In Spring 2006, the SSC released the GeRT (Germanium Reprocessing Tools) for Solaris, Macintosh, and Linux operating systems. It can be found in the Data Analysis and Tools section of this website. Most people will find the archived BCD products (filtered or unfiltered) are suitable for their needs, but with the GeRT, you can either ''clean up'' the BCDs, or reprocess the raw data. Examples of cases in which reprocessing from scratch would be important include the following: (a) the online stim calibration was not done optimally, e.g., extrapolated stim calibration or stim calibration corrupted by bright emission regions; (b) important new or updated calibration files and/or pipeline corrections not yet available online but important for a specific case; (c) optimization of the pipeline parameters for non-standard data and/or specific science applications for expert users.
8.2.4 Recommended steps for looking at your data
1. Using the BCDs, look at the online default post-BCD mosaics (*maic.fits, used to be *msaic.fits).
a. If this maic.fits file is ok, then you're done, although it is still useful to check the results by remaking the mosaic offline.
b. If this maic.fits file has bad outlier rejection, remake the mosaic from the online BCDs (using MOPEX).
c. If this maic.fits file has streaks/artifacts, then check the filtered mosaic (*mfilt.fits).
2. Look at the online filtered mosaic (*mfilt.fits), which was constructed from the filtered (*fbcd.fits) files.
a. If this mfilt.fits file has bad outlier rejection, remake the mosaic from the online *fbcd.fits files (using MOPEX).
b. If this mfilt.fits file has negative sidelobes, do 2-pass filtering offline starting with the BCDs and adjust the filtering parameters based on the extent of the source.
3. If stim response artifacts are present due to bright sources, then reprocess from scratch (the ''raw'' data) yourself using the GeRT and make new BCDs by masking out bad stim frames/pixels. You should correct for any multiplicative stim-response effects first and then correct for the residual additive slow response drifts via filtering.
The choice of filtered vs. unfiltered BCDs depends on the your scientific goals. Extended absolute background measurements should be made from the unfiltered BCDs. For point sources and small extended regions, you can use filtered products (optimized for your source, if necessary).
Saturated data close to the saturation limit could potentially be recovered by changing the pipeline parameters for slope determination. Note, however, that this will affect absolute flux calibration.
8.2.5 Note on point source photometry
Since the small field photometry AOTs do not dither with offsets much larger than the size of the MIPS-Ge PSF, temporally filtered fbcd data will underestimate the true point source flux values (by roughly about 10%). We recommend that you mosaic both the filtered and non-filtered data to determine whether or not additional offline processing is needed. You may find that the offline techniques described here significantly improve your data products.
To maintain point source calibration, it's possible to mask out the point sources before filtering. If you don't do that, we recommend that you measure the point sources before and after filtering to check its impact on the flux calibration.
