From SCHLABACH@FNALDFri Nov 10 17:09:15 1995 Date: Thu, 9 Nov 1995 15:37:50 -0600 (CST) From: SCHLABACH@FNALD To: PEKKA@FNALD Cc: "WHCDFM::SCHLABACH"@FNALD, schlabac@cdfsga.fnal.gov Subject: muon code changes in v7.12 All changes were in C$MUO and the C$INC/C$DIC files associated with C$MUO unless otherwise noted. Packages ******** Changes to the d-to-e reconstruction ==================================== Bank changes: CMUE/CMPE/CMXE/CSXE unchanged Filtering on error flags - The d-to-e modules now allow filtering on the error flags allowing one to pick off particular types of pathological events. This is of interest mostly to the detector experts, but it might be useful for validation. Inching toward standardization of the 4 packages - This is only a help to the librarian. Some speed up in db access - CMX/CMP got an order of magnitude speed up in 7.11. Some changes were added to CMU and CSX. Big improvements could still be made. Changes to the stub finding =========================== Bank changes: CMUS changes, CMUL, CMUR, CMRD unchanged Changes to the CMX stub finding algorithm (CDF note 2871) - Changes were made to the algorithm to increase the efficiency and improve the matching Version 2.0 of the stub bank was defined - More trigger information was added to the stub banks in a spare word for all 3 detectors; the definition of the "trigger tower" was changed for CMX to remove the low phi bias at the tower boundardaries (see CDF note 3027) - it now chooses the tower with the most hits on the stub rather than the first tower with two hits. This was not changed for CMU and CMP (it's not a problem, at least not at the level one can see) Changes to the track-stub linking ================================= Bank changes: CMUO, CMIO both change, energy summary/link banks change The previously undefined fidicual word is now filled using CMUSWM - the so-called fidicual word now contains matching information from the CMUSWM extrapolation and the hit mask (mimicing the CMUO type word) was added in a spare word. As noted below the cosmic ray word in CMIO is now filled in the same was as the CMUO word. The energy summary/link bank information added in 7.11 was improved. It now contains all the information to reproduce the energies used in muon clustering. Changes to cosmic ray filters ============================= Added MUONCOSFLT defined by C$DIC:CMCRFL_DICT.UIC - CDF note xxxx adds timing of various sorts and back-to-back stub-stub and track-stub cuts to try to get away from the track-based bias of the standard (used in production) filter Changed the standard cosmic ray filter to fill the cosmic flag word in the CMIO bank (in addition to and in the same way as the CMUOs) Simulation ========== CDFSIM (C$GMU) - The pulse width is now simulated for CMX and CMP. The CMX drift model now is the same as in the reconstruction. QFL (properly part of C$QFL but completely entangled with C$MUO) - Let me remark here that MUOQFL and CMUSWM out of phase and that CMUSWM is right. QFL will overestimate the CMP acceptance at the few % level. This should be fixed, but I haven't had time and haven't persuaded anyone else to do it. Well, I persuaded Jonathon, but he didn't do it yet. At least we know how ... Utilities ********* CMXBTC/CMPBTC ============= The unpacking of the raw data banks now adds the pulse width as well. This, in combination with the changes to the simulation, now makes the trigger simulation better - both for real and simulated data. CMUSWM ====== The fidicual volume is now correct for IA CMX - one trigger tower that didn't work has been removed. TCMDPT/TCMQPT ============= The unpacking was changed to reflect the changes in the CFT FIFO (64-128 deep). Those who don't know what this means don't care. TCMQPT now recreates some information that was not kept in TCMQ but can be inferred from information that was. (Jon said the latter was buggy; he buggered it; he fixes it.)