Counts are too High Compared to Linux IRAF and 3 Images in a Single Fits File about ccdproc HOT 4 OPEN

Thanks for the detailed report @kjkoeller. One question I can answer right away. I think that ccd_process creates, by default, an error image and a mask image for each output image. That is causing most of the size difference, too.

The other factor that affects size is what precision the results are saved with. Python defaults to 64-bit float which much more resolution than any CCD images actually need. I'm pretty sure IRAF defaults to 32-bit floats.

What settings did you use when you combined the images in IRAF? Was it a median combine, or mean, did you do any clipping, etc?

from ccdproc.

That is fine @mwcraig , but should that mark and error image be outputted with the new image? It might make it hard to do other photometry analysis on the images if there is 3 images imbedded in a single file. Also the file size is now enormous so having hundreds of images where each image triples or quadruples could be an issue as well.

The combination is a median and there is a high sigma clipping of 5 or higher.

from ccdproc.

I am a little confused as to why the error image and pixel mask are even created in the ccd_process to begin with, as they are never used again. From what I can tell it would be better to either comment these lines out or have them be a separate return.

https://github.com/astropy/ccdproc/blob/main/ccdproc/core.py#L217

from ccdproc.

Any updates?

from ccdproc.