LCD Weekly: Simu and Reco 6/7/05 --- [...] DW: Klongs into block of steel. 100 MeV kE -> 2.5 GeV depositted. LEP (Low Energy Parameterised) model. Not very good but surprised that is was "that bad". Could not reproduce in N03. G4 release on 6/30, which was to have had Bertini kaon extension. Discovered a bug, so it will have to appear as a patch. SLAC distribution of 7.1 will have the model in. Include a note on LCPhys page that will need the patch. Anybody who wants to try now, it is ready in a test release. Just finished writing and testing today. Switch to a more hermetic example. GN: Sensible for neutrons? DW: Old model was sensible for neutrons. Protons, neutrons and pions -> already used the Bertini Cascade, which is better than LEP. Bertini wasn't able to handle strange particles, but now can handle lambdas, cascades and kaons. RP: Machine backgrounds, pair backgrounds? NG: Machine backgrounds are being generated using GUINEA PIG. Glen White is generating samples of these @ UK. Have not put through any detectors yet. WHIZARD now reads in the GUINEA PIG files, and it handles the physics-type backgrounds. RP: Device that is not sensitive to time structure, need to look at whole train. If sensitive to time, look at particular crossing. NG: Do what did with "warm machine". Pile up 3000 events and do the analysis. RP: Each event will have all the pairs from that particular crossing. NG: In principle, could do it that way, but in practice, might "run out of patience". Since event has internal timing, just +30 nsec. NS: Usually, don't need the timing. Just decide how many backgrounds needed by resolution of tracking detector. RC: Need it for other detectors. Reading out a calorimeter and only get 5, 4, 8 -> need to write out so have it for individual cells. Beyond some number -> no more signal. To do this, need the signal. NG: Update on EM? DW: No. TK in Genoa. NG: Resolution is too large w/ Geant4 out-of-the-box. Range cuts lower, then approach 15%. Suspicion that range cut is not appropriate for this particular problem. NG: Uriel on Geant4 simulations? JG: Cutoff if two photons came from same pi0. NG: Working with hep.lcd software? JG: Yes. NG: By after Snowmass, hope to have all hep.lcd functionality. Started using any other detector descriptions? JP: Sent file to NG couple weeks ago. # of layers was one short in (?). NG: Send me update. JP: Recommmend to continue using JAS2 w/ hep.lcd? TJ: Not really. Can use hep.lcd with JAS3, too. Finish off something, then maybe stay with hep.lcd . New stuff, e.g. clustering algorithm, start to use org.lcsim. UP: Correction put in for clusterer? Can download code and it is now corrected? NG: Idea is to make weekly releases of org.lcsim after this meeting. Tag a release and make that available. UP: Schedule for realeases = every Weds? How will we know? (hep.lcd) TJ: Planning on weekly releases w/ org.lcsim. Announce on mailing list for this meeting. Could do the same thing for hep.lcd patches. NG: Tracking? NS: Fine tuning parameters. Very high efficiency = 99.5% . Still some tunings to be made. Sent list of events where hits are missing. Originally thought a reason that a pair was generated and both tracks had = origin point. Maybe assigning this point to only one track. Tracking shouldn't care, because should still assign this point. See two hits from the same MCParticle at distance of 3 cm . NG: Very low energy secondary may not create MCParticle. Can get a hit from a track that is not near physical track. TN: Detailed material into compact geometry. Ran into a few small inconsistencies/gotchas and trying to sort out. Trying to come up with class hierarchy for tracker digitization -> readout segments, measurement coordinates, charge dep on real strips. Waiting for a snippet of code about geometry into reconstruction. RP: Access entry and exit points in tracker hits? RP: Something going parallel needs this. NG: Failing of LCDG4 output that direction was not saved. No good way of saving direction. TJ: Maybe add to LCIO but can't currently count on it being there. SM: Need to do endcaps now. Pretty soon, should have MIP clusters that go into that volume (?). Need final output. Put into cluster interface to add hits to cluster. NG: Update from SK? SK: Error in JAR received. Tried Java 1.5, but didn't work. NG: JM on Far Forward Detector? JM: [slides] RP: Details of beampipe design are interesting for tracker. Where it flares out, cutoffs, etc. If 2 mrad, might do redesign. Depends on what is happening with the pairs. NG: Nice thing about 20 mrad -> smaller incoming beampipe, so that is the limitting aperature. RC: Problem with projective endcaps getting correct position back. TJ: JS doing work on vertex finder. A few small changes in org.lcsim to fix issues reported by JS. Added MCParticle browser. TJ: Should try to have weekly release notes.