Difference between revisions of "BCAL Reconstruction Meeting 2015-11-12"

From GlueXWiki
Jump to: navigation, search
(Agenda)
m (Text replacement - "http://argus.phys.uregina.ca/gluex" to "https://halldweb.jlab.org")
 
(7 intermediate revisions by 5 users not shown)
Line 38: Line 38:
 
# Commissioning Efforts
 
# Commissioning Efforts
 
## Effective speed (George)
 
## Effective speed (George)
## Time-walk corrections [http://argus.phys.uregina.ca/gluex/DocDB/0028/002884/001/HighPP.pdf (Noemi)]
+
## Time-walk corrections [https://halldweb.jlab.org/DocDB/0028/002884/001/HighPP.pdf (Noemi)]
 
## Time-walk parameterization (Elton)
 
## Time-walk parameterization (Elton)
 
##* [[Media:TAGH Twalk 110915.pdf | Tagger Hodoscope (Nathan)]],  See also [http://www.jlab.org/Hall-B/notes/clas_notes02/02-007.pdf CLAS-NOTE 2002-007]
 
##* [[Media:TAGH Twalk 110915.pdf | Tagger Hodoscope (Nathan)]],  See also [http://www.jlab.org/Hall-B/notes/clas_notes02/02-007.pdf CLAS-NOTE 2002-007]
Line 45: Line 45:
 
# Calibration
 
# Calibration
 
## Cosmics (Andrei)
 
## Cosmics (Andrei)
 +
### Attenuation lengths [https://logbooks.jlab.org/entry/3351612] (Mark)
 
## Muon decays  
 
## Muon decays  
### Data analysis (Will)
+
### Data analysis [https://logbooks.jlab.org/files/2015/11/3359798/nov12_cal_meeting2.pdf] (Will)
 
### Simulations (Andrei/Irina)
 
### Simulations (Andrei/Irina)
## Calibration Challenge
+
## Calibration Challenge (All)
##* From Sean: 'Just a reminder to please add the "neutral" starting values for the tables involved in this exercise to the main CCDB, under the "calib" variation, run 0. '
+
##* From Sean: ''Just a reminder to please add the "neutral" starting values for the tables involved in this exercise to the main CCDB, under the "calib" variation, run 0. ''
 +
##* What, When and Who?
 
# Reconstruction
 
# Reconstruction
 
# Any other business
 
# Any other business
Line 55: Line 57:
 
= Minutes =
 
= Minutes =
  
''Attendees:  ’’
+
''Attendees:  Elton, Simon, Mark, David (JLab), Curtis, Mike, Will (CMU), George (Athens), Sean (NWU), Ahmed, Tegan, Zisis (Regina)’’
 +
 
