Finished a model and moved it to Poser and some normals were inverted. Fine. Opened up the file (this was a single vertex object with several named polygon sets), flipped the errant normals. Back to Poser to find something in some other corner of the model had REVERSED ITSELF. This went on for a while. I got them all in the right direction eventually (was a bit like working a Rubic's Cube) but in the end spent more time than I had in modelling in the first place! Anyone else have similar experiences? An easier solution?

Make polygon normals consistent? That's what I do; then they're either all right or all wrong.

On Poser import you mean? Oddly enough, that didn't change anything. There were a couple a n-gons involved; that might be the root of the problem.

I have this problem constantly (C2) with models undergoing significant mods. Rarely, errant normals are normal (pun intended) after the file is closed and reopened. Mostly seems a result of emptying and refilling polys. Other data structures, such as named polys, seem to have no impact. Mysterious and maddening. Jeff

Poser can't handle polygons with more than 4 sides.