• Download HEASOFT
  • ******************************************************************************
                            RELEASE NOTES FOR HEASOFT 6.1
                                    July 24, 2006
    As for previous HEASOFT releases, configuration-related changes thoughout
    the package means that users who already have HEASoft v6.0.X installed will
    need to completely replace their software installation; there is no clean
    way to simply overlay this new release or to update an existing installation.
    For additional information about changes to the ATTITUDE and SWIFT packages
    since the release of HEASOFT 6.0, please see the release notes for HEASOFT
    v6.0.2 through 6.0.5.  Unless otherwise noted, the updates described below
    include changes made since the HEASOFT 6.0 release.
    New in HEASOFT 6.1 is support for the Macintosh Intel and 64-bit Linux PC
    (x86_64) architectures.
     - Fixed so that it looks for the EVENTS extension and doesn't just use the
       first extension.
     - Allow non-square chips by setting the internal image arrays to the maximum
       of the RAWXCOL and RAWYCOL sizes.  This is not required for ASCA but allows
       this tool to be used for Swift XRT and Suzaku XIS.
     - Fixed a problem with the way the exposure maps were computed, which led to
       a spurious pixeled map of the mosaic-ed exposure map.
    -  read_emap subroutine was rewritten: split the pixels and keep track of
       the number of times they're split.  There is no randomization done for
       the exposure map, only for events files (data and background).
    Attitude software changes since HEASoft v6.0.5:
    Revised tools
     * A minor change was made to SAA definition.
     * Save outcome and quality of correction as keywords in outfile.
     * Populate XXX_DELTA columns based on distance of detection (instead
       of catalog position) from XXX.
    CFITSIO Version 3.01 -  Main changes since the HEASOFT 6.0 release
       - a major enhancement to the CFITSIO virtual file parser was provided
         by Robert Wiegand (GSFC).  One can now specify filtering operations
         that will be applied on the fly to the pixel values in a FITS image.
         For example [pix sqrt(X)] will create a virtual FITS image where the
         pixel values are the square root of the input image pixels.
       - enhanced the 'col' extended filename syntax to support keyword name
         expressions like
            [col error=sqrt(rate); #TUNIT# = 'counts/s'],
         in which the trailing '#' will be replaced by the column number
         of the most recently referenced column.
       - added 2 new random number functions to the CFITSIO parser
         RANDOMN() - produces a normal deviate (mean=0, stddev=1)
         RANDOMP(X) - produces a Poisson deviate for an expected # of counts X
       - added fits_copy_cell2image and fits_copy_image2cell routines for
         copying an image extension (or primary array) to or from a cell
         in a binary table vector column.
       - added new fits_read_wcstab routine that serves as an interface to
         Mark Calabretta's wcslib library for reading WCS information when
         the -TAB table lookup convention is used in the FITS file.
       - changed the definition of fits_open_file into a macro that will test
         that the version of the fitsio.h include file that was used to
         build the CFITSIO library is the same version as included when
         compiling the application program.
       - cfortran.h has been enhanced to support 64-bit integer parameters
         when calling C routines from Fortran.  This modification was kindly
         provided by Martin Reinecke (MPE, Garching).
       - modified region.c so that it interprets the position angles of regions
         in a SAO style region file in the same way as DS9.  In particular, if
         the region parameters are given in WCS units, then the position angle
         should be relative to the WCS coordinates of the image (increasing CCW
         from West) instead of relative to the X/Y pixel coordinate system.
         This only affects rotated images (e.g. with non-zero CROTA2 keyword)
         with elliptical or rectangular regions.
       - fixed 2 bugs that could affect 'tile compressed' floating point
         images that contain NaN pixels (null pixels).  First, the
         ZBLANK keyword was not being written, and second, an integer
         overflow could occur when computing the BZERO offset in the
         compressed array.
    Changes for CCfits 1.5:
    - More thorough fix to PHDUT.h write function:  Now genuinely handles
      the stride vector parameter.
