comparison of run8 results

[Claude Petitjean <claude.petitjean@psi.ch>]

Dear friends,

since we have now the preliminary analysis reports of Tom and Steve at
hand, it is natural to compare their results and proposed corrections
in order to locate possible discrepancies that need to be resolved.

For comparison of the lifetime fits I use data analyzed at conditions
which are extensively treated in both reports:

- muon stops muSC/muPC1/TPC in good fiducial tpc volume,
   +-25 us pile up with muSC and muPC1XY-OR

- clean electrons fully tracked (eSCxePC1xePC2), ePC's with Cathode-OR

- fit time bin 100 ns - 24'000 ns

- impact parameter window 120 mm

Here follows the comparison:

                                          Tom                 Steve
1) fit of all prod-50 cleanfill
    data before any corrections
    total statistics                      1.56 E+09           1.632 E+09
    lifetime (lambda) [s-1]         455'436.3 +- 12.3   455'434.0 +- 12.1

2) remove identified mu-scatters        -3.1 +-  0.2        -1.1 (n.a.)
    corr. unident. mu-scatters           -1.4 +-  0.5        -2.0 (n.a.)
                                    455'431.8 +- 12.3

3) correct high-Z impurities           -18.7 +-  1.6       -13.9 +- 

                                    455'413.1 +- 12.4   455'420.1 +-

4) correct mu-d diffusion               -8.6 +-  1.1       -12.1
    (using nat. H2 calib.)          455'404.5 +- 12.5   455'408.0 +- 12.2

5) correct mu-p diffusion               -2.8 +-  0.5        -2.7
                                    455'401.7 +- 12.5   455'405.3 +- 12.2

6) remove bad muSC run periods          -1.9
    (not applied yet)
7) muPC1 inefficiency correction                            +1.1
    (n.a.)
7) cosmics errors adjusted              -3.1            (-0.2 incl in 1)
    (n.a.)
8) use only unique electrons            -0.4
    (n.a.)
9) pp-mu formation                     +11.4 +- 3.6
    (n.a.)
10) ortho-para transitions              +4.6 +- 3.2
    (n.a.)
11) electronics & beam structure    (+-?)6.8 +- 4.6
    (not applied yet)

	*************************

Here follows some comments:

the agreements are quite satisfactory, since most differences can be
understood:
- Steve sees less mu-scatters, but this may be due to a different
   track finding algorithm
- Steve has a smaller high-Z correction because Nitrogen is assumed
   as the main impurity. This needs to be modified eventually.
- there is a difference in mu-d diffusion corrections. This will
   need further discussions.
- why is the adjustment of cosmics error so different? Needs to be
   discussed
- the corrections due to pp-mu formation+ OP transitions need to
   be presented a.s.a.p., since it's a sizable effect.