Changes that have been made to CALSTIS in version 1.8 include:
1-d spectral extraction (x1d)
Several useful processing values determined during spectral extraction have been written to the trailer file. These entries, listed below, are now included the the x1d table file.
| A2CENTER | default or user specified y extraction location |
| EXTRSIZE | size of extraction aperture |
| MAXSRCH | search radius for cross correlation location of extracted spectrum |
| BK1SIZE | size of background box 1 |
| BK2SIZE | size of background box 2 |
| BK1OFFST | offset of background box 1 |
| BK2OFFST | offset of background box 2 |
| EXTRLOCY | y pixel array of extracted spectrum location |
To prolong the lifetime of the MAMAs, we are shifting the location of the spectral images in the cross-dispersion direction on a monthly basis (see the MAMA Phase II update). This offsetting start in January 1998, and the offset values are stored in the keywords MOFFSET1 and MOFFSET2. These keywords were not being properly read, so a zero offset was being used by calstis-6 (x1d). These keywords are now being properly accessed. Note for the first order data, this error does not impact your data (since the finding algorithm searches the entire detector for your spectrum), but for echelle modes, this error can cause orders to be mis-extracted in cases where the shift put the order beyond the search region. We are presently reprocessing all the impacted echelle observations, and will have the corrected files placed in the archive and email sent to the appropriate PIs so they can retrieve the recalibrated data. We expect the recalibrated data to be available in the archive in a few weeks. Note that this problem only affected GOs with echelle data taken after January 1998.
Wavelength calibration (wavecal)
If a science observation and its associated wavecal were taken with different apertures, the offset between those apertures was being taken into account twice, thus resulting in an improper wavelength calibration (typically of order 1 pixel). This problem has been corrected. GOs requiring accurate wavelength scales for observations where the science data and automatic wavecals were taken through different slits (which will be the case for observers using slits which are >= 0.5 arcseconds wide) will want to recalibrate their data.
Reference file processing for MAMA observations
The computation of the Doppler shift for convolving reference files originally had the incorrect sign; this has not been corrected. The affect of this sign problem is to reduce the signal-to-noise in the calibrated data for echelle and first order M mode data taken with the NUV-MAMA. FUV-MAMA and first order L mode data is unaffected (the former because no flat is applied currently in the pipeline).
A number of additional enhancements and minor problem fixes include:
- when processing a reference image which requires binning, calstis will no longer print warning messages about new keywords being added to the header (since the binned reference file is not saved).
- if an invalid detector name is in the header, calstis will now stops rather than assuming a default detector.
- if no output file name is given, the input file name is used to determine the output names. Calstis now properly modifies the input file name to remove the directory prefix if one is given.
- when running x1d, the output name can now be omitted, and the input name will be used to define the output name.
- the calibration switches are now case insensitive (e.g. OMIT or omit).
- it is now possible to perform flux calibration (during 1-d extraction) without performing background subtraction.
New tools released with CALSTIS version 1.8 include:
ocrreject - the cosmic ray rejection task, which was updated to allow for background subtraction, for multiple file input, and for new input parameters
tastis - a task to analyze target acquisition data, which prints the position of the target and reference aperture in the ACQ images, the flux of the target and slews in the coarse and fine locate phases, and for ACQ/PEAKs, the flux at each step of the peakup, the flux in the confirmation image, and the calculated slew. A warning is printed if there appears to be some possible problem with the acquisition.
daydark - a task to create a daily dark
these tasks are used to create a fringe flat:
normspflat
prepspec
mkfringeflat
defringe
pp_dads - this task is used to generate paper products, and has been significantly enhanced (particularly to show 1-d extracted spectra)