    - In BinTable::addColumn function for case of strings, it was outputing
      the column format incorrectly as "A<nChar>".  Should have been "<nChar>A".
    - In BinTable and Table constructor, now checks for columnUnits vector 
      size rather than assume it's filled in.  This is necessary if units are
      to be an optional parameter for the FITS::addTable function, as claimed 
      in the docs.
    - Modifications needed to fix compile errors on certain platforms with
      g++ 4.0.x.  These are functions that haven't been instantiated, which is
      why other compilers didn't complain about them.  In ExtHDUT.h read 
      function, needed explicit cast to size_t to make both variables in
      std::min call the same type.  Same fix made in PHDUT.h read function.
      Also in PHDUT.h write function, several calls to PrimaryHDU::writeImage
      had the wrong number of arguments.
    Changes for CCfits 1.4:
    - Fixed reading in of BSCALE and BZERO.  Previously these keyword values
      were not being saved in getScaling function.  HDUCreator::getScaling
    - Now reads in all floating-point keyword values as doubles rather
      than floats.  In conjunction, needed to allow Keyword casting from
      doubles to floats to keep things backward compatible
    - In ColumnT.h read function which takes row as a parameter, test and
      throw for case of row out-of-bounds of table.
    - Memory leak fix in HDU.cxx.  Everywhere insertions into m_keyWord
      map were taking place, needed to check for previously existing entry
      with same key name and delete if it existed.  FITS::read often does 
      insertions twice with the same Keyword, which was causing a memory leak.
      Also fixed a couple minor leaks in HDU::writeDate.
    - Exception safety / mem leak fix in FITS.cxx and FITSBase.cxx.  
      All FITS constructors (except copy c-tor) now use auto_ptr when creating 
      FITSBase* m_FITSImpl.  Also FITSBase destructor now checks if it needs
      to close file, indicated by non-zero m_fptr.  To make this work,
      FITS::close now resets fptr to 0 after it closes file.
    - Add another case of allowed casting of keyword values, this time
      from strings to int/float/double if the string contains an integer value
     - Fixed a Darwin-specific bug which generated a fault in "expand" mode. 
     - genrsp has received a major overhaul to provide greater flexibility.
       1. Response energies and channel definitions can now be read from input
          (ascii) files. This required adding new parameters resp_reln, resp_file,
          chan_reln, chan_file.
       2. If the response energies and/or channel definitions are not read from
          the file then the parameters to define them have been altered to set
          the number of energies/channels rather than their size. This will allow
          an extension to logarithmic intervals at some time in the future.
       3. The response energy parameter names now all start resp_. To avoid
          possible confusion the resolution parameters res_reln and res_file
          have been renamed to resol_reln and resol_file
       4. The resol_file input file format has been altered to allow multiple
          response peaks. genrsp can now build a response with multiple gaussian
          peaks from a given input energy (eg photopeak and escapes). At present
          the input file is ascii but a FITS format will be supported when a
          standard has been defined.
     - Added arfcol parameter to allow a column other than SPECRESP to be used
       when reading the ARF file.
     - version 3:
       Substantially changed the logic by which calibration conflicts are
       determined. Now for "base conflicts" (where cnam, reftime,detnam,
       telescop, filter all match) it will check all the boundary keywords
       for overlaps.  If a parameter name is missing it will alert the user.
       If a parameter name matches but the parameter has different UNITS, it
       will alert the user.
       For "base conflicts", if either boundary parameter array is completely
       filled with NONE, then there's a boundary overlap and the files conflict:
       For "base conflicts" where both boundary arrays contain at least one
       non-NONE entry, the boundary parameter names are checked to see if any
       match. If all the parameter names match, and the parameters all have
       overlapping values then conflict.eq.true; otherwise there's at least one
       parameter who's values don't overlap, so conflict=.false. Now for "base
       conflicts" (where cnam, reftime, detnam, telescop, filter all match) it
       will check all the boundary keywords for overlaps (not just the first one).
