Sept 26/23 PionLT/KaonLT Analysis Meeting Notes ----------------------------------------------- (Notes by GH & AH) Today: PionLT will be discussed first Please remember to post your slides at: https://redmine.jlab.org/projects/kltexp/wiki/Kaon_LT_Meetings Present ------- Regina - Nathan Heinrich, Garth Huber, Vijay Kumar, Nacer Hamdi, Muhammad Junaid, Alicia Postuma CUA - Richard Trotta, Tanja Horn FIU - Pete Markowitz Nathan ------ Hodoscope calibrations - time walk work took longer than expected, is finally done - time walk correction for the hodoscope calibration, using the formula (from Carlos's documentation) to fit the time difference between TDC and fADC. - correction is of the form: f_TW = 1/[ (a/t_thr)^c2 - (200/t_thr)^c2 ] where the second term is apparently used instead of c1 - some outliers w/small errors are due to bad fits, so they were removed from correction - all paddle data are divided into two periods, Winter 2021-22, Summer 2022 - some paddles have very clear divisions in the offset - previous work (last week) missed c1, now it's included - had to do some code fixes to get everything working - now moving to second part of calibration, Vprop - wil be asking Junaid for help in seeing how often Vprop needs to be redone - will push new offset params and code to GitHub tomorrow Junaid ------ - Still waiting on Nathan to push hodoscope calibration into the working software, will help him on calibs starting tomorrow - Next will be re-check calibration of the drift chamber, using new hodo calib Richard ------- Heep offsets - replay of data with new offsets to see if they worked - was running wrong branch, now re-running data with correct branch LT-sep preliminaries - issues with t-phi binning - getting zeros in certain bins, some histos had wrong ranges, now fixed - has a code to automatically equalize statistics between t-bins for low epsilon, using center+left settings - the code had some issues, for example was using a modulus and this gave some weird results, now fixed - worked on LTsep t-average calculation in parameterization - discussion about Data/MC ratios at Q2=2.1 - low epsilon ratios look reasonable before iterations, in the vicinity of 1 - high epsilon ratios have much larger range, possible bug that needs to be fixed - using 8 phi-bins, 3 t-bins - phi-bin #5 has very high Data/MC ratio (near 1000) - GH wonders if maybe one SHMS setting is missed in forming the ratio? - still some bugs in weight calibration script that need to be fixed - running the script causes unexpected changes to the weight ViJay ----- Heep offsets - started running Heep Coin SIMC Low Q2 data, to allow for in-plane offsets extraction Hodo efficiency calc - DG suggested adding an extra PID cut to be sure effs near 97% are not due to proton contamination - looked at code with Nathan, it's in hcana, not the report files - this is because it needs to make use of tracking info to select the right paddle and not all of this information is available in the report files - the code looked confusing, not sure how to implement a better PID cut - GH will see if DG, NH, VK are available for a separate meeting on the necessary code change Next steps: elastic cross section measurements Alicia ------ BSA cointime studies - cointime with PID cut, Q2=3, W=3.14 data - investigating the reason of asymmetry in the CT distribution - i.e. the accidental peaks are not flat near the prompt peak, but have a plateau about 30ns wide which are higher and the valleys between the peaks are more filled in - tight missing mass cut around neutron mass leads to more asymmetry - coincidence time vs. missing mass shows prompt and accidental bands - a large band at the neutron mass extending well into accidental events is also seen! - it is clear this band is the reason for the asymmetry in the CT plot - Even with accidentals happening at different times, the tracks are still well reconstructed and that corresponds to neutron mass - this means we have real e+pi coincidences corresponding to neutron MM (info coming from tracking algorithm), but for some reason the CT time is wrong up to 10-20 ns! - GH speculates this is due to the bad CT resolution in these 10.6 GeV data and the loose trigger timing that was used in this beginning KaonLT run - the CT problem was fixed mid-way through the KaonLT data, so Alicia will have to be careful to treat data before and after this correction differently - GH: assuming these are real good e+pi coincidences, as indicated by the MM peak corresponding to neutron mass, even though CT time is bad, we will need to apply a correction for the data lost by the tight CT cut - this should not affect the BSA data, but for sure the correction is needed for the L/T-sep data ***Richard and Ali take note!*** - for accidental subtraction, use bands far away from the prompt peak - to correct for good e+pi+n events removed by tight CT cut, need to take difference between random peaks close and far from the prompt peak - TH: Richard should do a cut study to see the effect in the K+ data - AP: it would be good if K+ and pi+ corrections are similar within errors, and if so, apply the pi+ correction to K+ data as the uncertainties would be smaller Nacer ----- - still going through Carlos Yero's thesis and Hall C papers from GH - looking to start looking at 3.8GeV KaonLT data soon Next Meeting ------------ - Tues Oct 3 @ 15:00 Eastern/13:00 Regina/12:00 Pacific - KaonLT will go first Join Zoom Meeting https://uregina-ca.zoom.us/j/94470211297?pwd=djBROEdWUENLUTF3ZFN1c0kvS0VIQT09 Meeting ID: 944 7021 1297 Passcode: 320362