Past Calorimeter Software Improvements

1
Past Calorimeter Software Improvements

• p11
• Non-linearity corrections
• Offline zero suppression (but lt not lt)
• Tweak of calorimeter layer weights
• Correction of ICD layer map
• Added better noise simulation file (symmetrized
  in phi)
• p13
• Further corrections for electronic weighting
  factors in ICR (5pF? 10pF)
• ICD weight increased 35
• Rewrite of pileup noise simulation
• GeV ? Raw ADC
• Individual noise channels used in MC
• Can add raw data or simulated noise
• Same noise file used for both simulation (LADC)
  and zero suppression (RADC)
• Correct offline zero suppression cut
• Correct switch of gain x8/x1 in MC and add
  persistence for all levels
• Apply zero suppression to MC

2
p13 Leftovers

• p13.00-07 packing at d0gstar was 16-bit
  corrected in p13.08
• Raw2Sim packing also corrected
• Do we need 32-bit information at DST level?
• Cleanup of weights kept in two separate
  packages (caldata and caltables)
• Maintenance nightmare?
• Ensure unpacking MC data is consistent with
  pileup
• NLS is not tracked correctly
• ICD channels need correction factor for 22pF/5pF
  from calibration
• Ongoing work in online software
• Cal_examine
• Cal_elec
• Hot cell killer
• L3CalCalib
• Pedestal download
• Online database storage

3
Future developments

• Obtain zero suppression thresholds, NLC
  coefficients, offline pedestal and gain
  corrections from calibration database
• Need to track stability
• Provide persistent data structures for all
  packages to use
• Allow for different weighting of layers (and
  etas?) for zero suppression
• CH layers may need different suppression
• Higher suppression at high etas suggested (eta
  gt 3.2)
• History tracking for different reco versions
  rather cumbersome to maintain
• Channel to channel calibration corrections for
  ICD
• NADA style block for T42 scheme
• Implement CH-seed weights
• Consistent handling of CalDataChunk for all
  object reconstruction

4
Improving Calorimeter Data

• Error checking
• ADC readout error flags
• Timeouts
• Event mixing between crates
• Set error flag on event
• Can flag bad data blocks if too high a rate for
  data quality?
• How does this affect luminosity?
• Data Quality and stability improvements
• In-spill unsuppressed data
• Useful for online monitoring
• Preprocess to create offline pedestal correction
  and kill hot-cells before reco
• needs to be in DB?
• What rate is needed and impact on online?
• Calibrate individual SCA channels and correct?
• Use d0cert packages in offline farms
• Need quick feedback back into the online