BatchPreProcessing Script and DSLR Images

[twade]

To all,

I finally got a clear night and captured some images of the Rho Ophiuchi region.  I used MaximDL to acquire my images which saves them as CFA.  For some odd reason, the script seems to be saving the file as an RGB file without color.  For example, the original CFA files sizes are 41,305 KB.  When the script saves a master BIAS, it grows to 247,801 KB.  Even when saving a 16-bit CFA to 32-bit, the resulting file size should only be 82,604 KB.  I've tried the script with "CFA Images" checked and unchecked -- both instances result in the same behavior.  Perhaps, I'm using the script in a way it wasn't designed.  Basically, I did the following:

Step 1: Selected all my Bias files and 480s Darks --> the script generated master files (master Bias and master 480s Dark)
Step 2: Selected my Master Bias and all my 5s Darks --> the script created master 5s Dark
Step 3: Selected my Master Bias, Master 5s Dark, and selected all my 5s Flats --> the script created master 5s Flat
Step 4: Selected my Master Bias, Master 480s Dark, Master 5s Flat, and all my Light frames --> the script generated both CFA calibrated lights and RGB (debayered)

While using master files, I did select the appropriate checkbox in the script dialog.  What's intereresting is the Light CFA calibrated files are the proper size, 82,604 KB.

Does anybody know what's going on here?  Does it really matter outside the wasting of a large amount of disk space?  I would much prefer the generated masters remain CFA since they are significantly smaller.

Thanks,

Wade

[twade]

To all,

I figured it out.  The files are so big because they are a combination of three files: the master file, rejection_low, and rejection_high.  I'll just turn these options off from now on.

Wade

[Juan Conejero]

Hi Wade,

Please note that the rejection maps are important because they allow you to check if the pixel rejection routines applied during master integration have worked well. There has been some discussion about this on the BatchPreprocessing script development thread, and we'll probably add an improved "Generate rejection maps" option in a future version of the script, so this option will be selectable separately for bias, dark, flat and light frames.

Besides this, you don't need to run the script four times. Just add your whole data set and run the script once; it will calibrate everything and generate all your master frames and final light frames in a single execution. Don't forget to save a process icon from the script.

[twade]

Juan,

Please note that the rejection maps are important because they allow you to check if the pixel rejection routines applied during master integration have worked well.

I fully agree.  I generally watch the process in the Process Console looking for too much rejection.  The defaults generally work pretty good.  Perhaps, the rejection files should be separated (i.e not embedded within the primary file).  This way, they can be deleted once they have been looked at.

Besides this, you don't need to run the script four times.

I'm in the experimental stage.   Once I get more familiar with the script, I'll let it do everything at once.

I really appreciate your help and expertise,

Wade