Collaboration comments to v3.0 of Pion HBT paper and responses
Here are the comments from the collaboration to version 3.0 of the pion HBT paper from p+p collisions, and the PA responses:
From the BNL group - (they sent in a pdf file of comments; we paraphrase their feedback, here)
1) suggest changing title to "PION femtoscopy in p+p collisions at RHIC"
-- we agree. Done. In fact, as per IOPP suggestion (below), we also replace RHIC with 200 GeV.
2) Why don't we normalize the correlation function to unity, but leave it at 0.82?
-- the main reason is that there is no "flat region" at large q, over which to normalize the CF. This is one of the main themes of this paper. We could normalize the plot so that the CF comes "near" unity, but this is really arbitrary. Very important: this normalization issue does NOT affect HBT radii or other fit parameters, since the overall normalization is an overall parameter. This is true for all HBT analyses.
3) <Pz>=0.32 seems too high for pions.
-- this comment is concerned with the EMCIC parameters listed around line number 532. Please note that it is the average value of Pz-squared that is 0.32. Probably even more important to this comment, please note also that this number is not for the measured pions, but for all particles emitted from the effective source. Take a look at the discussion and formalism in [44]. (See e.g. Tables 1 and 2 from that paper.)
4) Why does Figure 6 show fit parameters from zeta-beta fit, but Figure 7 does not?
-- The zeta-beta fit is only applicable for three-dimensional correlation functions. Figure 6 shows parameters to a fit to 3D correlation functions, while Figure 7 is only for one-dimensional correlation functions.
5) Figure 12 is missing a label on the y-axis of the lower panel.
-- thanks, this is fixed.
6) Clarify which formula-- equation 7 or 8-- is used to get the parameters plotted in figure 14.
-- thanks for catching it. it is Eq 8, and it says so now in the paper.
7) Explain the x-axis on Figures 13 and 14. Why <pT> and/or sqrt(2/3)|p|?
Yes, these x-axes were a big mess. We've cleaned it up; please take a look now. The issue was simply that NA22 and NA27 reported the total momentum (not transverse) of the pair. In order to compare it to ours, we assumed isotropic emission, and as a proxy say pT=sqrt(2/3)|p|. Not ideal, but it seems the best we could do, or?
8) make usage (italic versus roman) consistent, when writing "Au+Au", "p+p" "e^++e^-" etc.
-- done
9) When starting a sentence, use "Figure," but otherwise use "Fig."
-- okay done.
Comments from Spencer Klein 18 Jan:
This is a very nice paper, presenting a very carefully done physics analysis. Nice. Although I have no comments on the physics, I do have a few comments on the presentation. I acknowledge that many of these comments are a bit 'squishy;' feel free to ignore them if you don't like them.
Line 131 - 'soft sector' is pretty slangy.
Okay, now we have removed that term until we mention Pythia in the Discussion section. There, "soft sector dynamics" is discussed explicitly in the reference given.
On line 177, it's not clear what you mean by 'reconstruction system.' If this refers to tracking software, etc. say so.
Yes, this is essentially it. We say "reconstruction software," now.
There are a few spelling mistakes: 'measurements is misspelled (differently) on lines 139 and 204, 'briefly' is misspelled on line 395, and 'resonance' on line 750.
Thanks. Fixed.
Line 184+ seem a little overblown.
Hmmm, okay we have slightly toned down, replacing "We hope that our results may eventually lead to a deeper understanding..." with "These results may play help in understanding..."
Line 336 says 'we will use the name '\delta-q' when the above formula...' Is this Eq. (12) or Eq. (13)? This should be clarified.
Well, we use Equation 12 for 1D correlation functions and 13 for 3D correlation functions. This is clarified now in the sentence following the formulae.
Since this is not a letter, could you add a little bit about the trigger conditions to Section III? You later show that multiplicity affects the measured radii, and multiplicity can depend on the trigger conditions.
Okay, we have added some information about the trigger, based on BBC coincidence, at line 450.