 +
# Announcements: Accelerator is on track, beam to Hall D on December 4.  Magnet: final stages of leak checking, Helium system today or tomorrow, then N2 system which should be simpler.  Then the super-insulation, etc.  Cool down 1st week in December.  N2 will be connected back to BCAL whenever they don't need it anymore for the magnet.
 +
# Action Items
 +
## Emulation: David/Mike have developed preliminary plan on how structure will work, for at least the 1st attempt.  New factory for emulation objects, handling ok in DAQ code and get constants in DB. Long project. 
 +
## Sampling fraction implementation will commence soon.
 +
# Commissioning Efforts
 +
## Effective speed (George): Nothing new.  Cosmics+PTRACKFIND extracted values are different from cosmics+No_field values:  16.5 vs 17 cm/ns.  The first value is what we get for the 1st layer both from spring data runs with B field = ON and from cosmics when PTRACKFIND flag is used. The 17cm/ns is what we get from cosmics when the No_Field option is used instead of PTRACKFIND, as well as a strict cut on the data. Which are the correct values?  If we choose to "trust" the data from the 1st layer as the most accurate ones then other layers may need extrapolation, since there is no way to get rid of the layer dependence effect (up until now). Layer 1 is very stable for all cells. Still investigating.
 +
## Time-walk corrections [https://halldweb.jlab.org/DocDB/0028/002884/001/HighPP.pdf (Noemi)]  New function used (with an additional term) a la Nathan from Tagger TDC and link above from Hall B work.  Top right plot is with Mike's function and bottom right plots with new function; the latter seems to improve things at higher ADC values.  Work is worth it if it is not a huge overhead in constants for CCDB.  Elton: The function should be a function of ADC/threshold so this is why the 14 is included.
 +
## Time-walk parameterization (Elton): Noemi should look into these references below from the CLAS note.  The function and its derivative at the transition point lead to a new parameter.
 +
##* [[Media:TAGH Twalk 110915.pdf | Tagger Hodoscope (Nathan)]],  See also [http://www.jlab.org/Hall-B/notes/clas_notes02/02-007.pdf CLAS-NOTE 2002-007]
 +
## Layer efficiencies (Ahmed): picking up the thread.  Checking mod7 and mod8 differences.  In progress.
 +
## Up/down gain ratios for spring data (Mark): mostly done.  This will be removed from the list. See [https://halldweb.jlab.org/wiki/index.php?title=BCAL_Reconstruction_Meeting_2015-11-12&action=edit log entry] Gain ratios have been passed on to Will.  Full consistency between cosmics and data is now possible that a full replay has been done since Collaboration meeting, but has not been done yet.
 +
# Calibration
 +
## Cosmics (Andrei): Irina found memory issues in Geant simulation and re-simulating data.
 +
## Muon decays
 +
### Data analysis (Will): first plot is from data. 2nd plot: top one has points+hits, bottom one only points.  Page 3: investigate possible miscalibration of channels, but definitive conclusions are hard to draw; there is a phi dependence (only max shown here).  Next, Elton's ratio method in plot 4 (link:  [https://mailman.jlab.org/pipermail/halld-cal/2015-November/001172.html e-mail]).  Plot has points and hits. y has total energy, x is most energetic/ratio.  This plot is from cosmics run.  Why does the left not align with the right half?  The 0.5< are from 3 hits or more, so this leads to missing energy.  0.5 is 15MeV, and threshold is 2.5 MeV plus some attenuation factor. It would be interesting to see these in MC.  Compare to plot on p 7.  Last plot: there could be overlap between hits and points sets.  Plot it in 2D to see correlation.  Also, plot man vs total instead of using the ratio.  The key will be to understand the thresholds (which also change with time and channel).
 +
### Simulations (Andrei/Irina): nothing new; same problems as above.  Total energy plot (which caused the memory leaks) will be added to the fiber energy deposition. 
 +
## Calibration Challenge (All)
 +
##* From Sean: ''Just a reminder to please add the "neutral" starting values for the tables involved in this exercise to the main CCDB, under the "calib" variation, run 0. ''
 +
##* What, When and Who?  The tables are the timing offsets for TDC and ADC, time walk, and attend parameters.  Mike will handle the timing ones.  For the ADCs we want the gains; we could use the average for each layer and same for attenuation lengths. However, we should prepare for a run-by-run basis mechanism.  For BCAL is it a calibration thing or monitoring thing. The position is reported by the points, so we need to know which parameter extractions need a starting value and which do not.  Do we need to compare extracted values to some "nominal", average values?  This particular plugin should be run after all the times have been calibrated.  The gains can come out from the first pass, and multiple iterations can be used to refine things.  Start with current values of calibration.  The gain calibration is the most sensitive extraction to the procedure.  Use current values for these, timing can start with zero, atten len can use average values.  LED data should be added as well for monitoring (Noemi?).  Shaun: deadline is Monday for timing ones (offsets and t-w).  Will to do current values of gain into CCDB. JLab will put atten len in CCDB.
 +
# Reconstruction
 +
# Any other business: No updates on DocDB migration to JLab.

Latest revision as of 12:40, 14 March 2017

Video Conferencing Information

Meeting Time: 11:00 a.m. EDT

  1. To join via a Web Browser, go to the page [1] https://bluejeans.com/907185247.
  2. To join via Polycom room system go to the IP Address: 199.48.152.152 (bjn.vc) and enter the meeting ID: 907185247.
  3. To join via phone, use one of the following numbers and the Conference ID: 907185247.
    • US or Canada: +1 408 740 7256 or
    • US or Canada: +1 888 240 2560
  4. Upon connection all microphones are automatically muted. To unmute your mike on a Polycom or equivalent unit, enter *4. Unmuting on a computer is trivial as there is a microphone button than can be clicked.
  5. More information on connecting to bluejeans is available.

Participant Direct Lines

  • JLab Phone: in CC F326 is 757-269-6460 (usual room)
  • JLab Phone in CC L207 is 757-269-7084
  • Phone in the Regina Video-conference Suite is 306-585-4204

References

  1. BCAL Reconstruction Issues
  2. BCAL Reconstruction Algorithms
  3. Will's reconstructed energy info

Action Items

  1. Use pTrackFind for comics analysis (George)
  2. FADC Emulation code in November (Mike?)
  3. Low priority items
    1. z-coordinate determination from up/down amplitude ratio
    2. Sampling fraction tables (Tegan, Andrei)
    3. Code Cleanup

Agenda

  1. Announcements
  2. Action Items
  3. Commissioning Efforts
    1. Effective speed (George)
    2. Time-walk corrections (Noemi)
    3. Time-walk parameterization (Elton)
    4. Layer efficiencies (Ahmed)
    5. Up/down gain ratios for spring data (Mark)
  4. Calibration
    1. Cosmics (Andrei)
      1. Attenuation lengths [2] (Mark)
    2. Muon decays
      1. Data analysis [3] (Will)
      2. Simulations (Andrei/Irina)
    3. Calibration Challenge (All)
      • From Sean: Just a reminder to please add the "neutral" starting values for the tables involved in this exercise to the main CCDB, under the "calib" variation, run 0.
      • What, When and Who?
  5. Reconstruction
  6. Any other business