       Also updates checksum and datasums in 1st extension of caldb.indx file. 
       Also added check to make sure calibration files TELESCOP values in CALDB
       file match all those in the CIF, to avoid the "suzaku" vs. "sazuku" problem
       Removed option of including "duplicate" data files since the CBD block is
       now correctly checked.
       Added CHATTER parameter for more explicit feedback
     - Fixed a bug that could cause chimgtyp to skip a couple of keywords near the
       beginning of the header when copying keywords from the input to the output
       file.  It previously assumed that there were always 4 COMMENT keywords
       that should not be copied which gave the bibliographic references for the
       definition of FITS, which is not the case in general.
     - Added test so as to not duplicate the FITS definition COMMENT keywords
       at the beginning of the header if they already exist in the input file.
     - Fixed bug in the way null pixels in the input image are handled.
       Previously, if there were any null pixels in the input image, then a
       random pixel in the first few rows of the output image would be set to
       null instead of at the correct pixel location.
     - Increased the number of significant digits in output statistics.
     - Fixed improper handling of sky coordinate wrap-arounds which could
       result in removal of all events with DETX >= 182, not just those which
       were masked.
     - Fixed bug which reversed the WCS keywords for the X and Y axes of the
       output image.
     - Modified to ignore extra <CR> or <LF> characters at end of lines of text
       in the input ASCII files.
     - Added copykeys parameter to control copying of header keywords.
     - Modified to print the number of errors and warnings separately instead of
       just the total, then only printing a 1-line pass/fail report for each FITS
    FV 4.4:
        * Added capability to view/modify subset of FITS columns.
        * Added capability to re-order the display of FITS columns.
        * Expanded capability to sorting/modification of FITS columns parameters.
        * Tracking parent window view via vector columns in "Expand"ed view .
        * Allowed user to specify path for DS9 and XPA in Windows.
        * Fixed problem with printing a contour map on an image in MS Windows
        * Modified how the rotation angle in region files is interpreted to be
          consistent with DS9 (angle is w.r.t. the North direction in the image).
    FV 4.3 (April 2006):
    New features/Bug fixes since V4.2 (in HEASOFT 6.0) include:
        * Added capability to invoke future HERA modules.
        * Added capablity to calculate Flux on region.
        * Added capability to handle "Physical" attribute in region file.
        * Added capability to center (and resize) of ROI (region of interest) when
          the image is zoomed.
        * Added capability to allow parameters editor to stay until user decided to
        * Fixed running FTOOLS tools not copying physical par file problem.
        * Fixed sorting Fits table but displaying wrong slice image problem.
        * Fixed problem in rotation of coordinates resulting in wrong region
        * Fixed plotting of selected rows from FITS table problem.
        * Fixed ploygon region handle placement and dragging mechanisms problem.
     - Fixed bug in the CGRO tool bod2rmf.  For at least every tjd between
       10179 and 10464, the energy edges for channel 15 were wrong.
     -  Now uses 'CALDB' as the default parameter for the Swift clock correction
        file. The legacy ascii file ("swco.dat") is also still supported, however.
     - v1.6 handles Swift orbit file gaps
     - v1.4 handles leap seconds properly
     - Allow for a larger number of input files (100 instead of 10) since
       mathpha is no longer limited to 10 input files.
     - Modified so an EXTNAME of SPECRESP MATRIX is searched for as well as
       just MATRIX.
     - Fixed infinite loop which arises if one of the extensions doesn't have
       its EXTNAME set.
    fovdsp v2.5:
     - Suzaku HXD nominal position (DETX,DETY)=(-3.5arcmin,0.0) is marked for
       the Suzaku FOV.
     - TGT_RA, TGT_DEC and TGT_NAM parameters are added for target RA, DEC
       and target name.
     - "Segmentation error" was fixed when input catalog is not found.
     - Suzaku FOV (HXD-GSO, HXD-PIN and XIS) may be drawn.
     - "ipixeldist" is no longer used.
     - Fixed failure of 't' to work when sources are not plotted.
