Trying to do further QA for run 8 dAu I plotted the refmult distribution for BBC coincidence rate below 140KHz (Low lum) and above 170KHz ( High Lum) trying to see the effect of pile-up on the refmult distribution:

Query used:

<input URL="catalog:star.bnl.gov?filetype=daq_reco_Mudst,trgsetupname=production_dAu2008,production=P08ic,storage=nfs,filename~st_physics " nFiles="all"/>

Trigger Id :

is210020  = muEvent->triggerIdCollection().nominal().isTrigger(210020);     //VPD-ZDCE trigger for dAu

# of events:

653297 LowLum, (19329 with single vtx)

190968  HighLum, (174 with single vtx)

What is plotted above is the reference multiplicity distribution for events with BBC rate below 140KHz and above 170KHz  for all events (with pile-up) and for events with only one reconstructed vertex (no or little pile-up).

In the left plot HighLum_single_vtx has very few stats ( only 172) . One notes from the right plot that Hight multiplicity events (with more pile-up) shifts to higher refmult (consistent with Minuit picking up higher multiplicity pile-up vertices rather than real low multiplicity ones) .