My Flats - Artemis capture sanity check please?

[kirkster501]

First time I took flats last night with my CCD and my home made flats panel

Just to check my procedure please?

I took all my lights (well, 4 x LRGB of 120s each, it was 01:30AM and I had work today). Then I started on my flats. What I did was to create a new Artemis sequence. What I did was to try to get my flats to between 20000 and 23000 ADU as an average level of the whole flat (I read this is about the right level?). Since the brghtness of the flat panel is a constant, I have to adjust the exposures for each filter to get to that level of ADU. I was surprised at how small these exposure vaules were, between 0.022s and 0.08 (the longest being the luminance) also I noticed they were slightly different for each filter to get to between the ADU values mentioned. When I look at the flat afterwards they look correct at these exposure values.

So I set up a sequence in Artmemis, made sure the luminance channel is binned at 1x1 and the others at 2x2 (as per the lights) and fired off 10 each of flats of LRGB with a delay of 1s between each. I also kept the camera at the same temp the lights were made.

Appreciate a view if what I am doing sounds right please?

Generally speaking my workflow with CCD is taking shape. Its just so light until so late its difficult to get the data when I have to be up early and have no observatory.

Rgds,

Steve

[Freddie]

Sounds right to me, though you may want to take a few more. It's odd you have pretty much black in the top right corner. Do you use an OAG?

[kirkster501]

Thaks Freddie, no I don't. This is a ED80 scope with 0.85 FR/FF. Its not black but a definite darkening on all flats. So vignetting I think? Converting to jpg seems to make it darker though

Here is a light from the Red channel at 2x2 120s.

[carastro]

You don't need to do flats at the same temperature as the light frames, otherwise sounds OK.

Carole

[ollypenrice]

No, all is not well.

Firstly the luminance should, most obviously, be the shortest. The L filter admits about 3X the light of the R,G or B. The L should be the shortest. If it is the longest, you have a problem.

Secondly the flat you posted doesn't look right to me. You get a nose for these things and I may be wrong, but with a small chip I would expect a far more symmetrical illumination with a central bright region. What we see is a gradient. The bottom of the image does not have dark corners. When you see this, alarm bells should ring. On my own setups I have seen flats like this countless times and I know they are not right. Why they are not right is a different matter and not always obvious!! But they go in the bin.

I don't understand what you mean by 'an average level' in Artemis. What you should do is is look at the white point in the DISPLAY window and expose till that is around 20K. Maybe a bit lower, with experience. But 23K should be fine.

Of course, I don't know your setup but I sometimes do panel flats as you have done (nothing at all wrong with your procedure in principle) and when I get a result like that I know they are not going to work. I try again with a different system. The first thing I would do is slow your panel down big time with high tec typing paper...

Welcome to the dark side. You do everything right and get a wrong result. (Well, I think this is the case here.) Happens to me all the time!

Olly

[narrowbandpaul]

I don't know where the 20K ADU mantra comes from. The quality of your flats is dependent on the total number of electrons used to calibrate the flats. Ie the number of electrons per pixel times the total number of images taken. Surely you would want to maximise both terms by using more like 2/3rds of full well (~40K ADU). There might be non linearity effects but these should not even manifest themselves at 2/3rds full well. CCDs are very linear devices, it's one of their strong points. I can't see any reason to use less than 1/2 full well. I could of course be missing something, but this is my understanding of the flat fielding process from a more scientific point of view. Where does the 20K philosophy come from?

Hope that helps

Paul

[Russe]

I'm most interested in an answer to that question!!! Where does the 20K come from?