     - Fixed so bad channels that were written with grouping of 0 now have
       grouping of 1. This for conformity with standard and prevents xspec v12
       producing a warning message.
     - Changed error to warning if ROWID column is not found.
    extractor (updates since HEASOFT 6.0 release):
    extractor 4.66 - 
      - Added support for PANDA regions but only for the case of one sector and
        one annulus. For PIE regions fixed an incorrect definition of the angle
        and a bug that excluded all events with angle > 180 degrees. In all
        regions angles now take into account any CROTA set in the input event file.
      - Fixed a bug which caused incorrect time ranges to be written to the
        output xronos window file if the input event file had different GTI for
        each chip.
      - Added a keyword to lightcurve output to record the minimum value of FRACEXP
        accepted for bins written to the output.
      - Increased the limit on the number of GTIs allowed.
    extractor v4.62 -
      - Fixed bug that causes light curves from Swift UVOT data to have a TSTART
        keyword of zero and hence incorrect TELAPSE keyword.
    extractor v4.61 -
      - Fixed bug that produced an incorrect error message about inconsistent XCOLH
        and YCOLH when reading multiple event files.
      - Fixed bug in extract events when reading from multiple event files with
        different numbers of columns. If later files had more columns than the first
        one then the events from those files could have zero values for some columns
        even for columns that were in common between the files but in different
      - Fixed seg fault that arises if a line in a region file which includes a
        region specification also contains a comment marker (#).
      - Trapped the case of the grade column having no TLMIN/MAX. Somewhat
        arbitrarily assume them as 0 and 25 but this shouldn't matter as long as
        the real limits are within this range.
      - Fixed error introduced in v4.55.  When ellipse region was corrected to
        define the rotation angle in the correct sense, the corresponding code
        for box regions was broken.  The set_region and select_ellipse routines
        define the rotation angle in opposite senses.
    extractor - v4.56 -
      - Fixed problem that arises if the input event file has no GTIs and there
        are selection GTIs in use. The resulting event file was given the
        selection GTIs instead of the correct result of no GTIs.
      - Modified the region file reading routine to ignore the 'CR' character
        if it is the last character on the line.  If the region file was created
        on a PC, which uses CR LF as the 2 End-of-Line characters, then the CR
        is left over when read on a Unix machine, which uses only a single LF
        character to mark the end of line.
      - Fixed inconsistency between rotation angle defined in ds9 ELLIPSE and BOX
        regions and their use in extractor. It appears that extractor was treating
        these angles as clockwise from the x-axis while ds9 defines the angles as
        counter-clockwise from the same axis.
    extractor - v4.54 -
      - Fixed so that wtmapfix=yes writes -1 to pixels outside the region even when
        the image and wmap coordinates are the same.
    New tasks:
     - New task for performing coordinate system conversions between galactic,
       equatorial and ecliptic, either on a table or at the command line.
     - New task for performing arbitrary image calculations, including
       world coordinates and multiple image extensions
     - New task for joining two tables together using arbitrary matching
       criteria (similar to a database "join")
     - Create a new FITS image from a combination of 1 or more input images
    Updates to existing tasks:
     - Changed the support for X bit columns so that users always need to input
       the value of the unsigned decimal or hex integer that has the equivalent
       bit pattern.
     - Modified to show more significant digits when displaying differences
       between floating point values.
     - Modified so that it does not interprete any blank characters at the end
       of a line in the input data file as a zero value.
     - Added 'copyall' parameter to control copying of other HDUs.
     - Added 'lastkey' parameter for list of keywords to be updated or added
       to the output file.
     - Added 'skipbadfiles' parameter to allow files that cannot be opened for
       any reason to be ignored.
     - Fixed to handle NULL values properly.
     - Fixed a problem that caused type conversion overflow.
     - Added test to ignore logical (L) columns in binary tables when computing
       the statistics.
     - Modified so that it doesn't exit with error if the table contains a
       string (non-numeric) column.
