Re: [Brahms-l] qm04- first drafts

On Sun, 28 Dec 2003, Ian Bearden wrote:

> Hi Zhongbao,
> A couple of quick comments/questions...in line below:
> On 28/12-2003, at 21.07, yin.zhongbao@fi.uib.no wrote:
> 
> > Hello,
> >
> > I hope everyone will be back to work tomorrow :)
> >
> > What I found that the yields of charged particles from dAu and pp at
> > midrapidty are much higher than other experiments (PHENIX, PHOBOS and
> > STAR). For min-bias of dAu, the yields ~20% systematically higher than
> > PHENIX published, while our published data is lower than PHENIX  
> > especially
> > in low pt region. For 0-30% central dAu collision the yields I got even
> > higher than that PHOBOS publilshed spectra from 0-20% central events.  
> > The
> > reason I guess might be due to: scaledown correction and acceptance
> > correction. I would appreciate you if you could provide me some idea or
> > suggestion.
> Do you have some plots showing this?  In particular, it would be  
> interesting to see a comparison of your dAu spectrum
> compared to the published dAu spectrum.

I will show you later.

> >
> > For scaledown correction, it is clear that the inel trigger scaledown
> > factor is vertex position and centrality dependent from a  
> > two-dimensional
> > histogram. I think it is quite understandable since the inel trigger
> > efficiency has weak centrality and vertex position dependence.  
> > Certainly
> > if ignoring this weak dependence, the yield for central collisions  
> > will be
> > lowered, but I think that not the right thing to do. I would appreciate
> > if you could show your experience in how to deal with the scale down
> > factor correctly or reasonablly.
> again, it might be helpful to see the histogram to which you refer.   
> The  trigger scaledown factor is a constant for a given run, isn't it?

Sure, it is. But trigger scaledown correction might be vtx and centrality 
dependent. I will look at it run by run. Is centrality defined for each 
run instead of all data sample? 
   
> I guess you mean that the trigger efficiency is a function of VTX Z and  
> centrality (as you also say).  What is the dependence, and why is  the  
> trigger less efficient for "central" collisions?

I didn't say that it is less efficient for central collisions. On 
contrary, it is more efficient for central collisions.

> > I still doubt the acceptance map. The map generated from BRAG is very  
> > much
> > different from map generated from Claus' pure geometrical method (the
> > same cut on slats of tofw is applied). And the map from BRAG for B1000  
> > is
> > also very much different from that for B1050. The map generated from  
> > BRAG
> > doesn't match data at all. Then I would think it the current is wrong,  
> > I
> > mean much lower than 1050 actally, then the acceptance correction might
> > be larger, then the yield will be lower.
> I do not understand this.  Djamel showed a couple of weeks ago that  
> BRAG and CLAUS give similar results, though I do not have the mail at  
> hand.  Again, plots would be helpful. The fact that BRAG gives much  
> different results for B1000 and B1050  seems to me to be impossible, if  
> everything is working as it should.  Are you sure you are comparing  
> maps from these settings?  How can the current be wrong?  When you say  
> "doesn't match data at all" what do you mean?  sounds like you are  
> comparing data and MC from different settings, but I suppose you have  
> checked that?

Djamel only showed a map generated from BRAG for current of 1000 at 90 
degrees. It 
doesn't take much time to generate maps either from BRAG or from Claus' 
program. I wonder if you can verify what I said is true or not. Jens Ivar
have compared maps generated by BRAG for current of 1000 and 1050, and
also see the big difference. But may both of us have done something wrong.
So it is better if someone else can check it.

> > I would use constructed spectra from PHENIX pi^0 spectra for pp
> > collisions+Pythia  simulation as reference spectrum for R_AA of pi at  
> > both
> > mid-rapidity and forward rapidity.
> Why? And how can you use PHENIX at forward rapidity?  How does this  
> ref. compare to the one we have used for
> our published results?

Because I lack of conifidence on the spectrum get from our data especially
when I don't know the inel trigger efficiency for pp collision. It should 
also be vtx position dependent if what I saw the dependence is real in dAu 
collisions.

Using simulation you can do everything but it may not be realastic. You 
can get all kinds of ratios, for example, pi^+- at forward rapididty over 
midrapidy. pi^+- over pi^0 at both midrapidity and forward rapidity.
And ratios at the same rapidity for difference rapidity range etc. So
there is no problem to constructed a spectrum for forward rapidity.  
The reference spectrum published is for h^-. I haven't looked at it.

> >
> > By the way I have re-studied the RICH efficiency via simulation for  
> > 12A427
> > with a fiducial cut applied on track intersetion with RICH mid-plane.  
> > The RICH
> > efficiecy for pion at pt>3GeV/c is ~97% constantly.
> Thanks for some good news...:-)

Bad news sometime is better than good news. It can help us understand 
something better or make new discovery... :)

Best regards,
Zhongbao
-----------------------------------------------------
Zhongbao YIN                 Phone:  +47-55-582792 (O)
Address:                             +47-55-276803 (H)
  Fantoftveien 14G 466       E-mail:
  5075  Bergen                  Yin.Zhongbao@fi.uib.no
------------------------------------------------------


_______________________________________________
Brahms-l mailing list
Brahms-l@lists.bnl.gov
http://lists.bnl.gov/mailman/listinfo/brahms-l