Likewise, could you say something about how diffractive events are handled? Our trigger is sensitive to doubly-diffractive events (i.e. where both nuclei break up diffractively); previous STAR pp papers have accounted for the diffractive events, at least crudely. If I remember correctly, they were between 10 and 20% of the cross-section. Could you say a little bit about how these events affect the radii? It seems to me that, at least, this is a systematic uncertainty.
Actually, we do not correct for doubly-diffractive events in our earlier papers, but only say that they could give a contribution on order 11% for the yield integrated over all rapidity. At midrapidity, the contribution to the yield is significantly less, and the contribution of pairs at midrapidity will be much smaller yet. But the point is that we are measuring the radii for NSD, just as in previous papers (e.g. R_AA) we measured spectra for NSD. To the extent that DD contributes, it is included.
Line 437 'it was only a 1% of the femtoscopic radii' isn't very clear. Can you rephrase, with a little more clarity about what 'it' is, and where the 1% comes from?
Now we say in particular what was done, which was that we played around with PID cuts (cutting on dEdx, or nsigma), being very "strict" or opening up the cuts all the way, and saw very little change in radii (1%).
I would suggest looking at the section that compares our pp results with the worlds pp data, to see if it can be shortened, and if the promises that are made early on can be toned down - the conclusions to that comparison are (necessarily) not very crisp, and people may be left a little bit disappointed. The analysis is beautiful - very detailed - but it unfortunately doesn't lead to a clear conclusion.
Figs. 11, 12 and 15 cover a rather wide range of experiments. I'm wondering if a sentence or two about them would be worthwhile (I recognize that this contradicts my previou comment).
Line 679 'experiments did not follow a standard method of measuring and reporting multiplicity...' I don't think that there was a standard method than; this sounds a little negative toward the experiments. Can you rephrase?
We did tone down somewhat the promise at the beginning ("hope it leads to deeper understanding" etc); hopefully people don't feel too disappointed after that, but who knows. I guess I see what you mean about negative tone regarding past experiments, but, between you and me, I am really shocked when I look at these old measurements. I mean, did they not talk to each other at ALL? Did they not want to compare to each other at ALL? But okay, who knows what people will say about our papers in 20 years; indeed, who knows what they are saying now?
We have moderated the negative tone; I hope it's better.
LEP ran at several different energies, most notably at the Z^0 peak (~ 91 GeV) and above W^+W^- threshold (> 180 GeV), and, I believe, published HBT radii at both. Can you add the energies for the points to Fig. 15?
All of the radii shown are indeed at 91 GeV. We've updated the figure to include that information.
A few references - 33, 56 and maybe others, that appear to merit a collaboration label.
Hmmm, well we are trying to use "standard" bibtex directly from SPIRES, and for those papers, the collaboration (E735) isn't listed for whatever reason. But okay we have added them by hand. (I am thinking though that they will disappear again after it goes through the journal.)
Comments From Institute of Particle Physics (Huazhong Normal University) 18 Feb
1) Title: Femtoscopy in p+ p collisions at RHIC ->
Femtoscopy in p+ p collisions at \sqrt_{s} = 200 GeV ?
Okay good idea. In fact, as per the BNL group's suggestion, we preceed it with "Pion Femtoscopy..."
2) Abstract: line 1 and line 4, p+p -> $p+p$
line1/2 at 200 GeV -> at \sqrt_{s} = 200 GeV
Okay, done.
3) Line 111, up to 200 GeV -> up to \sqrt_{s} = 200 GeV
Line 112, d+Au -> $d+\mathrm{Au}$
Line 131, in the soft sector-i.e. at low pt. ->
in the soft sector, i.e., at low pt.
Okay, done.
4) Line 193, The paper is organized as follows ->
The rest of the paper is organized as follows
Well, we see your point, but prefer the wording as it is, so we didn't change it. Let us know if you feel very strongly.
5) Line 221, $Q_{inv}$ -> $Q_{\mathrm{inv}}$, everywhere,
$R_{inv}$ -> $R_{\mathrm{inv}}$, everywhere
okay done.
6) Line 228 and 236, “o,” “l” and “s” ->
“$o$” “$l$” and “$s$”
done
7) Line 430, 439, ... MeV/c -> $\mathrm{MeV}/c$, everywhere
done.