     - Fixed to prevent y axis centroid world coordinates being printed out
       for x axis also.
     - Added a few more digits of precision to the image mean and stdev values.
     - Fixed memory deallocation problem when input file is invalid.
     - Updated to warn about null-valued keywords.
     - Modified to print the number of errors and warnings separately instead
       of just the total, then only printing a 1-line pass/fail report for each
       FITS file.
    Latest QDP/PLT from Allyn Tennant, http://wwwastro.msfc.nasa.gov/qdp/
    Changes to QDP/PLT since our last update (2003-07-24):
    2006-04-10 - Fit Plot -100 (negative number) should now work (again).
    2005-09-22 - CURFIT now always writes out the "model insensitive to" error.
    2005-08-24 - YAX LIN now affects groups not windows (like XAX LIN).
    2005-08-15 - WM adds parameter numbers as comment, mo @ ignores comments.
    2005-06-29 - Fix use of GETLUN/FRELUN when using WEnvir.
    2004-07-22 - Add LAb # Pos x1 y1 To x2 y2 option.
    2004-02-27 - Fix bug that caused "LAb 1 VPort Curs" to fail.
    2004-02-24 - Fix bug that prevented same COD file from being used twice.
    2004-02-04 - In cod functions, "ASLO" no longer matches "ASin".
    2004-01-22 - Fix case where MA ? would be displayed in background color.
    - hrirpsf: version 2.1.2
      fixed uninitialized ierr error when getting INFILE parameter
      write PIXSIZE, BACKGRND & SUMTCOUNTS keywords in theoretical profile
    - pcrpsf: version 3.5.4
      fixed uninitialized ierr value
      corrected writing of PIXSIZE keyword in output file for theoretical profile
    Swift-specific software changes since HEASoft v6.0.5
    New tools
     * New task which adjusts a BAT image for known non-linear image distortion 
     * For source detection, batcelldetect should be used instead of this task
    Revised tools
     * This task finds and corrects anomalous Swift attitude "jumps" during
       the settling phase.
     * Bug fix for the case when the spacecraft does a "slew-in-place".
     * Bug fix to the way the final time bin was written, in some rare cases.
     * Bug fix to transfer the MJDREF* keywords properly, when combining GTIs.
     * Bug fix to remove buffer overflow error tickled by the new parameter library.
     * Primary HDU images are now labeled with HDUNAME instead of EXTNAME.
       This is still compatible with standard CFITSIO and DS9.
     * New time binning algorithm, MATCHLC, which allows the time binning
       of an existing light curve to be cloned.
     * Write NULL values in bins that have zero exposure (when minfracexp is 0)
     * Slight changes to error messages.
     * Bug fix: clustering of >2 sources during the PSF fit stage was done
       improperly, and is now fixed.
     * Bug fix: known bright sources which touch the border of the image were being
       improperly detected as new sources.
     * New feature: adds distortion correction to position reporting
       (the distortion map is pointed to by the 'distfile' which defaults to CALDB).
     * Internal changes of the use of the WCS library.
     * Primary HDU images are now labeled with HDUNAME instead of EXTNAME.
       This is still compatible with standard CFITSIO and DS9.
     * Catalog I/O is more forgiving if TUNIT/TNULL values are not present.
     * Primary HDU images are now labeled with HDUNAME instead of EXTNAME.
       This is still compatible with standard CFITSIO and DS9.
     * New feature: adds distortion correction to position calculation
       (the distortion map is pointed to by the 'distfile' which defaults to CALDB).
     * New feature: adds distortion correction to position calculation
       (the distortion map is pointed to by the 'distfile' which defaults to CALDB).
     * Primary HDU images are now labeled with HDUNAME instead of EXTNAME.
       This is still compatible with standard CFITSIO and DS9.
     * Additional error handling code.
     * Bug fix: when the earth is at the very edge of the FOV and exclusively in the
       southern hemisphere, the whole FOV was accidentally blanked.  Now fixed.
