Factor of 8/9 expressed as 0.8 instead of 0.89

[cgilhuly]

Maybe there's another numerical factor involved that I'm not aware of, but I think there may be a small typo here:

mrf/mrf/sbcontrast.py

Line 276 in 1de806f

sig_adu = np.sqrt(biweight_midvariance(im_fluct_in)) * 0.80 # 8/9 is area correction

[AstroJacobLi]

I did some tests on this factor: generate an image with Gaussian noise, mask out some areas, calculate the SB-contrast using this code, and compare the result with the input Gaussian noise. Although an area correction factor 8/9 sounds more reasonable, I found that a correction factor of 0.80 works better empirically. Anyway, I don't think it will make a huge difference regarding the SB limit (-2.5 * log(0.8/0.89) = 0.1 mag). Any thoughts?