Fed up with flats...

[Astroscot2]

Ive been out testing my newtonian and have tried taking a few flats which dont been to be working as per my expectations.

The camera is a SXVH694, Ive tried various exposures with average ADU values from 3000 - 24000 and everything inbetween - im finding that exposures resutling in an average ADU 3000 seem to work a better but are not solving all my issues i.e dust bunnies, as soon as I make a few tugs on curves the dust becomes an issue.

For those using this camera does an ADU of around 3000 seem right ?  should I be going lower ?  These are my first flats with this camera, ive taken flats many times before with various cameras and never had this issue.

Thanks

Mark

[swag72]

Your ADU for flats at 3000 is far too low. You do need a little experimentation to find the ideal, but mine is generally 21,000. I find that if I use flats with a low ADU then I introduce a lot of noise. 

[ollypenrice]

Your ADU for flats at 3000 is far too low. You do need a little experimentation to find the ideal, but mine is generally 21,000. I find that if I use flats with a low ADU then I introduce a lot of noise. 

I like the same kind of exposure as Sara. A couple of things come to mind, though. Are you dark-subtracting your flats? A master bias is all you need by way of a dark but if you don't do this you'll have the flats over-correcting.

You're sure no light is getting into the Newt during flats? The bottom end is a notorious source of leakage.

Olly

[Astroscot2]

Thanks both for geting back to me.

Sara I initially tried my flat with an ADU of 21000 and they were overcorrecing, I worked back and noticed the lower ADU average were producing better results 3000 was as low as I could get.

Olly - no im not dark subtracting my flats, should I ?. Olly can you explain how you would do this? do I need to take darks at the same exposure as the flats,  regarding bias,  I dont use these as a dark, should I?

Im using AA5 for calibration - Oly , can you show a screen of your typical calibration ?

I dont think im getting any light leakage from the Newt, the mirror cell is enclosed with fans .

Thank for your help

Mark

[Astroscot2]

Well I used flats with an ADU of around 22000 & my Bias frames as darks and dust bunnies no more.

Many thanks

[ollypenrice]

Thanks both for geting back to me.

Sara I initially tried my flat with an ADU of 21000 and they were overcorrecing, I worked back and noticed the lower ADU average were producing better results 3000 was as low as I could get.

Olly - no im not dark subtracting my flats, should I ?. Olly can you explain how you would do this? do I need to take darks at the same exposure as the flats,  regarding bias,  I dont use these as a dark, should I?

Im using AA5 for calibration - Oly , can you show a screen of your typical calibration ?

I dont think im getting any light leakage from the Newt, the mirror cell is enclosed with fans .

Thank for your help

Mark

I'm not at home tonight but will do you a screen shot when I am back. A master bias will make a perfectly good dark for any flat you are ever likely to take. No need to make special ones. It is vital to do this though.

Olly

[steppenwolf]

I'm with Olly and Sara, around 20,000 to 22,000 is what you should be aiming for. Olly's suggestion of using Bias frames as 'Darks' for your Flats makes really good sense as your flat frames will be very short so special darks for them will not be required.

[ollypenrice]

OK, I needed to make up this set of screen shots anyway to give to guests learning imaging here.

Firstly, how to make master dark and bias frames. AA5's preprocessing page 1 (below) is laid out on the assumption that you'll use virgin sets of new data for darks, bias, flats and lights. You are very unlikely to want to do this with a cooled CCD so it makes sense to prepare and keep masters. Here's how;

Once you have your master bias to use mainly as a dark-for-flats this is how to make master flats;

When you have both master darks and master flats you apply them like this. On page 2 choose Sigma if you have enough light subs (below 7 it probably isn't worth it) or median if you don't. For alignment choose Star Pattern Translation And Rotation and apply Normalize Background. On the last page you can experiment with the hot pixel and column repair if your camera needs it. Ours all do!

Finally there is an interesting alternative to darks which you can try in AA5. It works for me and is usually equivalent to, or better than, the classical dark method. It is also a time saver for me because the same calibration will work for images of different exposure times and since I find I don't need different flats for different filters it streamlines my mornings no end! Whether or not this will work for others I cannot say.

Olly

[Astroscot2]

Thanks Olly for the write up on the process, will give thix a try and confirm the outcome, once again many thanks.

Mark

[carastro]

Sorry to hijack thread, but seeing what Olly has put above.

I always put the Bias into the bias box and enable for flats.  Why do you put the master bias in the dark frames box?

Carole 

[ollypenrice]

Sorry to hijack thread, but seeing what Olly has put above.

I always put the Bias into the bias box and enable for flats.  Why do you put the master bias in the dark frames box?

Carole 

If you're asking about my screenshot called Making Master Flats then the Bias is only a bias in name. It isn't being used as a bias, it's being used as a dark for the flats. (At short exposures the difference between a bias and, say, a 1 second dark, is insignificant and shooting dedicated darks for flats really is a waste of time.)

If you're talking about my last screen shot, Carole, the one titled Using Bad Pixel Map Instead of Darks, then the Bad Pixel Map corrects the hot pixels and the remaining noise, which is mostly bias signal, is removed by a master bias treated as a dark. This system is outlined in the manual. I found it worked better than the normal dark frame method which sometimes does more harm than good for me. Darks can be quite invasive and counter productive sometimes. I think you need to see what works best for you.

When you put a bias in the Dark Frame Optimization box its job is to allow AA to scale the darks to match the lights. This should be a help to uncooled camera users who can never be sure that their darks are really at the same temperature as their lights. AA can use a bias to adjust badly matched darks to fit the lights.

This part of the programme has nothing to do with subtracting noise from your flats and won't do that.

Olly

[martin_h]

I use a similar method when stacking using DSS, I create a dark pixel map in Nebulosity and use it in Nebulosity to do its thing then drop the resulting frames into DSS as normal but in the Dark box I use a Bias...........it moans about it not being the same duration as the lights but I just ignore and carry on with the stacking.

[carastro]

Thanks for clearing that up Olly.  

Didn't notice the title bars.

Carole