GESD pixel rejection discussion and and TV Episode 1

Roger Smith

Hi Adam,

I re-watched Episode 1 discussion on GESD. In that discussion you talked about the probability that an outlier could occur on a pixel stack of 30 images. Your example making the case for a low value of ESD outliers talked about the low probability of 2 images both with a cosmic ray hit crossing on the same pixel. If nothing else wrong in that stack, then the fraction would certainly be low. That all makes sense. 

I have a bit different example....

In  the case of 30 images with 10 images with not so good seeing and exhibiting slightly bloated stars it seems that 1/3 of the images with possible outliers on a pixel stack pixel near the outside of the star. So I think more rounds of calculations by GESD would be warranted. What do you think?

I would like to run a test with my varying seeing I often have here. Hopefully I will do it soon.

Thanks,

    Roger