     * Primary HDU images are now labeled with HDUNAME instead of EXTNAME.
       This is still compatible with standard CFITSIO and DS9.
     * Additional error checking for bogus inputs.
     * Additional error handling code.
     * Additional error handling code.
     * Exclude nulls when estimating background (significant for genie images with missing packets).
     * Use pixel filtering instead of f*arith for image calculations.
     * When creating images from EVENT data, use the new columns in the WINDOW
     * Exclude parameter has new default value DEFAULT which is equivalent
       to exclude=ASPCORR:NONE,DUPEXPID.  The new special value DUPEXPID
       checks for duplicate EXPIDs and includes only the exposure with
       the largest window if there are two HDUs for the same exposure.
       This simplifies avoiding including exposures taken in IMAGE&EVENT
       mode twice in the sum.
     * "coinfile" parameter has new special value NONE which indicates to skip
       the coincidence loss polynomial correction (coincidence loss correction 
       is still performed, but as if the polynomial evaluated to 1).
     * Frametime parameter has new default value DEFAULT which indicates
       to use the value of the FRAMTIME keyword if it is available
       or fall back on 0.0110322 [s] if it is not.  The user can still
       provide an exact frametime with frametime=0.0123 for example.
     * The aspect corrections file is now FITS and includes a number of details 
       related to the exposure and correction for the benefit of analysis.
     * New required parameter "respfile" (filename or "CALDB") sets the value of the 
       RESPFILE keyword in the output pha file.
     * DATE/TIME keyword handling was improved (eg, flexible about DATE_OBS vs. DATE-OBS).
     * Handling of the TIME column from the Bad Pixel Calibration Files.
     * Handling of the TIME column from the Bad Pixel Calibration Files.
     * Pixels falling outside the Data Mode Window are set as not exposed.
     * Unfiltered event files processed by a previous run are cleaned up before 
       running stage 1.  (STATUS column set to 0 and BADPIX extension erased).
     * The KWM1STCOLNUM keyword in the Windowed Timing EVENTS extension header 
       get updated when the column shift is applied.
    - sec2time:
      Fixed bug which produced incorrect results when the computed (output) date
      is near the end of a leap year (e.g. 12/29/04).
    - sec2time / time2sec:
      Fixed bug which caused a segmentation fault when CALDB was input for the
      leapfile parameter.
    For full documentation on XIMAGE, refer to:
       * Added Intel Mac and 64-bit Linux support
       * Improved text rendering in PostScript output
       * Grid command output more closely matches pre-4.2 appearance
          (e.g. no abbrevation of grid line labels)
       * Parameter file (ximage.par) replaces functionality of
          ~/defaults.def file
       * Changed rotated region convention to match that of ds9
       * Improved ra_dec_to_pixel command to accept all common RA/Dec formats
       * Suzaku PSF approximation added
    For complete change history of XIMAGE see:
    Changes since the Ximage 4.3 release in HEASOFT 6.0 also include:
      - Changed default equinox to 2000
      - Improved handling of exposure maps by utilizing the VIGNAPP keyword.
      - Improved handling of missions with unknown psf.
      - Added psf/pileup option to analyze piled-up sources.
      - Corrected Swift XRT psf and vignetting functions.
    Updates since HEASOFT 6.0 release:
    xselect 2.4:
     - Fixed a bug when using filter column on the spectrum column (eg PHA or PI).
       Note that there is a conflict if such a filter column command is used in
       combination with filter pha. In this case the filter pha is ignored
     - Fixed filter time UT which was not working at all.
     - The automatic response generation option for save spectrum now uses
       the badpixel data for the XMM EPIC. At present this requires that
       there only be one event file from which the spectrum was extracted. If
       there are multiple event files then the user is advised to do an
       "extract event" then save the spectrum. If the SAS is not set-up locally
       then xselect will automatically call Hera to do the response and arf
     - For Suzaku XIS set the the quality of channels 1-81 3290-4095 to bad. This
       means these channels will be ignored in xspec when the "ignore bad" command
       is used.
