7. Examine calibration results

While 'bass-o-matic' grinds through the data, it accumulates a few histograms. After it is done, we should have a look at some of them to make sure everything worked right. Actually 'Hist' can display the histograms in real time as the data is being processed.

The histograms of special interest are listed below, in the order they appear in the list: (NN in the histogram names is the iteration number)

• "1 pi0 mass"

  fig-pi0mass.ps "1 pi0 mass"

• "2 pi0 width"

  fig-pi0width.ps "2 pi0 width"

• "3 cc width"

  fig-ccwidth.ps "3 cc width"

• "5300 Mass wcuts 2 Gamma"/
• "13100 tag hadrons: X-distance"
• "13200 tag hadrons: Y-distance"
• "13300 tag hadrons: XY-distance"
• "3xxxx missing mass"
• "40NN chisq"
• "11NN 2-gamma eff. mass, 2 gamma events"
• "12NN 2-gamma eff. mass, 3 gamma events"
• "14NN pi0-pi0 eff. mass, 4 gamma events"
• " 6NN newold CC"/
• "20NN 2 gamma eff mass, 1-3-3"
• "21NN 2 gamma eff mass, 1-3-3 PS"
• "22NN 2 gamma eff mass, 1-1-1"
• "23NN 2 gamma eff mass, neutral"
• "24NN 2 gamma eff mass, 0-2-2"
• "25NN 2 gamma eff mass, 0-4-4"

The best way to see if everything is okey is to compare them with the histograms from some other, already calibrated group.

There are a few things to watch for, though:

• the pi0 mass: the pi0 mass should not jump around and stay within 1 MeV from the nominal pi0 mass.
• hadron rejection: both X and Y "distance" histograms should peak at zero and be symmetrical. An assymetry in the histograms means that the LGD is "in the wrong place", that is, the LGD position encoders are not calibrated correctly. This condition should be immediately fixed by changing the offset (in cm) of the LGD position encoders. This procedure is not documented here. Sorry.
• missing mass plots: should peak at the mass of the proton/neutron squared (approximately 0.88 GeVˆ2).
• 2 gamma effective mass plots: you should see the "pi0" peak at the pi0 mass and the "eta" peak at the eta mass. The eta peak is expected to be sharper in the all neutral data. There ought to be no "extra" peaks.
• pi0-pi0 eff mass: in the all-neutral data, the "K-short" and "f2" peaks should be clearly visible. If there is no all-neutral data in the calibration data set, the "f2" peak will be gone but the "K-short" peak should still be visible.

Once all the histograms look right, the group can be considered "calibrated".