My overhead concern is, if the NSImage view image data part, already got/contains representation info about what kind of image type like 'PNG', 'GIF' e.t.c. it is. Or if that happens, adding this and compression ( if saved ) info, during saving time?
Then you have no need to worry. The generated PNG data or GIF data, already contains such info inside the data. As I already noted, no extra space when saving data.
Today I will start reading, again, NSIMageRep and the rest of the family and try to get that better in my head.
Mac OS progamming has a long history, and many still-useful documents and samples are marked as archived and even now, no direct replacements for them are provided... I hope Apple would listen to our voices.
Anyway, if you find some difficulty on using NSImageRep or something, there are many readers who can help you in the forums.