteined, I make sure to update all morphs as well. I guess what I'm saying is, I'm no pro modeler, but I have made a fair few outfits that work quite nicely.
What I'm pretty confident about...
1. The textures should be linked to the mesh. Even CTU seems to think so. If you wish to double check, I included the package for the blue uniform pictured above.
2. The specular looks like it should. At least, it looks like speculars of non-problematic CAS parts. (Quick picture included - dds file can be found in package)
3. It seems to be independent of what the container was cloned from. Now, the container clone (package) I use is almost never the same as any of the mesh clones I derive my new mesh from. This could be the issue, but it's never been a problem for me in the past. I put the resources in both a package cloned from the base game tuxedo, and a package cloned from the base game pajamas, and I have the exact same problem. And since this is occurring with two different parts, it isn't any kind of hash instance conflict, either. I would think.
I really think it has something to do with the mesh itself, but I don't know what or how to fix it.
Any insight would be fantastic ! I really don't want to have to start over this silly thing from scratch!
(I know there are some clipping problems with the morphs, and the trim and star pin overlays look gross. Want to fix this much more glaring problem first.)
(Disclaimer: The design of this uniform came from a manga, which is why it's so... saccharine. I mostly made this set because the elementary version is really cute and I decided to just go the whole way.)
EDIT: I do not use TSRW in any capacity. I do direct package editing. Clone with S3OC, replace morphs, meshes, and textures with mine in S3PE.