What makes a good flat?

Ladd Lindsay

OK, so this is partially a acquisition question, but also relates to processing.  What makes a good flat?  I take my flats and do all the integration.  They look like they're good to me but they don't adequately calibrate my light frames, leaving dust donuts that I can't remove and impacting the rest of the processing.  Any general guidelines for the characteristics of a good flat?  I've tried getting an ADU of 22,000 and 30,000.  I generally reject any flat with an ADU below 0.2 or above 0.6 in pixInsight.

Thanks

- Ladd

Ladd Lindsay

Answering my own question here.  After further research, I discovered that my filter wheel had a "loose" piece of dust between the filters and the CCD that would move around as the wheel rotated.  So, that dust was in one spot during the flats and moved slightly during acquisition.  Taking everything apart and getting rid of the dust has resolved the specific problem I was encountering.

Adam, if there's a way to remove this "discussion", feel free.  Thanks

BlockHead

Flats are particularly sensitive to both hardware (acquisition) and software troubles.

You may have a perfect flat- but if the bias you subtract from it is not good...then the flat will not do its job.  All kinds of things can happen. Flats need to characterize the detector exactly- and the number one way to tell if they are "good" is to apply them. You cannot tell by looking at a flat on it own (in general) if it is good or not. A difference of 22K - 30K ADU will not make a big difference..and is very rarely the source of a flat issue. One diagnostic tool is to divide one flat by another taken at a different time (or under different conditions). This will show differences in illumination- it does not however, tell you which flat is better than the other. 

Anyway... just some ramblings on flats.

I will be releasing a video soon that shows an issue that I *thought* was related to flats...but really the issue was something else!

-adam