In response to Pat Broos' comment on using CIAO2 with the
old data, I have put together a test tar file with geometry
data corresponding to the old pixlib "pix_sim_table_flight<date>thru<date>.par"
files. It's in
http://hea-www.harvard.edu/~jcm/asc/dist/geom/old_processing.tar
and it has a README to tell you what to do (one pset, it's not challenging).
I haven't tested these files beyond verifying that they contain the values
I expect, and I won't have time to do so for a while,
but I thought since there are smart people out there who want this (hi Pat,
Roderick) I would make it available right away so that if you want to
you can try it out and give me feedback (e.g., it reproduces what you were
getting with CIAO1, or alternatively it gives results that look suspicious
to you...).
I include the README below.
Jonathan
==================================================
You can run into problems with coordinates using data processed early
in the mission. You must use geometry data consistent with the
position of the optical axis assumed in the aspect solution.
If you have old data, this means you have to replace the standard
geometry file with an alternative one containing the old, incorrect,
optical axis positions.
This tar file contains two replacement geometry calibration
files, tel*.fits, and two replacement geom.par files
to illustrate their use.
The old file is
$CALDB/data/chandra/tel/bcf/geom/telD1999-07-23geomN0002.fits
We don't want to replace it in CALDB, we just want to tell
CIAO's pixlib library to find the new file on this particular
occasion.
To use the file
telD1999-07-23geomN0002_T2.fits
which you've put in your directory /mydir, just do
pset geom instruments="/mydir/telD1999-07-23geomN0002_T2.fits"
and then
plist geom
to check that the pset worked ok. Look at geom2.par for what
your parameter file should roughly look like - only
the first parameter has been changed, that's where all of the
telescope geometry info is. All the other parameters will still
be set to "CALDB" and will pick up the standard file.
When the exposure map code (for instance) calls pixlib, pixlib
opens the geom.par file and finds your new geometry file.
NOTE 1:
The "T1" file is for data processed from 1999 Jul 23 to 1999 Nov 29.
The "T2" file is for data processed from 1999 Nov 29 to 2000 Mar 1.
NOTE 2:
WHAT COUNTS IS THE ***PROCESSING*** DATE
**** NOT !!!!! ****
REPEAT NOT THE ***OBSERVATION*** DATE.
If you have data observed by Chandra in 1999 but REPROCESSED since
Mar 2000, you don't need to and indeed MUST NOT use these files,
just use the standard ones.
NOTE 3:
Even if you have data from an old processing, you shouldn't really
use these files. Instead, you should throw your old data away
and get the new processing. Or, just get the new aspect solution
and/or aspect offsets files to make your exposure maps with. (the detector
coordinates in your old event file will be wrong but for many purposes
you won't use them anyway).
For problems, contact Jonathan McDowell (jcm@cfa.harvard.edu)
This archive was generated by hypermail 2b29 : Tue Feb 12 2013 - 01:00:04 EST