8) III Analysis details, are there any nhit, dca cuts in your analysis?
We have left most of the cut details to reference [49] (our long HBT paper), since a main point is that we do identical analysis. However, we now mention explicitly that we require at least 10 hits and a DCA less than 3 cm.
9) Line 481, GeV/c -> $\mathrm{GeV}/c$, everywhere
done
10）.In line 689, is it "pT-dependence" or "mT-dependence"?
Well, since we only discuss one particle species, they are equivalent.
11) Line 749, 808, $Au + Au$ -> $\mathrm{Au} + \mathrm{Au}$
Line 782, 803, 805, p+p -> $p+p$
Line 803, A+A -> $A+A$ or $AA$
done
12) Line 812, 813, why we need "-" here?
Now that you point it out, that sentence wasn't very good. We've reworded it, without any "-".
13) Fig 1, could we make the label "q_{o, s, l} [GeV/c]" larger? The
label size for different Fig are different, Should we make them same ?
We have made the label larger on Figure 1.
For the other figures, we think they are readable, so can remain for now.
14) Fig 3 and 5, $A_{22}^{RE}$ overlaped with data points.
Okay fixed.
15) Fig 7, caption, $R_{inv}$ -> $R_{\mathrm{inv}}$. Same thing in the
table V, VI, VII.
Okay fixed.
16) Fig 8, try this if PAs used root,
<dNch/deta> -> #LT dN_{ch}#/deta #GT, also, $R_{inv}$ -> $R_{\mathrm{inv}}$.
Nice, thanks on the #LT, etc!
17) Fig 9, $Au+Au$, $Cu+Cu$ -> $\mathrm{Au} + \mathrm{Au}$, \mathrm{Cu}
+ \mathrm{cu}.
Done.
18) Fig 11, 12, <dNch/deta> -> #LT dN_{ch}#/deta #GT. same as Fig 8.
Done.
19) Fig 14, the label for the x axis should be a little bit far away
from the axis.
Done.
20) reference [1] - [4], the journal information should be added.
(Please check all references.)
Okay, done. There were a few more that needed journal info added.
Actually, it seems we don't need the arxiv
number if the paper has been published.
well, we are doing straight bibtex from spires. The arxiv info can be included, or not, according to the style file. PRD has their own style file, so will format it all at once.
reference [14], why we need website here?
beats me! I've removed it manually. (spires bibtex had it included, for some reason!)
Comments from JINR (Jan Fedorisin, Peter Filip, Richard Lednicky and Stanislav Vokal for JINR)
Abstract: What about to put "200 GeV" in one line?
Done.
109: "The extraordinary flexibility of the machine"
Since e+A and p+A collisions are not yet possible,
it can better be reformulated as:
"High flexibility of the machine"
Hmm, usually I'm not much for cheerleading, but in this case, I strongly want to keep the wording. With such a range in energy, polarization possibility (transverse and longitudinal), and asymmetric systems, RHIC deserves to be called "extraordinarily flexible."
139: measuresments -> measurements
fixed. Thanks.
There are many comments on the term "HBT," which I collect here:
144,398: "HBT radii" -> femtoscopic radii
There is a huge amount of "HBT" garbage which, in most cases,
can be removed or substituted:
199,246,396,531,550,611,
689,701,710,760,774,787 : HBT -> femtoscopic
247,551,567,574,602,613,
614,629,636,667,675,716,
719,732,740,748,766 and
caption of fig.9 : remove "HBT"
675-676: lower HBT radius, at a given multiplicity ->
lower radius obtained by E735 at a given multiplicity
162: HBT correlations -> Femtoscopic Bose-Einstein correlations
I am certainly a subscriber to the "femtoscopy" nomenclature to describe two-particle correlations arising from the spatio-temporal proximity of particles at freezeout, especially when one is discussing such correlations in general, i.e. for various particle types. However, I see no reason to entirely expunge the term when discussing pions in particular, when this term has seen historical use for pion correlations and when our paper connects to historical data. After all, it's not like "HBT radius" is "wrong." It's just a term. We even explicitly acknowledge this early on, by using the term in quotation marks.
