AuAu39GeV

This is a continuation of the embadding vs. real NFit mismatch investigation.

1) The production which was done with 15 embedded particles / event showed a mismatch in the NFit
 

2) By changing the number of embedded particles to 0.05*refMult we got a much better matching (plots are for lfspectra embedding)

3) In order to see the reason behind this dependence on the number of particles we looked at the NFit in different refMult bins. The following plots are for

π+ AuAu39GeV  (0<pT<5.0, -0.5<η<0.5,  0.05*refMult embedding)

refMult = 0-50	refMult >350	no cut on refMult

4) π+ AuAu200GeV  (0<pT<20.0, -1.0<η<1.0,  0.10*refMult embedding)

refMult = 10-150	refMult >400	no refMult cut

5) I conclude from the above table of plots that there is a problem with embedding in peripheral events. The reason that we don't see the mismatch in 0.05*refMult or 0.10*refMult is because we get much less number of particles embedded in low refMult events and by scaling our histograms by (1/Ntrk) the contribution of the tracks embedded in peripheral events is diminished. So the mismatch is there whether we use 0.05*refMult or a fixed number of particles per event.

QA @

AuAu39GeV: http://www.star.bnl.gov/protected/heavy/mstftsm/emb_refMult_bins/lfspectra/

AuAu200GeV: http://www.star.bnl.gov/protected/heavy/mstftsm/emb_refMult_bins/auau200/

--
Mustafa^2