Log entry time 09:46:32 on April 29, 1998

Entry number 120

From volmer@cebaf.gov Wed Apr 29 09:30:43 1998
Date: Tue, 28 Apr 1998 17:37:14 -0400 (EDT)
From: Jochen Volmer 
To: Vardan Tadevosyan ,
    Ketevi Assamagan , Dave Mack ,
    Rob van der Meer , Kelley Vansyoc ,
    henkb@nikhef.nl
Cc: Jochen Volmer 
Subject: replay_2

Gents,

it looks like I have a reasonably stable replay_2 engine now. I will
delete all rzdat-files, hbook-files, paw*kumac-files in $WORK/replay_2/paw
and the outfiles tonight and start a new run-through on the farm Linux and
Sun machines.

Specifications:

- To preserve disk space, I implemented a cut on the HMS coincidence
  timing which only takes effect on the ntuples, but not on the hbooks.
  Here I cut off all events with ccointime_hms<-5.0 ns. This will get rid
  of most of the protons, but not of all of them.

- Matrix elements, offset files and detector calibration files are
  up-to-date.

- Software Upgrades:
  - Kelley's Energy loss correction effects h_physics, s_physics,
    c_initialize, total_eloss
  - Henk's c_physics with new definition of the lab frame, and some
    stabilization
  - Vardan's s_correct_cal to correct for the wrong sign of the
    attenuation correction
  - Dave McKee's h/s_dump_tof, h_tof to correct for the flake scintillator
  - Kelley's hms/sos_calorimeter.cmn, h/s_tracks_cal, to disable the
    calorimeter tracking fiducial cut. We have got fortran files that
    will do the job within PAW. You can find them in
    /home/volmer/PAW-fortrans/h/scal_fv.f. On information how to use them,
    please contact me directly.
  - c_keep_results to reduce the size of the ntuples.
 
Important notes for the analyzer:

- The timing is stabilized for most runs. This will make timing cuts a lot
  easier and more reliable. The real coincidence timing peak should now be
  located at ccointime_hms=0.0 (cointime=0.0 in the ntuple).
  For this replay, I cut out all events with ccointime_hms<-5.0 to save
  space. A survey of the coincidence timing for all replay_2 runs should
  be done to determine runs where the correction is not good enough.

- In the coincidence ntuples, up till now the HMS and SOS SHTRK and PRTRK
  had been E1 and ET, that is not normalized on the particle momentum. In
  this version HSSHTRK, HSPRTRK, SSSHTRK and SSPRTRK will actually be the
  normalized versions. A survey of these quantities should be done.

Regards,

	Jochen