Stacking stacks with different flats

[cfpendock]

In this country it is rare to be able to image a complete run of images in one evening.  So for example, for luminance, I might get a couple of hours in one night and then nothing for a few days, then a few more, and so on.  Perhaps even three or four sessions just to complete the set for one filter.

 

Despite not taking the rig apart in the time between sessions, I invariably find that dust motes have moved, and so I normally take flats for each session.  When stacking I use the flats for each session to calibrate the appropriate lights.  So then I might end up with two or three stacked images for each filter.

 

What is the best method for stacking the calibrated stacked images into a single stacked image for subsequent processing, eg a simple average or what? 

 

I normally use DSS for calibration and stacking, but I guess the methodology would be the same for other software.

 

I should be very interested to hear how other imagers deal with this problem.

 

Chris

[alexbb]

With DSS, you can load the images to multiple tabs for each session and you can select for each session its flats. I'm not sure of the final result's quality if you have different quality sky conditions (ie. thin clouds, fog, etc. not the same for each session).
http://deepskystacker.free.fr/english/technical.htm#filegroup

My intuition tells me that choosing "Auto adaptive weighted" stacking method should handle better this, but it either hangs, either it doesn't produce any result. I stick with Kappa Sigma clipping. Be sure you check the background calibration for this:
http://deepskystacker.free.fr/english/technical.htm#Stacking

[johnrt]

This is easy in Pixinsight which is where I calibrate my data - just select which subs you want to calibrate, then the flat (and bias etc) you wish to use, let it run, then swap out for another night's light images and select a different flat, let it run again. You can then go on to align and combine after you have calibrated the batches taken needing different flats.

[cfpendock]

Thanks Alex, especialy for the links - I will do some experimenting with DSS.

Thanks John, sadly I do not have Pixinsight, although I am tempted by the idea of selecting the flat you want to use for particular subs.

Chris

[alexbb]

2 hours ago, cfpendock said:

Thanks Alex, especialy for the links - I will do some experimenting with DSS.

Thanks John, sadly I do not have Pixinsight, although I am tempted by the idea of selecting the flat you want to use for particular subs.

Chris

You can select a set of flats (or a master flat) for each group also in DSS. You don't need to specify which flats should be used for each sub, instead you should specify a group. It's described in my first link.

Just add a set of lights into DSS and then another tab at the bottom will appear (named "Group 1" if I remember correctly). There you will load the lights from another sessions together with their corresponding calibration files. If the temperature is ~the same, you can reuse the darks and the bias. If you don't provide calibration files for the second group, the ones provided in "Main" are used. Actually, if you select the same darks or bias or flats in another group, they won't even be shown in the group since they are already the default ones from the Main. After you load subs into "Group 1", "Group 2" becomes available and you can load there files from another session.

[cfpendock]

Just done it.  Brilliant. And it seems you can select a light form any group to be the reference image. Thanks very much Alex.

Chris

[alexbb]

Yes, indeed, you can select any light as the reference image.

You're welcome!

Clear skies,

Alex