     - Modified mkfbin and hkbin so that if an event lightcurve has been
       extracted then the mkf or hk parameters will be binned to match the event
       lightcurve and the RATE will be added to the mkfbin/hkbin output
       file. plot mkf and hk then enables the user to plot correlations between
       housekeeping parameters and the event rate.
     - Added filter mkf and filter hk options as aliases for select mkf and
       select hk. Added save mkfbin and save hkbin options.
    xselect v2.3:
      - Sensitivity to case of mission, instrument, etc. removed.
      - Fixed error that datamode keyword was not allowed to be different for
        different instruments.  Trapped case of a blank string being entered for
        a filename.
      - xsl_chandra_acis_makeresp:
        mkacisrmf wants the spectrum file as the input of the wmap parameter,
        not the weighting file produced by mkwarf (as mkrmf did).  Modified
        call to mkacisrmf to include all required parameters.
      - Trapped out the case of returning an obsno greater than the available size
        of the obscat array. This prevents a possible bug that can lead to input
        data files being deleted.
      - Fixed bug that can cause make obscat to delete input data files if there
        are more than 8 types of instrument available and the one chosen is > 8.
        Reset the size of the array to 20.
      - Suzaku-related changes:
        Default mission is now Suzaku.  Automatically group XIS spectra by a
        factor of 4 and add the DETNAM to the displayed info for the HXD so we
        can see whether an event file is PIN or GSO.
      - Corrected failure to propagate MJDREFI/F everywhere it was required which
        led to incorrect label at the top of lightcurve plots.
      - Fix so that FILTER TIME FILE works for files whose EXTNAME is either GTI
        or STDGTI. Previous version worked for GTI only, which gave an error when
        using output from maketime.
      - Fixed so looks for the first GTI extension instead of just assuming it
        is the first extension when doing a filter time file.
      - Fix for bug under Linux in the plot mkf command. The xparm string was being
        overwritten leading to junk being passed to plt as the label for the X axis.
                                       XSPEC 11
    xspec v11.3.2t: 
      Please see http://heasarc.gsfc.nasa.gov/docs/xanadu/xspec/xspec11/bugs.html
      for details on the xspec 11.3.2t patch.
                                       XSPEC 12
    Xspec version 12.3.0:
    Xspec version 12.3.0 incorporates the fixes and enhancements from all of the
    patch files (a-ap) for v12.2.1, in addition to the following key changes:
    Enhancements and Modifications:
    - Full support for models with multiple responses (previously only 
      available for Integral/SPI data formats).  This replaces and enhances
      the v11 technique of specifying '/b' for background models.
    - New command "energies."  This is a more general and flexible 
      replacement to the "extend" command of v11.  
    - 6 new models added to the standard XSPEC models library.
      Additive:  diskpbb, laor2, nsatmos
      Convolution:  kdblur, kdblur2, rdblur
    - First-time support of XSPEC12 on Windows platform via Cygwin (except
      for initpackage and lmod commands for loading model libraries, see 
    - First-time support on Intel Mac platform.
    - New "show" command option, "show plot", displays all current 
      "setplot" settings.
    - New handling of zero variance during chi-square fits.  Previously a
      minimum variance was ALWAYS applied to zero-variance channels for any
      type of fitting.  Now it is only applied for chi-square fitting.  
    - Gaussian randomization is now applied in "fakeit" for non-Poisson
      type spectra.  Previously Poisson randomization was applied in all cases.
    - Enhancements to output of "tclout" flux and lumin options.
    - The "weight" command's model weighting option is now working.
    - The STAMP option for the "log" command is now implemented for the 
      first time in v12.  
    - During C-statistic fitting, the Poisson data warning is now issued
      only when the data are not Poisson.
    - Fix to "undo" command for partial data removals.
    - Fix to variance calculation for case where Poisson errors are 
      selected and data (source and/or background) is given in RATES.
    - For (e)emodel and (e)eufspec in setplot wave mode, the "e" multiplier
      is now in energy units, not wavelengths.
