Illumination corrections
Elden Paul Iverson wrote on Feb 06, 2008
Mike-
What is the difference between mkskycor and mkillumcor and between mkskyflat and mkillumflat? The help files seem to say approximately the same thing.
Also how would an illumination frame be applied? My regular reduction steps are the following:
1. zerocombine
2. ccdproc dark frames with Zero.fits
3. darkcombine
4. ccdproc flat frames with Dark.fits
5. flatcombine (by subset)
6. ccdproc object frames with Flat*.fits
I have been re-reading Phillip Massey's reduction guide and he seems to apply the combined bias to ALL frames and then the combine dark to ALL frames etc. Wouldn't this essentially apply a bias frame multiple times through the combined bias and then later through the combined dark and flat (with a similar thing happening with the dark)? I could be wrong since this is simply how I was taught to do a reduction but I don't know.
Finally, if I make an illumination frame from my flats or objects, should I be using ccdproc with all of its parameters (i.e. have it use the Zero.fits,, Dark.fits, and Flat*.fits) or should I turn these off and make an illumination frame without allowing the task to call anything from ccdproc?
Thanks, none of my professors had a concrete answer for me so I am hoping you can help.
-Paul
What is the difference between mkskycor and mkillumcor and between mkskyflat and mkillumflat? The help files seem to say approximately the same thing.
Also how would an illumination frame be applied? My regular reduction steps are the following:
1. zerocombine
2. ccdproc dark frames with Zero.fits
3. darkcombine
4. ccdproc flat frames with Dark.fits
5. flatcombine (by subset)
6. ccdproc object frames with Flat*.fits
I have been re-reading Phillip Massey's reduction guide and he seems to apply the combined bias to ALL frames and then the combine dark to ALL frames etc. Wouldn't this essentially apply a bias frame multiple times through the combined bias and then later through the combined dark and flat (with a similar thing happening with the dark)? I could be wrong since this is simply how I was taught to do a reduction but I don't know.
Finally, if I make an illumination frame from my flats or objects, should I be using ccdproc with all of its parameters (i.e. have it use the Zero.fits,, Dark.fits, and Flat*.fits) or should I turn these off and make an illumination frame without allowing the task to call anything from ccdproc?
Thanks, none of my professors had a concrete answer for me so I am hoping you can help.
-Paul
Francisco Valdes wrote on Feb 06, 2008
Hi Paul,
I see you question is quite some time ago and so I appologize for not giving you a response.
mkskycor and mkillumcor are basically the same. The differences are minor and the two options are provided more for completeness than functional differences. I'm sorry if this actually makes it more confusing.
The way you apply a correction is really up to you and it doesn't really matter. The main difference is whether you want to things in multiple steps or not. While things were set up so you could create and apply both dome flats and illumination corrections in a single step I think it is more common for people to use ccdproc through dome flats. Then spend time on either dark sky flats or illumination corrections which are then applied in a second pass through ccdproc.
The distinction between dark sky flats and illumination correction (especially like mkskycor) is whether there is enough dithered data without large sources to determine corrections to the dome flat at the individual pixel level or just some smooth correction to flatten structure caused by illumination effects and sky color.
I hope this helps; if not for your early Feb work then for the future.
Yours,
Frank Valdes
I see you question is quite some time ago and so I appologize for not giving you a response.
mkskycor and mkillumcor are basically the same. The differences are minor and the two options are provided more for completeness than functional differences. I'm sorry if this actually makes it more confusing.
The way you apply a correction is really up to you and it doesn't really matter. The main difference is whether you want to things in multiple steps or not. While things were set up so you could create and apply both dome flats and illumination corrections in a single step I think it is more common for people to use ccdproc through dome flats. Then spend time on either dark sky flats or illumination corrections which are then applied in a second pass through ccdproc.
The distinction between dark sky flats and illumination correction (especially like mkskycor) is whether there is enough dithered data without large sources to determine corrections to the dome flat at the individual pixel level or just some smooth correction to flatten structure caused by illumination effects and sky color.
I hope this helps; if not for your early Feb work then for the future.
Yours,
Frank Valdes
Last post on Feb 06, 2008