
Delivered-To: oxborrow@dsri.dk
Date: Tue, 7 Jan 2003 14:18:56 +0100 (MET)
From: Stefan Larsson <stefan@astro.su.se>
Subject: Anodes dead or alive
To: oxborrow@dsri.dk, njw@dsri.dk
Cc: stefan@astro.su.se, peter.kretschmar@obs.unige.ch
MIME-Version: 1.0
Content-Transfer-Encoding: QUOTED-PRINTABLE
Content-MD5: nz6w4gfFRxos32dnh/VONQ==
X-Virus-Scanned: by amavisd-new with f-prot at smtp.su.se

Hi Niels Jrgen & Carol Anne,

I am coming back to the issue of dead anodes in data reductions.
We decided that these should be indicated with the STATUS flag
but as I said in a mail (early Dec) that does not solve the problem
for me. When I start to think about this there are more and more
questions and thoughts popping up. I would appreciate any
comments from you to guide my further thinking. I must 
admit that the following is a list of rather loose thoughts.
(Added at the end a few relevant emails from a month ago)

1. How do we localize a dead anode and specify its position?
   RAWX, DETX or some model fit?
   I assume that position corrections will be wrong near a
   dead anode?
   What about the onboard corrections? They will also be wrong
   near dead anodes. 
   Are there plans for changes of those onboard corrections?
   A diagnose/recalibration by observing without onboard correction?
   (or corrections specificly created to localize the anodes in th
   pixel binned data)
   Could that allow a more accurate localization of anodes?

2. Application:
   We talked about indicating the locations of dead anodes in
   the detector map. That map is in pixels (uncorrected) and the event
   positions are in mm (corrected).
   Can I simply "backcorrect" the event positions by taking out the
   correction from JMXi-CORX-MOD?
   Is that the ONLY position correction done?
   (does it e.g. include the CDELT1 (or 2) scaling factor in 
   JMXi-DETE-MOD?)
   Are the calibrations source positions in -DETE-MOD directly comparable
   with the corrected event positions?

3. Is the detector pixel map a good localization of the anode positions?
   How accurate can one expect it to be? Can we do better with something
   else?


Stefan

> Delivered-To: stefan@astro.su.se
> Date: Thu, 5 Dec 2002 09:38:42 +0100 (MET)
> From: Carol Anne Oxborrow <oxborrow@dsri.dk>
> Subject: Re: Dead Anodes
> To: oxborrow@dsri.dk, jerome@dsri.dk
> Cc: peter.kretschmar@obs.unige.ch, stefan@astro.su.se, 
smaisala@astro.helsinki.fi, njw@dsri.dk, silvia@dsri.dk, jerome@dsri.dk, 
nl@dsri.dk, sb@dsri.dk, cbj@dsri.dk
> MIME-Version: 1.0
> Content-MD5: 0Sl+59gTCVSCd1Acz4kMDw==
> 
> Dear Jerome,
> 
> I suspect there is not a direct one-to-one mapping between RAWX and anodes - 
> what I've sent round are pixel Xpositions (RAWX) that suddenly show a drop,
> whereas the hardware guys tend to work either in cathode position in mm or in
> anode numbers. I was hoping that there was some clever person out there who
> knows how to change between the these position conventions. If not I'll have 
to
> do it myself, which means it won't get done anytime in the next few days.
> 
> Can anyone else help with this problem?
> 
> Best wishes,
> 
> Carol Anne
> 


> Delivered-To: stefan@astro.su.se
> Date: Wed, 4 Dec 2002 14:09:45 +0100 (MET)
> From: Carol Anne Oxborrow <oxborrow@dsri.dk>
> Subject: Re: STATUS flag
> To: stefan@astro.su.se
> Cc: peter.kretschmar@obs.unige.ch, stefan@astro.su.se, 
smaisala@astro.helsinki.fi, njw@dsri.dk, silvia@dsri.dk, jerome@dsri.dk
> MIME-Version: 1.0
> Content-MD5: SQGcvH3DNO9w4MmM0MiVpg==
> 
> Dear Stefan,
> 
> all the pixel-by-pixel information is in JMXi-DETE-NOD, which is in IMOD 
group.
> This is the table I use to check positions against. In theory, once I've 
raised
> the various status flags for each individual event, there's no need for anyone
> further down the pipeline to correlate the events to the DETE table again, but
> you can always look in there is you really feel the urge.
> 
> Best wishes,
> 
> Carol Anne


> Delivered-To: stefan@astro.su.se
> Date: Wed, 04 Dec 2002 14:03:26 +0100
> From: Stefan Larsson <stefan@astro.su.se>
> Subject: STATUS flag
> To: njw@dsri.dk, silvia@dsri.dk, juhani.huovelin@astro.helsinki.fi, 
smaisala@astro.helsinki.fi, peter.kretschmar@obs.unige.ch, oxborrow@dsri.dk, 
jerome@dsri.dk
> MIME-version: 1.0
> Content-transfer-encoding: 8BIT
> 
> Hi SDAST,
> 
> Although I was strongly in favour of our decision on the
> STATUS flag I have realized it does not provide me with
> all the information I need. By connecting the information 
> of e.g. dead anodes with events I can not find dead anode
> areas without events. With the count rates we have it is 
> possible to have zero counts in a pixel (in particular
> at a dead anode!) and then I can not tell that this is
> a pixel I should exclude. The pixel-by-pixel information
> is something I need when I compute the open and closed
> detector areas (mask open fraction) and when I apply a
> background model which also is defined in terms of detector
> area. 
> If I need to read this information my self is that
> all in the instrument model? 
> Are dead anodes defined in terms of corrected or uncorrected 
> coordinates? From Peters arguments I think we concluded that
> the latter is physically more correct but is this also so in
> practise?
> 
> The last questions are probably for Niels Jrgen and Carol Anne
> but the rest of you are welcome to comment as well.
> 
> 
> Stefan


