Actually some of it isn't inefficient, some of the models just have more surfaces.
Take for example, the bearcat and the confed banshee..
Both are similar sized (banshee is bulkier). The uvmapping area size is the same (i've tried to keep all fighters the same size). The bearcat takes two textures, banshee 4.
I think the amount of surface area plus the general geometry of that area determines how effectively I can work. long thin stripes and small pieces can be squeezed into tiny areas, whereas big huge chunks (like wings) tend to be more fussy (they either fit or simply don't) On some of the fighters there was just a small amount that just couldn't fit into any existing texture and using a second one would have been pointless, so i cheated and shrunk it down a bit. No sense in including a nth
texture and not even 1/4th used.
Also I've just noticed something else, the panels themselves have gotten smaller and more numerous
ugh and photobucket strikes again
Let me attach it to this thread directly.
[attachment deleted by ninja]