May 20, 2015 Calibration
GlueX Calibration Meeting
Wednesday, May 20, 2015
11:00 am, EDT
JLab: CEBAF Center, F326
Connection Using Bluejeans
- To join via Polycom room system go to the IP Address: 220.127.116.11 (bjn.vc) and enter the meeting ID: 630804895.
- To join via a Web Browser, go to the page  https://bluejeans.com/630804895.
- To join via phone, use one of the following numbers and the Conference ID: 630804895
- US or Canada: +1 408 740 7256 or
- US or Canada: +1 888 240 2560
- Upon connection all microphones are automatically muted. To unmute your mike on a Polycom or equivalent unit, enter *4
- More information on connecting to bluejeans is available.
- New Offline Wiki Page
- Calibration/Reconstruction Uncertainties
- Subdetector Reports
- Simulations (Sean/Mark I.)
- Calibration Processing (Sean)
Attending: Sean (NU); Curtis, Naomi (CMU); Mahmoud (FIU); Simon, Kei, Beni, Nathan, Adesh, Mark I., Elton, Eric, Alex B., Fernando, Luke, Eugene, Mike S., Justin, Alex S. (JLab)
- Mark Ito has reorganized the offline software wiki page, and asked for any comments. Naomi suggested adding links to information on making plugins and Subversion.
- Paul M. wanted to remind everyone to work on integrating uncertainties in measurements and calibrations into the reconstruction code and CCDB.
- Curtis asked if there were templates for calibration uncertainties in the CCDB. Sean said that some detectors have them, some don't. Experience has shown that the most effective procedure is to produce the results and a concrete plan for adding the constants to the CCDB.
Adesh reported that he is currently working on estimating the number of pi0's in data. The cuts he is currently using are very tight and he is investigating what happens when they loosen them. He also tried some of the suggestions made in the last calibration meeting for improving his fits to the pi0 spectrum, and is still finding a pi0 mass that is ~10 MeV below the known mass. Such a shift was not seen in the Fall 2014 data.
Another improvement to the analysis is the rejection of clusters matched to charged tracks - this does indeed improve the signal-to-background ratio that is seen. Other analyses in progress include a combined analysis of the Fall and Spring data, and studying electron showers.
Elton reported that Will is continuing his pi0 calibration studies, now looking at pi0's with one shower in the BCAL and one in the FCAL. He sees some shifts in energy that may be due to showers that are not entirely continued in the BCAL, and is studying them. George is looking at the effective velocities - he has some preliminary results and he is looking at their systematics.
Simon is updating his alignment calculations with some recent long cosmic runs. The data definitely seems to be of higher quality than before. In the drift range where the efficiency is flat, he is seeing resolutions of ~100-150 um, with very small residuals. Mike S.'s work on the alternate alignment with Millipede is on-going.
Mike's overall timing calibrations for the spring data are on-going. He's integrated the RF timing reference signal into his procedure, which helps a lot. The higher-rate runs are trickier and need some more analysis.
Simon has added Lubomir's alignment constants into the CCDB, and is studying their effect. Lubomir is updating these calibrations using Spring data.
Beni reported on his work on TOF calibrations. His procedure now obtains timing offsets for all paddles, including single-ended ones, and ADC versus TDC timing. He's been looking more at the ADC data - finds the timing isn't bad, though it could use some improvement, and is able to extract some preliminary attenuation corrections. There is still a lot to do, as the histograms have lots of features that are not trivial to understand.
Eric reported that the latest results have been revisiting the propogation correction calculation based on some comments that Matt S. made at the Collaboration Meeting about their fitting method. The fitting procedure has been completely rewritten, and similar results are obtained.
Nathan has been looking at the spring data that was taken with simultaneous PS+FCAL/BCAL triggers. He has skimmed the PS triggered events out from these runs, and the EVIO files can be found at:
The initial PS counter energy calibrations have been refined, and Nathan is working on improving some of the timing calibrations, e.g. there is some residual timing offsets between the arms of the PS.
Alex Somov reports that he has more time to work on the PS calibrations now, so he has started to look at them in more detail. He is also working on generating field maps for the PS magnet using TOSCa, and has a student starting to help with this work starting June 1.
Alex is looking at the results Nathan showed at the Collaboration Meeting about the overlap of the TAGH & TAGM energies. Nathan pointed out that this is probably due to the fact that the PS energies are not correctly calibrated.
Mark I. has run the production jobs for simulations corresponding to the Spring running conditions. The jobs are nearly all done. Justin has looked at them and identified some areas for improvement, especially for the diamond runs. Likely another iteration will be needed before they are ready for general consumption.
There is some initial documentation on the Spring 2015 Commissioning Simulations page, Sean is in the process of updating it to reflect the latest results.
Sean presented an initial proposal for streamlining calibration processing/development. The primary goals are two-fold: (1) regularly run production and development calibration plugins at once to make our batch farm usage more efficient; (2) automatically run calibration codes on the output of these plugins.
This proposal was discussed. Some general comments that were made:
- Calibrations for most detectors will likely focus on the spring data, since the quality of this data is much higher than what was collected in the fall.
- An exception might be the calorimeters.
- This data might be small enough that we can keep all/most of it pinned to the cache disk, and not have to worry about tape access as much.
- With the smaller data size, we could run this processing more often (every other day?).
- It would be good to have a system like this before next and/or production data taking.
- Most current calibration work right now is fairly small scale/in development, and might not benefit from this system.
- Much of the work with a system like this is expected to be in administering it and integrating the different plugins/procedures.
- It is easier to test and develop a system like this on a small scale, and work out the issues as you scale it up.
Sean will work to develop an initial version of the system for testing on the spring data.
Elton and Eugene brought up the issue of obtaining a rough calculation of the overall normalization or amount of data that was collected. Two important pieces to this puzzle are the analysis of the pair spectrometer data and of the trigger efficiency. Alex Somov said that he will work on these.
N.B. In the online meeting, Sean asked if the DAQ live times were saved anywhere. The exact numbers are not current saved anywhere - David L. had developed code to save them and other useful information to the data stream, but this code was not yet enabled for data taking. David had previously developed a method for calculating the live time by looking at the trigger time distribution, so this information can be extracted.
Sean and Dave Mack have discussed methods for using physics processes to normalize the data, but we are far from being able to use these given the current state of calibrations and amount of spring data.