wrote:

"Roger N. Clark (change username to rnclark)"
wrote:

In my testing of photoshop on real images, I find the following
equation:

PS = int(IP/2),

Are you sure it isn't "PS = int((IP+1)/2)"?

A minor danger is that if IP is 0xffff, the above will map to 0 if
done in 16b arithmetic. "IP 1" (M. Clarke's equation) is a single
instruction that doesn't suffer from overflow problems. But if wider
arithmetic is allowed, then:

PS = rint(IP*(32768.0/65535.0))

or its integer equivalent may be used.

Note that PhotoSlop is "just an image editor" (a large, complicated,
extensible, useful one to be sure), so precise stuff like you are
demanding will probably never make it high on the priority list at
Adobe, where most of their users are graphic artists, not by-the-bit
technician types. Can MaximDL and similar handle non-astronomical
imagery? It's internals are probably _alot_ more formal (linear
images, etc).