Furthermore, there is another reason why wholesale replacement of "HBT" with "femtoscopic" is a bad idea in this paper. In this paper, we frequently make the distinction between "femtoscopic" and "non-femtoscopic" correlations, a distinction which is not often discussed and about which significant confusion can arise. We mention the radii frequently in the paper, and if we start calling all "HBT radii" as "femtoscopic radii" when we are at the same time referring to "femtoscopic" and "non-femtoscopic" correlations, it will be simply too much.
For both of these reasons, we argue to continue the liberal use of the term "HBT" in this paper.
144: the mass -> the transverse mass
done
146: [13] -> [13,10]
done.
165-166: "less clear and well developed" perhaps should read
"less clear and not well developed"
Good suggestion, thanks.
181-183: into the context both of ... and
(as much as possible) into the context of previous ->
into the context both of ... and
(as much as possible) previous
Okay, except we preserved the "of"
224-225: Pratt-Berttsch "out-side-long"
coordinate system [23,24] ->
"out-side-long" coordinate
system [gra'77,pod'83, 23,24]
add earlier references:
gra'77: P. Grassberger, Nucl. Phys. B 120, 231 (1977).
pod'83: M.I. Podgoretsky, Sov. J. Nucl. Phys. 37, 272 (1983).
(particularly, see eqs. 12,13 in [pod'83])
We have added the Podgoretsky reference, but it not actually clear to us that the Grassberger reference is relevant here. Probably I just missed it, but does he really propose an out-side-long (or equivalent) coordinate system and a fit with three radii? I do see some discussion of q transverse to P and q parallel to P, and cigar shapes, but that's kind of a tenuous connection. There is no (or?) suggestion to analyze in that frame.
283: separation distribution. ->
separation distribution [led'79].
add pioneering reference on the topic:
led'79: R. Lednicky and MI. Podgoretsky,
Sov. J. Nucl. Phys. 30, 432 (1979)
done.
301: vanishes. ->
vanishes [cso'91].
add pioneering reference on LCMS:
cso'91: T. Csorgo and S. Pratt,
Proc. Workshop on Heavy Ion Physics,
KFKI-1991-28/A, p.75.
You are right, we need a reference there. But actually, we found an even earlier one, which is even published and available:
S. Pratt, T. Csorgo, and J. Zimanyi, Phys. Rev. C42, 2646 (1990)
Probably the conference talk was based on that publication.
314: points. ->
points and over the angles of the relative
three-momentum \vec{q} in the pair rest frame.
done.
380: C_F(q). ->
C_F(q), especially taking into account their
rather small values of the radii in p-p collisions.
done. We actually word it "...especially considering the small radius values in p+p collisions."
395: breifly -> briefly
done
431: their the -> their
oops, thanks. done.
481: ...[0.00,0.12] GeV/c.
Why such a range? No comment in text.
Well, there's no reason. It's just to make a picture. In heavy ion papers, one usually projects over a region of 35 MeV or 50 MeV (often without even mentioning it). It's just that's where the bump is.
492: to present -> of present
No, we really want it to say "to present." But we see the confusion-- it wasn't well worded. We have changed "...the determination to present results..." to "...the desire to present results..."
509: ... Black dotted and purple dashed curves ...
and
511: ... Red curves
Colors are OK in the online text,
but how it will be seen in the paper copy?
Well, mentioning colors doesn't harm anything, but you are right--we agree that there should not be ambiguity. Hence, we have changed the one about red curves to "Solid red curves..." That way, the reader who really only uses a B&W printer can cue on the words "dotted" and "dashed" and "solid."
574: R_{inv},extracted -> R_{inv}, extracted
done.
566-591: Perhaps it should be explicitly mentioned why
dNch/deta=7 is chosen to select the events in
low and high multiplicity classes
Okay. Now we say "The choice of the cut was dictated by the requirement of sufficient pair statistics in the two event classes."
652: STAR's -> of STAR's
done. thanks.
651-657: this paragraph requires an improvement,
description of figures 11 and 12 is not very clear.
