Bias vs Super Bias re. Hot Pixels

[HomerPepsi]

Hi,

I guess I just need some clarification on bias as a whole here. When calibrating do we want to subtract the hot pixels that are in the bias frames from all frames (darks, flats, lights)? Or is it just to subtract the underlying sensor pattern?

Super Bias is obviously a much cleaner bias showing only the columns (in my case) and horizontal banding patterns. When using it, there are no hot pixels to be subtracted from the other frames.

[Carlos Milovic]

Hot and cold pixels are better handled with a dark frame. They do not beheave as a fixed value input, but rather they are non-linear pixels. So, I would recommend just using dark frames of the same exposure times as the lights, if hot pixels are your main concern.

[HomerPepsi]

Hmm maybe I was not so clear.

My question is what happens with the hot pixels that a present in the bias? My normal bias has hot pixels. My super bias does not. If I am calibrating darks, flats and lights with a bias that has hot pixels, are they subtracted?

[dnault42]

I think what you are referring to are 'stuck' pixels.  These are pixels that are always at minimum or maximum ADU.  These will never properly calibrate even with dark frames.  You need to either rely on dithering and rejection algorithms to remove these or use the CosmeticCorrection or DefectMap tool to replace these pixels with data from the pixels around them.

Hot and cold pixels either fill their charge wells significantly faster or slower than the norm and as long as they don't reach the max ADU dark frames will calibrated them properly.

Regards,
David

[HomerPepsi]

Thanks for the insight, David!