iraf.net

I am doing the step of the wavelength calibration of the data reduction.
In the manual (A User's Guide to Reducing Slit Spectra with IRAF) it is mentioned that the first step is the extract a comparison spectrum from the object.
I am not sure which comp goes with which object. These are the files that I have:comp3036.fits comp3040.fits comp3048.fits comp3052.fits comp3059.fits
comp3037.fits comp3041.fits comp3049.fits comp3053.fits comp3060.fits
comp3038.fits comp3042.fits comp3050.fits comp3054.fits comp3061.fitsobj3033.0001.fits obj3043.0001.fits obj3062.0001.fits
obj3034.0001.fits obj3046.0001.fits obj3067.0001.fits
obj3035.0001.fits obj3047.0001.fits
obj3039.0001.fits obj3051.0001.fitsAs you can see after observing each object I was taking 3 comps... do I match the first comp of each set with the first obj of each set? Do I need to use all comps and all objects in this step?

This is something of a judgement call based on the science. Because there are a variety of ways people have wanted to relate objects to comps there is the general task refspectra. One of the more common and probably the one to use if you are not sure is the nearest in time. The refspec task can do this for you, provided you identify a time keyword or ordering keyword, or you can do it manually with hedit or a table.So there is no right answer and I suggest the nearest in time.Frank Valdes

Thank you for your reply, I have 2 questions about this:1. So I don't need to use all comps in this step, correct?2. Do these comp images depend on the position of the telescope? The reason I am asking this is that the "night assistant" at the telescope used to remind me to take comps after observing each object before I move the telescope to the new position.

Does my question make sense?

For spectrographs attached to telescopes (as opposed to spectrographs coupled to telescopes by fibers) there is a thing called "flexture"---as you move around the sky, different parts of the instrument are tugged on differently by gravity, and there will be shifts in the wavelength as a function of pixel. Usually these shifts are small, at the 1 pixel or so level. You might beneift from a recent review of astronomical spectroscopy; see
http://adsabs.harvard.edu/abs/2010arXiv1010.5270M