Okay, we've reworded that paragraph somewhat; hopefully it is better now.
680: Nch -> N_{ch}
fixed.
694-700: Please explain why <mt> and Sqrt(2/3)|p| variables
can be used on the same horizontal axis to compare
Tevatron and STAR data.
Yes, these x-axes were a big mess. We've cleaned it up; please take a look now. The issue was simply that NA22 and NA27 reported the total momentum (not transverse) of the pair. In order to compare it to ours, we assumed isotropic emission, and as a proxy say pT=sqrt(2/3)|p|. Not ideal, but it seems the best we could do, or?
706: lonngitudinal -> longitudinal
oops. fixed.
727: "Additionally Alexander's explanation ..."
sounds too personal. Please replace by
"Additionally, the arguments from [64,65] ..."
Yes, much better, thanks.
736: HBT -> Bose-Einstein enhancement
done.
750: resonannce -> resonance
fixed.
759-764: The sentence starting with "With a very similar ...."
is significantly emotionally NEGATIVE.
The same content can be expressed by a more
appropriate and more neutral statement:
"With a very similar model, Humanic [71] was able to
reproduce femtoscopic radii measured at the Tevatron [34]
only with strong additional hadronic rescattering effects."
Wow, we didn't mean to be emotionally negative, but we see your point. Your suggestion is much better and is in now.
832: remove "Russian Ministry of Sci. and Tech." (no more exists)
Okay, except you need to tell this also to Liz (I guess), since it comes directly from the official STAR acknowlegements document which one gets from the author tools page:
http://www.star.bnl.gov/central/collaboration/authors/
Comments to figures:
Fig.1: Offset of horizontal axis title should be enlarged.
It touches label 0.3.
Ok, it is better now.
Perhaps add links to Eqs. used for the
fits directly on the figure:
measured CF
standard fit (11) \Omega=1
\delta-q fit (13)
\xi-\beta fit (14)
EMCIC fit (15)
In a way it's a nice idea, but we'd rather not put in equation numbers (which may change) into the plot.
Figs.1-5: Fonts in X/Y- descriptions should be larger
and unique;
offset of horizontal axis titles should be enlarged
Ok, it should be a little better now.
Fig.7: Offset of vertical axis title should be enlarged
(it touches label 1.5).
Okay, it is not touching now.
Fig.9: Zeros on vertical axes of the upper plots
overlap with 8's of the lower plots.
fixed.
Fig.10: Y-descriptions should be larger (as on Fig.9.)
They are a little bigger now.
Fig.12: on the left vertical axis (upper half)(lower half)
substitute R_G by R_B/2
and on the left vertical axis (lower half)
add R_G [fm]
on the right vertical axis remove R_G=R_B/2 and R_G
and move up the descriptions of the points
Actually, you got confused because we neglected to put on the axis title on bottom panel. It is tau*c. So, the top panel has the radii (RG or RB/2 if only RB is available), and the bottom has the tau parameter if it exists. Sorry about that. It should be less confusing now.
General suggestions to figures:
1) increase little bit labels offsets for vertical axes
in almost all the figures, especially small ones. Seems the labels
touch the axes in some cases, when small zoom is used. This concerns
also label offsets on horizontal axes in Figures 2,3,4,5.
done.
2) use italic font for writing math symbols on the plots;
Okay.
3) If it is possible, improve the organization of the sections
(mainly IV) containing figures and tables could be improved.
Sometime figures are placed too far (even two pages) from the
text where they are referred to.
What about to put together Figures 2,3,4,5 as well as
Tables I,II,III,IV and V,VI,VII? Firstly, they are related and,
secondly, it would make comparison among them easier.
well, we will play around with it a little bit, but as you know, LaTex has a mind of its own and any little games you play to put figures on a particular page will get all messed up when someone else uses a different style file or some text gets added or something. We'll do a little bit, but mostly it is for the journal to do that formatting.
Some questions:
1) There is not much information in the paper describing the
fitting procedures.
Okay, now we have put into each table, the chi2 and ndf for each fit. We also mention (just before Section II.A) that maximum-likelihood was used, though chi-square fits give almost identical results.
