Frequently Asked Questions
The routine spec1d_ascii.pro will take a list of spec1d FITS files as it argument and return ascii files containing the flux, inverse variance, and wavelength for the 1-d object spectrum. More information regarding this routine can be found by reading the documentation:
IDL> doc_library, 'spec1d_ascii
(2) How do I determine what arc lamps were turned for a given arc frame?
This can be done by simply examining the header of the arc frame as follows (where d1003_0035.fits is the arc filename):
IDL> hdr = headfits('d1003_0045.fits')
IDL> lamps = sxpar(hdr, 'LAMPS')
IDL> print, lamps
(3) How long does it take for the pipeline to reduce a single DEIMOS mask?
Running on a machine with a 1.2 GHz processor and 2 GB of RAM, it takes roughly 6 hours to reduce a single DEIMOS mask with 3 science frames and ~125 slitlets.
(4) Are internal flats preferable to dome flats for working with the pipeline?
Yes, for working with the pipeline, internal flats are superior to dome flats. The scattered light issues present in internals are also present at non-negligible levels in dome flats. Internal flats, however, provide a higher signal-to-noise level which is useful in correcting such illumination issues.
(5) How do I view the 1-d object spectra?
Using the routine fill_gap.pro, spec1d files can be accessed easily. The syntax for the the fill_gap routine goes as
IDL> ss = fill_gap('spec1d.xxxx.nnn.ooooo.fits', /keyword)
where the /keyword specifies which extraction the user wishes to pull from the spec1d file (e.g. /boxsprof, /boxcar, /horne, or /optimal). Note that presently the pipeline employs the /boxsprof and /horne extraction algorithms by default. To see which type of extractions are present in a spec1d file, use the GODDARD routine fits_help:
IDL> fits_help, 'spec1d.xxxx.nnn.ooooo.fits'
This will print to the display the contents of the spec1d file. The output from the fill_gap routine will be an IDL structure containing tags indentical to those of the spec1d extensions (see the spec2d primer for details on the format of spec1d files). To simply plot the spectrum, do the following:
IDL> splot, ss.lambda, ss.spec
(6) How is exposure time handled in the pipeline?
Very early in the pipeline the counts in the science frames are scaled to counts per hour. This is done before sky-subtraction and before combining the science frames. The frames are combined using an inverse-variance weighted algorithm so as to properly consider variations in exposure time between science frames.
(7) What arc lamp species are recommended for use with the pipeline?
Our tests suggest that for DEEP2-like settings (7800Å central wavelength, 1200-line/mmm grating), the best results are obtained from a 1-2 second exposure with the Ne, Ar, Kr, and Xe lamps on. This gives a good number of strong lines on both the blue and red sides of the detector.
For 600-line data, the same combination of lamps (Ne Ar Kr Xe) produces good results; Cd, Hg, and/or Zn could be added to yield additional lines at the bluest wavelengths, but these are not absolutely necessary for a quality wavelength solution. Note that Cd Hg Ne Xe Zn yields few lines towards the red, and thus is likely to be inadequate for the current version of the pipeline.
Regardless, please be sure that the Ne lamp is warmed up and yielding strong flux; it has been intermittent at times, but is critical to the wavelength solutions.
(8) Will the spec2d pipeline reduce data taken with the DEIMOS long-slit?
There are several long-slit DEIMOS slitmasks for general observing use. These slitmasks have ~12 slitlets placed along the mask so as to form a single long-slit (while also maintaining a structurally sound slitmask). The spec2d pipeline has been tested on several sets of observations using various long-slit masks and shown to be successful at producing reduced spectra.
(9) Is the spec2d pipeline affected by the DEIMOS FITS header DEIMOS glitch? (for a detailed explanation of the aforementioned glitch, see the DEIMOS NEWS website (refer to the 2003-Nov-21 update).
The GODDARD routines employed in the pipeline to read FITS files handle the problematic keywords correctly. This glitch in the DEIMOS files should NOT affect any pipeline user.