Minutes

Attendees: Elton, Simon, Mark, David (JLab), Curtis, Mike, Will (CMU), George (Athens), Sean (NWU), Ahmed, Tegan, Zisis (Regina)’’

  1. Announcements: Accelerator is on track, beam to Hall D on December 4. Magnet: final stages of leak checking, Helium system today or tomorrow, then N2 system which should be simpler. Then the super-insulation, etc. Cool down 1st week in December. N2 will be connected back to BCAL whenever they don't need it anymore for the magnet.
  2. Action Items
    1. Emulation: David/Mike have developed preliminary plan on how structure will work, for at least the 1st attempt. New factory for emulation objects, handling ok in DAQ code and get constants in DB. Long project.
    2. Sampling fraction implementation will commence soon.
  3. Commissioning Efforts
    1. Effective speed (George): Nothing new. Cosmics+PTRACKFIND extracted values are different from cosmics+No_field values: 16.5 vs 17 cm/ns. The first value is what we get for the 1st layer both from spring data runs with B field = ON and from cosmics when PTRACKFIND flag is used. The 17cm/ns is what we get from cosmics when the No_Field option is used instead of PTRACKFIND, as well as a strict cut on the data. Which are the correct values? If we choose to "trust" the data from the 1st layer as the most accurate ones then other layers may need extrapolation, since there is no way to get rid of the layer dependence effect (up until now). Layer 1 is very stable for all cells. Still investigating.
    2. Time-walk corrections (Noemi) New function used (with an additional term) a la Nathan from Tagger TDC and link above from Hall B work. Top right plot is with Mike's function and bottom right plots with new function; the latter seems to improve things at higher ADC values. Work is worth it if it is not a huge overhead in constants for CCDB. Elton: The function should be a function of ADC/threshold so this is why the 14 is included.
    3. Time-walk parameterization (Elton): Noemi should look into these references below from the CLAS note. The function and its derivative at the transition point lead to a new parameter.
    4. Layer efficiencies (Ahmed): picking up the thread. Checking mod7 and mod8 differences. In progress.
    5. Up/down gain ratios for spring data (Mark): mostly done. This will be removed from the list. See log entry Gain ratios have been passed on to Will. Full consistency between cosmics and data is now possible that a full replay has been done since Collaboration meeting, but has not been done yet.
  4. Calibration
    1. Cosmics (Andrei): Irina found memory issues in Geant simulation and re-simulating data.
    2. Muon decays
      1. Data analysis (Will): first plot is from data. 2nd plot: top one has points+hits, bottom one only points. Page 3: investigate possible miscalibration of channels, but definitive conclusions are hard to draw; there is a phi dependence (only max shown here). Next, Elton's ratio method in plot 4 (link: e-mail). Plot has points and hits. y has total energy, x is most energetic/ratio. This plot is from cosmics run. Why does the left not align with the right half? The 0.5< are from 3 hits or more, so this leads to missing energy. 0.5 is 15MeV, and threshold is 2.5 MeV plus some attenuation factor. It would be interesting to see these in MC. Compare to plot on p 7. Last plot: there could be overlap between hits and points sets. Plot it in 2D to see correlation. Also, plot man vs total instead of using the ratio. The key will be to understand the thresholds (which also change with time and channel).
      2. Simulations (Andrei/Irina): nothing new; same problems as above. Total energy plot (which caused the memory leaks) will be added to the fiber energy deposition.
    3. Calibration Challenge (All)
      • From Sean: Just a reminder to please add the "neutral" starting values for the tables involved in this exercise to the main CCDB, under the "calib" variation, run 0.
      • What, When and Who? The tables are the timing offsets for TDC and ADC, time walk, and attend parameters. Mike will handle the timing ones. For the ADCs we want the gains; we could use the average for each layer and same for attenuation lengths. However, we should prepare for a run-by-run basis mechanism. For BCAL is it a calibration thing or monitoring thing. The position is reported by the points, so we need to know which parameter extractions need a starting value and which do not. Do we need to compare extracted values to some "nominal", average values? This particular plugin should be run after all the times have been calibrated. The gains can come out from the first pass, and multiple iterations can be used to refine things. Start with current values of calibration. The gain calibration is the most sensitive extraction to the procedure. Use current values for these, timing can start with zero, atten len can use average values. LED data should be added as well for monitoring (Noemi?). Shaun: deadline is Monday for timing ones (offsets and t-w). Will to do current values of gain into CCDB. JLab will put atten len in CCDB.
  5. Reconstruction
  6. Any other business: No updates on DocDB migration to JLab.