WBPP choosing a sigma for CC

Ian Hudson

I am just redoing some calibration frames (specifically darks) using NINA.  I have the files names set up so that they report the number of stars in order that I can (when taking light frames) delete very low numbers.   So.......  Taking darks, NINA was reporting stars (250-300) and I mused 'Is that a good target value for Cosmetic Correction ?

Not of dark frames !!!!   But if NINA is spotting about 300 stars which aren't there i.e. they must be hot / warm pixels ?  

Or have I had too much coffee this morning ?

BlockHead

Well... 250-300 is a very small ... extremely tiny population of pixels. Generally hot pixels number in the thousands. That being said, there is nothing wrong with this. 

You should run CC (the new version) with the default value of 10 and look in the log to see how many pixels were affected. A value of 10 is pretty aggressive and you will likely want to raise the value. (See my video on this for why making a *larger* number means less aggressive).

Keep in mind the main point of CC is to reduce the errors in measurements and registration that can happen with hot pixels in place. If they are not bothersome and you have *many* frames that are *all dithered*... you can be very conservative about CC (or even drop it).

-the Blockhead

Ian Hudson

Thank you Adam

Ian Hudson

Anyone reading this....... ignore my thesis.  Looking at a dark frame (which NINA reported as having 300 stars) there are thousands of hot pixels.