No dark calibration=vignetting?

[BrendanC]

Hi all,

As a test, I just decided to try processing an image without darks. This was after processing one from quite some time ago before I started using them, in which I was surprised at how noise-free the image was.

What I got was not what I was expecting - really severe vignetting right around the central object.

I'm confused. I thought that's what flats were supposed to correct? Which I used. And dark flats.

I didn't think darks had anything to do with this - just removing sensor and read noise.

Or am I wrong? Again?

Thanks, Brendan

Post edit -  I guess this should be in the image processing section, apologies, mods please feel free to move 

Edited April 8, 2021 by BrendanC

[vlaiv]

Yep, not doing dark calibration means messed up flat calibration.

Calibration is not about noise removal - it is about signal removal. You don't remove signal - well, odd things happen.

Flats correct vignetting that is result of light obstruction - it only affects light coming in from the front of the telescope. If your image contains both light signal and dark signal - flats try to correct both, but since dark signal is not affected by obstruction as it is not coming from the front of the scope - you are correcting thing that is not messed up - and you end up messing it up - if you understand what I mean.

In math terms - it is a bit like this:

Pure light case:

light * 0.7 / 0.7 = light

(light that has been reduced to 70% of its original value - gets divided with 70% flat  and you end up with original 100% light)

Now imagine you have following scenario:

(light * 0.7 + dark) / 0.7 = light + dark / 0.7

since you did not remove dark - it ended up in final image - but "corrected" for vignetting that it does not have - so it get corrected "to the other side" - and that is again vignetting.

[BrendanC]

Thank you Vlaiv. Once again, you give the definitive answer!