Request to preserve SAT mask in stacks

[Michael Wood-Vasey]

When stacks are made from input warps, it would be useful to preserve the reason pixels are masked from the warps. While this might not always be well-defined, or might conflicts, the particular case of recording saturate stars should lend itself well to this. In general stars that are saturated will be saturated in all of the input warps and this information could be carried through the to the stacks.

This came up particularly because my students we're hoping to blend in the center of the saturated stars and were trying to look for MASK=16 in the ref stacks, but found that instead the cores of bright stars were given as MASK=8 (blank), which is unfortunately degenerate with masked pixels in general missing data regions which one wants to be blank in the images.