    - A new implementation of model command parser should fix various
      parsing problems that arose with certain complicated model expressions.
    - No longer using a start-up xspec script file to launch executable.
      This had been causing problems on Linux with the ctrl-z suspend mechanism.
    Changes since the Xspec 12.2.0 release in HEASOFT 6.0 also include:
    Xspec version 12.2.1:
    Version 12.2.1 incorporates the fixes and enhancements from all of the 
    patch files (a-ck) for v12.2.0, in addition to the following key changes:
    - The Mac OS X Tiger build difficulties that had been causing runtime
      crashes have now been solved.  XSPEC libraries that were built as
      shared on other platforms are now also shared on OS X Tiger.  
    - OS X Tiger users should now also be able to dynamically load 
      their Fortran local model libraries, even if they share the XSFunctions
      COMMON block.
    Enhancements and Modifications:
    - Runtime HELP can now be accessed ONLINE in html format, rather than
      just with the distributed PDF documents.  The selection for this is made
      in the new Xspec.init file.
    - Users can manually re-seed XSPEC's pseudo random number generator
      at any point in the program using the new "xset seed" command option.
      The default initial seed is taken from the time at program start-up.
    - Uses new global_customize.tcl file (in Xspec/src/scripts) for
      storing customized start-up commands that are to be run for all
      system-wide users.  Previously these commands had to be appended
      to the bottom of the xspec.tcl file, and hence had to be re-applied
      if xspec.tcl was ever patched.  
    - Model and model-related plots (ufspec and its variants) now always
      default to logarithmic scaling for both axes.
    - New model components added, additive: nsagrav, multiplicative:
      zdust and zsmdust.
    - Please see http://heasarc.gsfc.nasa.gov/docs/xanadu/xspec/issues/issues.html
      for details on the patch files a-ck.
    xstar v2.1kn6
    Fixed an error which causes the wrong inital radius to be calculated when
    the constant pressure option is chosen. Also changed the units label on
    the ionization parameter to remove inconsistency with constant pressure
    Minor changes in the radiation transfer algorithm.
    Fixed bug which affected high ionization models which included nickel.
    This caused segmentation faults, and was caused by an incorrect data type
    flag in the atomic data for He-like Ni.
    Changed step size algorithm to prevent stepping beyond the
    column density specified in the input. This will not be accurate for
    constant pressure clouds in which the temperature is changing rapidly.
    Added the effect of photoionization and heating by line photons generated
    elsewhere in the cloud. These are photons which have already escaped the
    local region close to the point of emission.
    Changed the step size computation algorithm in order to account
    for the process of emission. That is, the step size is now is based on the
    length scale for significant change of both absorption and emission.
    Fixed several errors in the atomic database, notably affecting
    N-like ions. These affect some of the density sensitive lines in low
    ionization models.
    Updated the manual, in the chapter in the Physics of xstar,
    describing in more detail the radiation transfer algorithm.
     - Efficiency changes (to eliminate repeated reading of leapsecs file)
    fcollect (xtefilt):
     - Fixed problem that caused xtefilt/fcollect to fail for data taken on 29
       February 2000.
     - Portability-related changes for 64-bit Linux and Intel/Mac
     - Portability-related changes for 64-bit Linux and Intel/Mac

    Help lines: FTOOLS or xanprob@athena.gsfc.nasa.gov

    Web page maintained by Bryan K. Irby

    HEASARC Home | Observatories | Archive | Calibration | Software | Tools | Students/Teachers/Public

    Last modified: Tuesday, 23-Oct-2007 16:44:01 EDT

    The HEASARC welcomes your participation in a brief survey to capture how users access and utilize HEASARC data, software, and services. The outcome(s) of this survey will be used to guide, prioritize, and plan our activities and development in the coming years. It contains 18 questions, generally takes just a few minutes to complete, and your answers will remain totally anonymous. We thank you in advance for your valuable feedback.