The processing of NUV imaging data is depicted in Figure 3.6
. It is an abbreviated version of the pipeline that only involves those steps which identify bad data and linearize the initial counts. No absolute flux calibration is performed and no background is identified or subtracted.
The final data products for NUV imaging data are the flt
files described in Section 2.4.2
. Like the x1dsum
files, an fltsum
file is created even if only one exposure is processed. However, since no shifting is performed for imaging observations (see Figure 3.6
), the fltsum
file is a simple exposure-time-weighted mean of the individual flt
files (and it is identical to the flt
file if only one exposure contributed to it). The DQ flags image of the fltsum
file and, for that matter, all of the individual flt
images, are identical. This is because the only data which make it into an flt
image are free of temporal or event flags (see Section 2.7
). Consequently, in the absence of shifting, all of their spatial flags should be identical.
Although imaging data are not flux calibrated, a crude calibration can be performed using the total count rate from the flt
file and one of the two keywords provided in the count rate extension header. The header keyword PHOTFLAM
is appropriate for a source spectrum that is flat and featureless across the MAMA detector band when measured in units of power/area/wavelength and PHOTFNU
is appropriate for a source that is flat in power/area/Hz units. The values provided for the PHOTFLAM
keywords depend on the specific combination of mirrors and apertures used in the observation.