STAR Correlations and Fluctuations PWG

[ShinIchi Esumi <esumi.shinichi.gn AT u.tsukuba.ac.jp>]

Dear Youquan (Wensong and Anna, Vinh)

Thank you very much for showing your very nice studies on the pion (and kaon) 

correlations. Could you (Youquan) please show us the correlation function 

in 2D distribution like p11, c2(Qinv, SL), which is the ratio of "real/mix”, or 

subtracted distribution “real - mix” after normalization? I would like to see 

them for different centrality cases as you have mentioned 0.3 cut is good enough 

for all centrality selections. This is also because Anna/Vinh uses 0.6 cut on 

the same parameter with the similar(?) data set, although it might be different 

because of other conditions like FMH(R?) etc, and TPC/iTPC difference as 

Grigory says.

There were some discussion about delta_phi or data_phi* cut in both of your

talks. delta_eta (or delta_theta) cuts are needed simultaneously with delta_phi 

cut. I suppose you know about phi* which is the local (position) angle within 

the TPC, that is quite different from phi direction at primary vertex, this is also 

the reason why we say the delta_phi* cut is supposed to be better than delta_phi 

cut (because the splitting and the merging are both happening inside the TPC 

detector, not at the primary vertex). It’s quite easy to get phi* from phi, which is 

just given by pT and charge of the track at a given radius in the TPC. So 

“delta_eta && delta_phi* cut” or “delta_theta && delta_phi* cut” could be easily 

be replaced for both of your cases and it should be better. You would see more 

splitting effect at smaller radius at about 70cm (inner TPC radius) and more 

merging effect at larger radius at about 150cm (middle - outer TPC radius). Since 

you are already based on the SL cut, you would most-likely need this  delta_phi* 

cut on larger radius, since you would like to remove the merging effect with this. 

I would also like to remind and suggest you to plot and look first in order to 

determine the cut values on the “delta_eta (or theta) && delta_phi* cut”, based 

on what you really see on these 2D correlation plots (eta, theta and phi cut do 

not have to be the same) in addition to the final “momentum” correlation function 

in Qinv etc in order to determine the final cuts. 

By the way, FMH(R) (fraction of merged hit or row?) cut has beed also used in 

previous star hbt analysis, which I do not know much about it, have you tried to 

use the FMH(R?) cut? I would like to see the some qualitative and quantitative 

comparison between “SL && FMH cuts” and “delta_eta && delta_phi* cut at two 

TPC radial positions” or "SL && delta_eta/phi* cut at one radius", where all of 

them might work as we like/expect to remove both splitting and merging, properly.  

Best regards, ShinIchi

PS : Grigory, I could not find Anna's and Vinh’s E-mail address, so please 

forward this to them, if you think this is useful for them, too. Thanks for the 

(continuous) discussion...