I have one overall comment: good job! This is starting to look pretty useful. I have a few comments below for those who may be interested. Ian > -----Original Message----- > From: owner-brahms-dev-l@bnl.gov [mailto:owner-brahms-dev-l@bnl.gov]On > Behalf Of Claus O. E. Jorgensen > Sent: Wednesday, August 22, 2001 4:38 PM > To: brahms-dev > Subject: C1 results > > > > The C1 software gives nicer results now, and I would like > to commit it soon. You can download some plots at: > > http://www.nbi.dk/~ekman/trackDistToHit.ps.gz > http://www.nbi.dk/~ekman/blobenergy.ps.gz > http://www.nbi.dk/~ekman/eVsP.ps.gz > > I have a few questions and consideration I would like to > share with you. > > - C1 geometry: I align the geometry by looking at the > distance between track (p>4) intersections with C1 backplane > and tubes with a clear signal. The (2D) distribution show > a flat peak around 0,0 (is the geometry is right). The > width of the peak gives me the blob radius > (width = tubewidth + blobradius) - the blob radius is around 5 cm. > A tuned geometry (for 6 deg) can be found in > /direct/brahms+u/ekman/SixD.geo at the rcas machines. > > - Finding a C1 hit: I project FFS tracks to the C1 backplane > and find tubes that are hit by the blob (defined by the > trackintersection and the blobradius). I then sum the > energy of these tubes. This gives me a blob energy > with a pion peak (see plot). The separation gets better > for higher momentum (see plot), but I'm not sure if it's > good enough - what do you think? > As you point out later, this will be helped by also requiring hits in TOF1. My feeling (at least with last year's data) is that requiring TOF hits cleans things up quite a bit. Also for low momentum, fewer photons are produced. I don't know what effect to expect, but perhaps this contributes. You might also want to try doing this, but not allowing the 'crappy' tubes to contribute. > - Adc Gain calibration: I've made a rough adc gain calibration > of the C1 tubes by looking at the energy spectrum when > a track with p>4 hits in the middle of the tube (3x3 cm). > Some of the spectra look really nice, but some of them are > really crappy - I hope to improve this when I get more stat. > I don't know if we should apply a threshold on the energy > before we call it a signal in the tube, or if it's a > better idea only to cut in the blob energy. For now I cut > if the energy is < 0.2 * adcgain (this makes the eVsP plot > look very nice), I didn't use the pedestal width. The problem > is that the signal is not clearly separated from the pedestal. > Any ideas? I really think that you MUST do this for each track. That is, make the blobenergy as you describe above, and if that is above some value (which we still need to determine), then we have a BrC1Pion=kTRUE (at least if we are below the K tresh.) > > - Code: I've done all this in a BrC1PidModule, and I planned > to make a BrC1Pid class. How does this fit into the new > track/particle/??? classes (Flemming, Peter, Jans-Ivar)? > What data members should be in a BrC1Pid class (Yes/No, > Confidence Level, energy...???). > > I think there is still room for some improvements (by making > more precise gain calibration and more precise geometry). > The result will probably also get better if hits in TOF1 > are required. We can try this now, if you want. Ian > > Cheers, > > Claus > > > +-------------------------------------------------------------+ > | Claus Jørgensen | > | Cand. Scient. Phone : (+45) 33 32 49 49 | > | Cell : (+45) 27 28 49 49 | > | Niels Bohr Institute, Ta-2, Office : (+45) 35 32 53 07 | > | Blegdamsvej 17, DK-2100, E-mail : ekman@nbi.dk | > | University of Copenhagen Home : www.nbi.dk/~ekman/ | > +-------------------------------------------------------------+ > > >
This archive was generated by hypermail 2b30 : Wed Aug 22 2001 - 10:59:24 EDT