Are .bmp files really that bad?

[Combs, Richard]

Dick Margulis wrote: 
 
> However, .bmp is limited to a 16-color palette. This may or 
> may not be problematic, depending on whether color is 
> involved in the first place and depending on what colors are 
> used in the interface.

Dick is usually such a reliable source of information, I'll write that
one off to not enough coffee. :-) 

A 16-color palette is 4-bit color depth. The term "palette" is
appropriate only for 4- or 8-bit (AKA indexed color) images, which use a
table, or palette, to define the 16 or 256 colors (24-bit colors) that
they contain, and pointers to the table entries to specify the pixel
colors. 

I have lots of 24-bit BMPs (16 *million* colors). I don't know if the
BMP format supports 30-bit or 36-bit color -- probably not, but who
cares? ;-) 

That said, BMPs are absurdly large compared to PNGs and have no
advantage over the latter for most purposes (the only exception I can
think of is the legacy WinHelp format, for which BMPs are the format of
choice). 

One caution with PNGs: Don't use 8-bit (indexed color) PNGs in FM. The
space saving over 24-bit PNGs is negligible, and each unique indexed
color in your PNGs gets added to FMs color definitions as an entry like
"RGB 000,155,233." Since each PNG contains 256 such definitions, and
they don't necessarily overlap, you can end up with thousands of color
definitions in your doc. 

HTH!
Richard


------
Richard G. Combs
Senior Technical Writer
Polycom, Inc.
richardDOTcombs AT polycomDOTcom
303-223-5111
------
rgcombs AT gmailDOTcom
303-777-0436
------