DSLR-friendly version of BatchPreprocessing script?

[M Covington]

I'm starting to use the BatchPreprocessing script for Canon DSLR work and wishing I had a slightly different version of the script with defaults more suited to the task.  Specifically:

"Add Bias" "Add Darks" "Add Flats" "Add Lights" should default to DSLR raw files only, not all file types, and should accept files even though there is nothing in them marking which kind of image they are.  (Right now I have to add all four kinds with "Add Custom.")

"CFA image" should be checked, or even presumed (doesn't need a checkbox).

Maybe a few other little things.

It would not be hard to make a modified edition of the BatchPreprocessing script, or maybe add a checkbox that would flip all of the defaults to be more DSLR-friendly.  Is anyone working on this?

Going further, I would wish for a calibration algorithm that uses lights, darks, flats, and flat darks, but not bias frames, since that is the way I'm doing it with DeepSkyStacker.

If I only had time!   

[dzso.bacsi]

Hi,

well, I am far from being an expert in PI, but

"Add Bias" "Add Darks" "Add Flats" "Add Lights" should default to DSLR raw files only, not all file types, and should accept files even though there is nothing in them marking which kind of image they are.  (Right now I have to add all four kinds with "Add Custom.")

No, you don't have to. Simply select the appropriate files when you "Add lights", etc. It will accept files even though there is nothing in them marking...
And why should it default to .cr2? Or .nef? Or .pef? Or ...? I use Master Bias (.xisf). Sometimes Master dark (.xifs). How would you do that with "DSLR default" file extensions?

"CFA image" should be checked, or even presumed (doesn't need a checkbox).

If you use Pure RAW settings, I think it is selected automatically... It is checked for me... (Format Explorer/DSLR_RAW/Edit preferences/Pure RAW)

It would not be hard to make a modified edition of the BatchPreprocessing script, or maybe add a checkbox that would flip all of the defaults to be more DSLR-friendly.  Is anyone working on this?

Well, do you? It would not be hard after all... 

Seriously, if you expect certain PI behaviour "since that is the way I'm doing it with xxx", good luck! 

Cheers

[M Covington]

Thanks for your input.  On trying again I find that "Add darks" "Add lights" etc. indeed work correctly.  I don't know why I had to use "Add Custom" earlier.  I'm going to correct my online notes.

I didn't mean for my initial message to sound like a complaint.  The BatchPreprocessing script is very handy and serves PixInsight's primary goal very well, which is to be versatile.  I just found myself wishing for a slightly different script and thinking about making one.

As for wanting to use flat darks because that's the way I do it in DeepSkyStacker, the idea behind that is that

(Lights - Darks) / (Flats - Flat Darks)

is an exact calibration, whereas

((Lights - Bias) - (Darks - Bias)) / (Flats - Bias)

is only an approximate calibration because it assumes that the flats contain nothing but signal and bias.  This is close enough to the truth to be practically indistinguishable, but it is an approximation.  See http://www.covingtoninnovations.com/dslr/newdslr/#Arithmetic.

[dzso.bacsi]

Oh, I am really sorry, I did not read it as a complaint.
It just struck me, that meanwhile I have your site bookmarked and know your very useful work there, your description and hence the "request" was totally against my experience with BPP behaviour.

As for the calibration process: I leave it to much more knowledgeable people like you. What I found, that there is no 100% consensus on flat-darks or even "proper" calibration steps with DSLR. However, based on my own experience, the BPP calibration - or recently for me the manual way - leaves me with very good results. It yields definitely better results than the DSS process I used before PI.

If we only had time... 
     

[M Covington]

Yes.  It works very well as it is.  I may come up with my own modified version of the BatchPreProcessing script in the future.

One thing I've noticed about digital image processing is that people often come up with complicated processes that are mathematically equivalent to much simpler processes.   The script does not have that problem, but some of the folklore about how to use PixInsight does.

Is there any (copyright or proprietary) objection to my making a modified BatchPreProcessing script, acknowledging the original authors and adding comments to clarify what I changed, and then sharing it with others?

[M Covington]

Actually, looking back at my 2 formulas, if I simply use flat darks in place of bias frames, everything comes out exactly right... without a change to the script.