A typical reduction.

You have one or more pairs of `objects' and `sky' frames for your astronomical target, the same for a `standard' star. You also have one or more flats (halogen lamps up to 2.5 $\mu$m and sky frames beyond) and dark exposures. The first thing to do is creating a normalized flat (FLAT/IRSPEC) and to store it, together with the dark, with SET/IRSPEC. Then use SKYSUB/IRSPEC which does a `clever' obj-sky subtraction aimed at minimizing the residual sky lines in the output frame. The output frame is also flat-fielded and corrected for the fixed pattern of bad pixels in the array (you can use the default pattern or redefine it using DEFINE/IRSPEC, see also BADPIX/IRSPEC). You can then correct the tilt of the spectral line using RECTIFY/IRSPEC. The 2D frame can then be wavelength calibrated using CALIBRATE/IRSPEC; you can take advantage of the on line mechanical wavelength calibration - - usually precise within 1 pixel - or use reference line(s) (often the OH airglow lines in the sky image) in case very accurate wavelengths are needed. Once you have produced the 2D spectrum of your `standard' star, use RESPONSE/IRSPEC to determine the instrumental response - i.e. the conversion factor between counts/sec and flux units - after having defined the star intrinsic flux by means of STANDARD/IRSPEC. The resulting response frame can be used to flux calibrate the spectrum of your astronomical target using FLUX/IRSPEC, and this is the end of the reduction.

There are also two commands of more general interest which allow you to subtract, row by row, a polynomial fit of the continuum from an image (SUBTRACT/IRSPEC) and to `cleverly' merge together 1D spectra at different wavelengths into a table (MERGE/IRSPEC).

A last comment concerning integration times and count units. The IRSPEC images are the average of a number (NDIT) of read-outs each with a given integration time (DIT). Therefore, the counts level does not depend on NDIT and whenever you are asked to use frames taken with the same integration time it means that they must have been taken with the same DIT.