After rotdet2sky has been run for the QPB, and either or both of the SPF and SWCX backgrounds, all of the primary components for a background-subtracted and exposure-corrected image are ready.
Before a final image is produced, it is reasonable to combine the
MOS1, MOS2, and pn data if more than one data set exists. This
can be done using the task combimage, which simply adds the
individual components from the three instruments with proper scalings
for different filters and different instruments, and even
from multiple exposures from the same ObsID. For example:
combimage prefixlist="1S001 2S002 S003"
withpartbkg=true withspbkg=true withswcxbkg=true
elowlist=350 ehighlist=1100 alpha=1.7
prefixlist="1S003 2S004 S005" is a list of the prefixes (exposures) to be combined, withpartbkg=true controls whether QPB images are processed, withspbkg=true controls whether SPF images are processed, withswcxbkg=true controls whether SWCX images are processed, withcheese=true and cheesetype=t controls the source masking, elowlist=400 is a list of the energy band lower limits, and ehighlist=1250 is a list of the energy band upper limits. If multiple bands are combined the resultant output files are labeled with the lowest low energy and the highest high energy. The specified energy bands should be ordered by increasing energy.
The task combimage scales the data sets to the response of the MOS2 medium filter by using a look-up table to scale the exposure. The look-up table was created using PIMMS assuming an power law spectrum with H I cm absorption. The assumed value for is set by the parameter alpha=1.7 Combining MOS and pn data for objects with radically different spectral parameters (different from the assumed spectrum) may produce artifacts in the final image. The artifacts are most easily seen in the CCD gaps where the surface brightness in the gap, which will be covered by a different instrument (pn or MOS), will be brighter or fainter than the adjacent regions. Varying the input parameter alpha may alleviate this effect (which in our experience has been rare).