STSDAS Version 3.12 Release Notes
June 2010
This release of STSDAS contains changes to the pipeline code for nearly
all the instruments (ACS, STIS, NICMOS, WFC3 and COS). In addition,
the final software used by the ST-ECF to calibrate FOS data has been
included strictly as a record of the processing performed to generate
the FOS products available from the HST archive.
The specific changes are detailed in the following sections.
Platform Support
================
Binaries for this release were built on Red Hat Enterprise Linux 4
and 5, Solaris 5.8, and Mac OS X 10.4 (PPC and Intel architecture).
Binaries were built with IRAF 2.14, except for Solaris, which were
built with IRAF 2.12.2a. IRAF is available from http://iraf.net.
SYNPHOT
=======
There are no changes to the SYNPHOT software in this release.
SYNPHOT users should be aware that we are presently developing
the pysynphot package as a successor to SYNPHOT. Pysynphot is a
re-implementation of the SYNPHOT functionality. It is not a simple
port into Python, but provides improved algorithms that address known
shortcomings in SYNPHOT. It does not require IRAF, and is suitable
for either library or interactive use.
An alpha version of pysynphot (v0.80) is included in this release that
provides nearly all of the core SYNPHOT functionality. Once the
core functionality is supported (presently planned for Spring 2010),
SYNPHOT will be formally deprecated.
Additional details about pysynphot can be found in the release notes
for the STScI_Python package.
HST_CALIB
=========
STPOA
-----
This package represents the final state of the software used by
the Space Telescope European Coordinating Facility (ST-ECF) to
produce the statically archived FOS products and has been included
strictly as a record of the final calibration processing that was
applied to FOS data. There is no expectation that this software will
compile and run, but it does provide useful insight into how the FOS
data was calibrated for those working with that data. This package
includes a final version of CALFOS which supercedes the version
in HST_CALIB.FOS.
FOS
---
This software is being included as a record of the software used at
STScI to initially calibrate data from the FOS. The version of CALFOS
in this package has been superceded by the version in the STPOA package.
ACS
---
CALACS Version 5.1.1:
The following changes were made to CALACS to support
ACS data after its repair during HST Servicing Mission 4.
- This version of CALACS implements the changes to the ERR array
initialization for post-SM4 ACS data based on the algorithms provided
by the ACS Calibration team at STScI. Basically, if the image is
not a bias frame, the Poisson contribution to the ERR array should
be determined by differencing the associated superbias from the SCI
array, rather than subtracting a constant header-derived bias level.
- Updates were made to ACSREJ to add a new parameter (newbias)
to support processing of post-SM4 bias images. Changes were also
made throughout CALACS to automatically determine this parameter
value and pass it to ACSREJ, which used this value to change how
the error arrays are computed for CRJ products created with post-SM4
bias images.
- The interpretation of the crsigmas parameter for ACSREJ was updated
to use the latest version of 'strtor' from CALSTIS and to wrap the
input from the user in single-quotes.
- Specification of include file was added to allow CALACS to compile
properly under Solaris and Mac OS X.
- Corrected logic in 'donoise.c' to properly handle BIAS frames
taken pre-SM4.
MultiDrizzle
------------
- Help file for MultiDrizzle had a few inaccuracies cleaned up.
- Default parameter types for outnx and outny parameters corrected to be integer instead of real.
NICMOS
------
CALNICA Version 4.4.1:
- When calnica did the dark correction with a *tdd.fits file that
has the TFBLOW & TFBHIGH keywords in EXT=0, but lacks these two
keywords in EXT=1 and EXT=2, then the wrong shading correction
was being applied. This was corrected. Bad dark imset numbers
were also being reported and were related to this problem, this
has been fixed as well
CALNICB Version 2.6.3:
- Updated CALNICB to update the logic to handle pattern positions
without any members. Logic was added to 'n_combniter.c' to account
for a pattern position having 0 members when updating the DQ flags
STIS
----
CALSTIS version 2.30:
- A function that compared the values of keyword CCDAMP in the files to be
combined for cosmic-ray rejection was actually getting default values, so
the intended consistency check was not being done.
- A variable used during interpolation of time-dependent sensitivity values
was not being initialized. This would not have resulted in incorrect
results, just a slower start-up.
- Parameters that are used in scaling CCD dark images depending on detector
temperature are now gotten from the header of the dark reference file,
rather than being hard-coded. This change will be needed for calibration
of post-SM4 CCD data with new dark files.
COS
---
x1dcorr:
- A new task to re-run the 1-D extraction step independently of calcos.
The input is a corrtag file, and the task creates new flt and counts
files as well as the x1d file.
splittag:
- A new task to split a corrtag file into several output files of the
same format, based on time. The input file may be split at fixed
intervals of time, or the user may specify an arbitrary set of times.
The 'x1dcorr' task can then be used to extract a 1-D spectrum from
each of the output files.
WFC3
----
CALWF3 Version 2.1:
- Fixed the calculation of the IR "sampzero" (zeroth-read effective exposure
time) for IR sub-array exposures, so that the value is always recomputed
regardless of which processing steps are set to perform.
- Updated the IR FLATCORR routine to use the mean gain of all 4 amps when
applying the gain conversion to grism exposures. Now all images, direct and
grism, use the mean gain instead of quadrant-dependent gains.
- Wf3rej has been updated to only issue a warning message instead of stopping
when the spt files associated with the input images are not present.
- Dummy "drz" files are no longer produced by calwf3, because real ones are
now produced by MultiDrizzle.
- The IR ZSIGCORR routine has been modified to be more robust in catching
and flagging saturated pixels in the zeroth and first reads of an exposure.
- The computation of IR ERR values from up-the-ramp fitting has been corrected
to properly account for read noise and Poisson noise. Previous calculations
produced a severe underestimate.
- Wf3rej has been updated to properly parse embedded blanks in the "crsigmas"
parameter string.
- The order of IR processing has been modified so that NLINCORR precedes
DARKCORR, ensuring that the non-linearity correction and saturation checking
are performed before the pixel values are modified by dark subtraction.
- Wf3rej has been modified to ensure that the DQ flags associated with
detected CR's get propogated to output _flt files.
- The IR DARKCORR routine has been updated to compute and populate the
MEANDARK keyword for each readout.
