When using eis_prep on slot data (i.e., data obtained with the 40" and 266" slits) care should be taken with how the radiometric calibration is performed.
Consider 40" data in the Fe XII 195.12 line. The image formed in this line will extend approximately from 194.7 to 195.5 angstroms in the spectrum. Since it is is the Fe XII that mostly contributes to this image, then the effective area that should be used for the radiometric calibration is that which applies to the Fe XII line. However, the EIS calibration software does not know that Fe XII dominates in this wavelength window, so what it does is take the effective areas at the edges of the window and averages them. That is, it assumes that the window has the correct size to match the slot width on the detector and that the effective area does not change much over the window.
This generally works fine for window data with the 40" slot. For the 266" slot, however, it generally will not work. The Fe XII image will extend from 192.2 to 198.0 angstroms. Averaging the effective area at these wavelengths gives a value that is 14% lower than the effective area at 195.12.
The method also does not work for full CCD spectra obtained with the slots as a single effective area (the average of the two ends of the detector segment) is used for the entire spectrum.
An alternative method of doing the calibration is as follows.
1. Run eis_prep with the keyword /noabs (in addition to the regular keywords) to keep the data in DN intensity units.
2. Find the exposure time (exptime) for the observation (e.g., use the getexp method).
3. Run the routine eis_slot_calib_factor to obtain the DN to erg cm-2 s-1 sr-1 conversion factor.
As an example of using eis_slot_calib_factor, consider the case of Fe XII 195.12 and an exposure time of 10 seconds:
IDL> f=eis_slot_calib_factor(wavelength=195.12,exptime=10.0)
The output, f, is a scalar that you multiply your slot intensity image by to convert from DN to erg cm-2 s-1 sr-1.