**NEW PLOTS**:

1. Rechits with quality flags 0,1 set are no longer rejected; bit 0 because it rejects PMT hits and we want to study them, and bit 1 because it was optimized for center of sample running, and interfered with analysis of 50/50 data.
2. The 2TS Algo in use for these plots now agrees with official reco in terms of what phase represents t=0.
3. The minimum hit threshold was changed from 5GeV to 20fC and plots were redone in fC.
4. Hits with peak sample in sample numbers 2-5 were included in the analysis.
5. The Rechit reflagger was used to identify PMT hits, and the reflagger was configured to use the PET algo (default settings) for both Long and Short fiber channels.
6. Matched hits are both required to exceed the 20fC threshold; if one depth exceeds 20fC and the other does not, the larger hit is thrown into the "unmatched" collection. Unless otherwise specified, all plots are of matched hits only.
7. No time slew corrections are currently applied (same as before).

• Delta T between channel depths as a function of ieta and L/S energy sharing:
  

• Plots showing how the 2TS algo converts a distribution of energy-weighted 2-sample sums into it's corresponding hit time distribution:
  

• Timing per category of rechit. Note that unmatched hits have a similar profile to matched hits - so are these PMT hits or not?:
  

• 2TS Charge Sum distribution for long/short and different rechit categories - PMT hits identified by the reflagger set to the PET algo for both L and S:
  

Selection/Reco criteria:

1. Running on "GOODCOLL skim" files /store/dataCommissioning10/MinimumBias/RAW-RECO/...; runs 132961-132968, lumi sections as indicated on this twiki page. (NB: Some of the plot titles below indicate both runs, others indicate only one. All of the plots though have both run's data in them except the run info plots.)
2. Require L1 trigger bit mask: "0 AND (40 OR 41) AND NOT (36 OR 37 OR 38 OR 39)")
3. "OneGoodVertexFilter" and beam scraping filter also applied (note that the GOODCOLL skim only applies this and the previous requirements in an "OR" configuration).
4. Require BX=1
5. Reco window considers time samples 2-5
6. Time Reconstruction done for HF with Igor's new 2TS algorithm (also documented on the HFScan2010 twiki), except for time averages performed on averaged pulse shapes. There is a known offset of ~20ns between the new algorithm and the official reco algorithm due to a different choice of time t=0 in the lookup table internal to the reconstruction software.
7. Reject rechits with quality flags 0,1,18, or 19 set (see Jeff's twiki)
8. Only hits above 5 GeV considered
9. Only hits with peak sample in TS3 or TS4 considered
10. Igor's PET algorithm for determining PMT hits from two paired hits (L,S) from a given tower causes a different histogram to be filled. (The algorithm is implemented as defined in a talk given by Dinko a few weeks ago.)
11. Non-PMT hits with matching hit in the other depth (and both above threshold) are considered "verified", otherwise they are "unconfirmed"
12. The metric TS4/TS3+TS4 is formed from the digi in fC (no ped subtraction, no gains applied) for verified, unconfirmed, and PMT hits.

Run 132961-8 basic information:

Basic HF Rechit info:

Timing, energy, occupancy, polar plots:

Timing maps, depth comparison as a function of eta/energy

• Average event-by-event timing differences (Tshort - Tlong) per eta ring as a function of eta and energy:
  

Jitter phase per channel per event using the ratio TS4/(TS3+TS4):

Digi detail for HF+ Quadrant 1, depth 1, ieta=30-34:

Digi detail for HF+ Quadrant 1, depth 2, ieta=30-34:

Digi detail for HF+ Quadrant 2, depth 1, ieta=30-34:

Digi detail for HF+ Quadrant 2, depth 2, ieta=30-34:

Digi detail for HF+ Quadrant 3, depth 1, ieta=30-34:

Digi detail for HF+ Quadrant 3, depth 2, ieta=30-34:

Digi detail for HF+ Quadrant 4, depth 1, ieta=30-34:

Digi detail for HF+ Quadrant 4, depth 2, ieta=30-34:

-- PhilDudero - 14-Apr-2010