3/05    How do 16-bit and 32-bit color schemes work?  Which bits do what?
        Thanks.
        \_ 16: generally, either 5-5-5 (1 bit unused or alpha channel),
           6-5-5, 5-6-5, or 5-5-6.  32: 8-8-8 (8 bits alpha).  If you
           have really expensive hardware, sometimes 10-10-10.  These
           are RGB implementations, obviously.  You can also use the
           16-bit value to reference a TrueColor (24/32 bits) palette.
           \_ Other strategies are separate 32-bit into 3 10-bit numbers,
              and then each number is an index into a different table to
              retrieve primary color value.  Also, you have to consider
              possible double buffering.  Not all systems use RGB either,
              your color model may be HSV.  There are equivalent matrices
              to convert one color model to another.  Also, note that
              there are perceivable colors not representable by a positive
              combination of RGB.
              possible double buffering and z-buffering.  Not all systems
              use RGB either, your color model may be HSV.  There are
              equivalent matrices to convert one color model to another.
              Also, note that there are perceivable colors not representable
              by a positive combination of RGB.
           \_ What is "alpha" then?  Thanks again.
              \_ It provides a way to combine (or blend) RGB values together
                 to achieve special effects such as transparency.  In normal
                 applications, pixel values are overwritten, but that does
                 not have to be.  You can combine old RGB value with new
                 RGB value and blends them together.