CALCOS v2.11, v2.11a, and v2.11b Release NotesCALCOS v2.11 will be installed in the STScI Archive pipeline on or about September 14th, 2009. This version will be installed with OPUS 2009.2j release. CALCOS v2.11 has not been installed in STSDAS yet, but will be distributed with the next release of STSDAS, which is anticipated to occur in late November 2009. Although there were only a few changes between CALCOS v2.10 and 2.11, this version is considerably different to the previous version of the COS calibration software installed in the pipeline CALCOS 2.8. Since then CALCOS have gone through versions 2.8.1, 2.8.2, 2.8.3, and 2.8.5. For a complete list of the changes that went into these version, please check their release notes. ---------------- CALCOS v2.11 ---------------------------------- Relevant PRs for v 2.11: OPR: 62895 Allow CALCOS to use a corrtag file as an input OPR: 62997 CALCOS may need to update LAMPUSED OPR: 63346 CALCOS incorrectly determines active area in some cases OPR: 63347 CALCOS needs to initialize average shifts Relevant PRs for v2.11a: OPR: 62270 CALCOS must update ASN_MTYP when copying x1d to x1dsum Relevant PRs for v2.11b: OPR: 62270 CALCOS must update ASN_MTYP when copying x1d to x1dsum OPR: 63475 CALCOS sign error when averaging shift1 for NUV stripes Relevant Trac tickets for v2.11: 417, 425, 437, 438 Relevant Trac ticket for v2.11a: 386 Relevant Trac ticket for v2.11b: 386, 445 v2.11 Two bugs having to do with the data quality (DQ) image have been fixed. One is that the offset and smearing for the Doppler shift were not being done if the input was a corrtag file. The other is that the wrong column in the corrtag table was used when determining the range of wavecal shifts for FUV data. The range is determined from the pixels within the active area, and the effect of using the wrong column is that pixels with zero shift (because they were actually outside the active area) were included, thus resulting in a range of wavecal shifts from zero to a possibly large shift, which in turn resulted in very wide regions being flagged as bad in the DQ image. When the input is a corrtag file, it is no longer necessary to specify an output directory if the input file is not in the current directory. The input and output directories must still be different (for corrtag input), but the output directory may be the current directory. For tagflash data, if there were no lamp flashes at all, calcos now sets the LAMPUSED keyword to NONE before writing an empty lampflash table. If there are no flashes but that wasn't determined until after further processing, the dictionaries of average shifts is now initialized with zero values. v2.11a A change was made to 2.8c to update ASN_MTYP for the case that there was only one x1d file that was copied to x1dsum (i.e. rather than averaging multiple x1d files to make the x1dsum), but I forgot to copy that change back to more recent versions of calcos. This is the only change from calcos 2.11 and 2.11a. v2.11b There was a bug in updating keyword ASN_MTYP in fltsum headers. The keyword is only present in the SCI header, but calcos was trying to update it in the ERR and DQ headers as well. When the shift in the dispersion direction is found for only one or two of the NUV stripes, the shifts that are found are averaged and used to assign values to the shifts that were not found. Since the stripes do not move together, a rough correction for the skew between stripes is included. There was a sign error in this correction, however.