Date: Tue, 17 Sep 2002 11:07:19 +0100 (BST)
Cc: firstname.lastname@example.org, email@example.com, firstname.lastname@example.org
Subject: Response to SSC SPR SSC-SPR-2977
Your SPR entitled
"attcalc does not properly put sky coords of source image into background image"
has been rejected. Please see the response section for further details.
SPR no.: SSC-SPR-2977
Date: 6th September 2002
Task: attcalc (version 4.27.1)
Title: attcalc does not properly put sky coords of source image into
Originator: P. Rodriguez on behalf of D. Buote (email@example.com)
I am analyzing an extended source & would therefore like to
use the standard background templates provided by D. Lumb
for background subtraction. I would like to do this in sky
coordinates if possible. As Pedro explained to me, it is
possible to fill the (X,Y) columns of the background template using
which I have attempted to do with the following command:
buote@nephos: attcalc eventset='temp_mos1_bkgt.evt'
attitudelabel='fixed' refpointlabel='user' withatthkset=no
atthkset='atthk.dat' withmedianpnt=no imagesize='0.36'
nominalra=1.98850000500000E+02 nominaldec=-1.63850000000000E+01 -w 5
5 fixedra=1.98844293987737E+02 fixeddec=-1.63573613020474E+01
Here "temp_mos1_bkgt.evt" is just the mos1 background
template with standard screening. The settings for some of the other
where the RA_PNT etc. are taken from the source events
The problem is that these keywords are not propagated
exactly the background events file. As Pedro explained
to me, attacacl adds new keywords (ATTSRC, ATTRA,
ATTDEC, and ATTPA) from which the following keywords are
TCRPX, TCRVL, TCDLT, TLMAX, TLMIN, TDMAX, TDMIN
It is these keywords that must be the same for the background and
image file in
order for the "image"
coordinates to be exact. However, this is not what I find:
I do find the following exact correspondences between the source
events file and
the background events file after performing attcalc.
RA_PNT ==> ATTRA
DEC_PNT ==> ATTDEC
PA_PNT ==> ATTPA
But the other keywords are not in general identical:
TCRPX, TCDLT are identical in both events files.
But the following are not:
TCRVL6 = 1.98844293987737E+02 / [deg] coord. at X axis ref. pixel
TCRVL7 = -1.63573613020474E+01 / [deg] coord. at Y axis ref. pixel
TCRVL6 = 1.98850000500000E+02 / [deg] coord. at X axis ref. pixel
TCRVL7 = -1.63850000000000E+01 / [deg] coord. at Y axis ref. pixel
TLMAX4 = 19910 / Legal maximum for DETX
TLMAX5 = 19820 / Legal maximum for DETY
TLMAX4 = 19905 / Legal maximum for DETX
TLMAX5 = 19814 / Legal maximum for DETY
TLMIN4 = -19768 / Legal minimum for DETX
TLMIN5 = -20305 / Legal minimum for DETY
TLMIN4 = -19762 / Legal minimum for DETX
TLMIN5 = -20299 / Legal minimum for DETY
TDMAX4 = 19905 / Actual maximum for DETX
TDMAX5 = 19704 / Actual maximum for DETY
TDMAX6 = 43623 / maximum projected image X axis data
TDMAX7 = 43120 / maximum projected image Y axis data
TDMAX4 = 19910 / Actual maximum for DETX
TDMAX5 = 19710 / Actual maximum for DETY
TDMAX6 = 49042 / maximum projected image X axis data
TDMAX7 = 44820 / maximum projected image Y axis data
TDMIN4 = -19768 / Actual minimum for DETX
TDMIN5 = -20195 / Actual minimum for DETY
TDMIN6 = 1906 / minimum projected image X axis data
TDMIN7 = 3366 / minimum projected image Y axis data
TDMIN4 = -19762 / Actual minimum for DETX
TDMIN5 = -20189 / Actual minimum for DETY
TDMIN6 = 8742 / minimum projected image X axis data
TDMIN7 = 8936 / minimum projected image Y axis data
OK. I assume these keywords differences account for the problems I
trying to associate extraction regions for both the source and
images. Let me summarize my problem:
My goal is to define a region on an image and then extract a spectrum
region. The important thing is that I DO NOT WANT TO USE XMMSELECT --
I want to
use my own programs to define the region of interest.
I want to define an extraction region on the
source image and have that region apply also to the background image
(X,Y) coordinates have been filled using attcalc as we have
I have created a circular region on the source image and have saved
to an ascii region file, src.reg, using ds9. I have saved this file
"image" coordinates and "SAOIMAGE" format.
Here's my problem: If I display the background image in ds9 and then
src.reg, it does NOT correspond to exactly the same region in
coordinates that it did on the source image. It is close (maybe about
displaced), but it is not exactly the same.
I assume these differences arise because of the keyword
differences noted above.
Message of length 5685 truncated