I am glad that we now have alignment offsets that are likely to represent
the real offsets within the DC system.
I didn't recognize the decision on doing all tracking in the Hall Sector
Coordinate System.
It's actually much easier to do the tracking in the TSCS (as before)
and - if you want - transform the results into the HSCS or HCS.
The most important part of the tracking is the simulation / reconstruction
of the track WITHIN THE MAGNETIC FIELD. The field map is defined in the
TSCS, so all track swimming is done in this coordinate system (up to now:
modulo a rotation of the coordinate axes (=Bogdan's system, that may
easily be changed)). To do all the tracking in the HSCS means to
transform the data achieved in every 'swimming' step from TSCS to HSCS.
Isn't it more convenient to transform the DC geometry data into the TSCS
which is allmost the same as HSCS except the torus misalignment.
Please take these remarks into consideration when changing the GSIM,
RECSIS geometry.
Franz
On Mon, 26 Jan 1998 SCHUMACHER@ernest.phys.cmu.edu wrote:
...
> Off-line Folks:
>
> As of this afternoon (2pm 1-26-98) the CLAS drift chamber
> alignment using survey data is in place in the generally-used map
> (DC_GEOM.map). To use this information successfully you need to
> remake your Recsis, after checking out or updating the dc area in your
> workspace. The alignment changes have been tested over the weekend
> (by Stepan) and have been shown to significantly improve our W
> resolution, as well as removing some false correlations seen in the
> data. Appended is the full text of the CLAS_note describing what was
> done. Read it if you care about the details of drift chamber alignment.
...
>
> ISSUES IN USING THE ALIGNMENT DATA AS GIVEN:
>
> We have recently decided to do all tracking in the Hall Sector
> Coordinate System (HSCS). Thus, RECSIS takes the numbers given below
> in the HSCS and translates and rotates each of the sectors away from
> their ideal positions accordingly. The numbers given are in every
> case the displacement or rotation needed to move the object in
> question from its ideal location to its measured location in the
> HSCS. While displacements and offsets are given for the torus, they
> are presently not used in the tracking. Thus, we are ignoring the
> roughly 1 mm shifts in the torus centroid, as well as the 4 mrad
> rotation of the torus. We presently believe that neglect of these
> effects is negligible in terms of the results of tracking. The TSCS
> and TCS are not used.
^^^^^^^^^^^^^^^^^^^^^?? really not used?
> Results returned by the tracking code is
> therefore also in the HSCS or the HCS, as appropriate, for direct
> matching to the outside detectors.
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^that's ok.
>
...