Diffs from version 3.2 to 3.4
The following features have been implemented related to ISAAC support:
- jitter can now use more than one bright object to do
cross-correlation. A dedicated section (DetPeak) is now in the
jitter.ini file to set this up.
- New command called isaac_cal, used exclusively for
ISAAC spectroscopy data reduction. This model is used as a first
estimate to compute a calibration solution within MIDAS. This
utility is not an eclipse command indeed, it has been added
to this distribution for consistency. Actually, it can compile on
its own without need of the libraries.
- New command called is_illum to compute illumination
frames. Still in development...
See is_illum man page
- New command called is_zerop to compute zero points
automatically from the associated ISAAC data.
See is_zerop man page.
- New command called is_twflat to compute ISAAC twilight
flat fields. Flat-field construction is now robust to sources in the
See is_twflat man page and
flat man page.
- All files reduced for ISAAC now produce headers with the right
FITS keywords for later use (PRO dictionary).
There is new code to handle FITS files! Reading FITS keywords
actually does not need all the knowledge of what can be in the FITS
format. A few system calls do the job quite well to retrieve values,
now the accesses are really fast. The previous code to access FITS
values and change them is still there, but any routine needing only
read-only access is now faster by magnitudes!
Many bugs have been corrected in the adonis-related routines,
thanks to a tremendous help effort from Eric Gendron in
Observatoire de Paris/Meudon. Thanks for this help, it is most
fitsort now knows how to align its outputs, by adding
blanks to fit the sizes and make them readable for a human being.
Any software that reads tables in ASCII format should still be able
to cope with the outputs of fitsort. Have a look at the
fitsort manual page.
New command called peak, that is meant to supersede one
day all detpeak, fwhm, and photfits
commands. In one command, you can detect bright objects, compute
some basic statistics (min, max, mean, median in a zone around each
object), compute FWHMs, photometry (with background subtraction),
and why not Strehl ratio. The command still needs some testing, do
not hesitate to report bugs ASAP! Have a look at the
peak manual page.
flat now features a robust method to get rid of star
signal in twilight frames. The linear fit is now rejecting outliers
caused by star signals, in a very reliable fashion. Computation time
has jumped from a few seconds to 15 seconds per frame (5 minutes for
20 input frames), which is still reasonable given the amount of
needed computations. You can also request a flat-field fitting after
dark subtraction, in which case you only need to fit a slope and not
a full linear parameter set (y=ax+b). It is much faster, but of
course depends on the quality of the dark you provide for
subtraction. flat has an ISAAC alias: is_twflat
which is the same executable with some FITS keyword handling added
for the ISAAC case. Have a look at the
flat manual page and the
is_twflat manual page.
There are no more verbose, debug, or memory
option on any eclipse command. All of these are now set
through environment variables which need to be set on your user
account. See the installation procedure in the distribution to know
how to set your account to work with eclipse from now on.
- New command called std. It is actually a standard star
database browser. The database is internally hardcoded in the
program itself, it will be later offering the possibility to load a
database from an external ASCII file or a URL. You can browse the
stars by position, name, magnitude, etc. Have a look at
std man page
Back to eclipse home page