ProcessVertices()...

Is the ProcessVertices() function slow? I ask becuase of the following: I want to speed up the dreadfully slow X-file code that MS has provided (and used in Wolfgang Engel''s tutorial) by adding vertex buffers. It''s no problem to make the source buffer, but the destination buffer is a whole other ball game... If I make a call to ProcessVertices(), it only lets me do a range of vertices to transform, light, etc. I want to transform vertices in the same order that they are listed in the index list. I could do it as such: for each polygon in the index list { ProcessVertices(curr polygon); } but is this really, really slow? Does ProcessVertices() call a Lock() on the source or destination vertex buffers? Has anyone else really been using VBs? Vyvyan