2) It is mentioned at the end of section III that the
systematic errors consist about 10% (for STAR data).
10% of what (in femtoscopic radii ?) ?
Yes, it is the radii. In the manuscript, we say this explicitly now.
3) Were some systematic errors included in fitting process or
only statistical errors were used?
Only statistical errors were included in the fitting process.
4) Which fitting method was utilized? Chi-square, maximum
likelihood, or something else?
5) If chi-square method was used, what are chi-square values
of the fits?
It was max-likelihood (which we say now just before Section II.A), though chi-square give essentially identical results. We now include the chi-square of all fits, in the tables.
6) The number of parameters varies significantly mainly in
non-femtoscopic parts of fitted functions as described in
section II, subsection C. It would be interesting to see and
compare the quality of the fits (chi^2/ndf) and see
covariances (correlations) of the fitted parameters.
The chi-square values are now given in the tables. We don't plan on making covariance contour figures.
7) Are the non-femtoscopic effects taken into account
when obtaining values shown in Table IX?
Ah, thanks. In those fits, Omega=1, consistent with all previous fits of this kind. We say this now in the caption.
8) Why the centrality classes 0-5% and 50-80% are chosen for
the data presented in Figure 10?
No special reason to choose precisely those two. Choosing a central one and a peripheral one gives the message that the ratio is flat for all centralities.
From Bedanga - email 22 Feb 2010
I have read the nice paper and have only
few comments as noted below.
1. Abstract:
two-pion correlation functions from p+p collisions at 200
GeV.
--> may be good to add $\sqrt{s}$
two-pion correlation functions from p+p collisions at $\sqrt{s}$ = 200 GeV.
Okay done.
2. We say:
Our results are put into the context of the world dataset
of femtoscopy in hadron-hadron collisions.
--> however in Fig. 15 we also do comparison with e+ e- data.
Unless we want to focus only on hadron-hadron collisions here.
Well, yes, you are right, we put the e+e- data there just since it was available. But it is not our focus, so we don't need to mention everything in the abstract.
3.
Line 165 &166 :
although the theoretical interpretation of the results is less
clear and well developed.
--> This seems to be not very clear.
It Sounds like it is less clear and at the same time well
developed. Is it want we mean't.
You are right, it was poorly worded. Following the suggestion from Dubna, we have worded it "less clear and not well developed."
4.
In Section V, 197 - 198
we put these results in the context of previous measurements
in Au+Au and elementary particle collisions
--> We make comparison with pp and pbar-p collisions, not sure
if we should call them elementary particle collisions.
Okay. We change it to "...in Au+Au and $p+p(\bar{p})$ collisions."
6.
281
Historically misnamed the “chaoticity” parameter, it generally
accounts for particle identification efficiency, long-lived decays, and
long-range tails in the separation distribution.
---> So do we mean to say that the lambda parameter has no
connection to Chaoticity ?
No. Just that it's never been shown to be due to anything other than the items listed. I'll just take away the "misnamed" and put "called."
7. 437 - 438
The small contamination due to electrons and kaons
impacts mostly the value of l obtained from the fit while it
was only a 1% effect of the femtoscopic radii.
--> Reference will be useful, or are they estimates from this
analysis.
They are estimates from the present analysis. We indicate this now with some more text. See around line 441, where we start "By varying the cuts on energy loss..."
8.
X-axis labels for Figures for example 1,2,3 are small.
Yes, there were several tweaks of figures requested and needed. We have revamped essentially every one. This should be fixed now.
9.
It may be good to give a chisq table for the various fits
shown to the data in Fig. 1 - 4, this may provide a better
idea which fits work well.
Specially when we discuss "None of the functional forms
perfectly fits the experimental correlation function, though
the non-femtoscopic 515 structure is semi-quantitatively reproduced by
the ad-hoc .....
Okay, we have actually included chi2/ndf now for ALL of the results, in the tables. Good idea.
10.
The dNch/eta in the Fig's 9, 11 and 12 are corrected for efficiency or raw
values from STAR ?
Those are the raw values.
While in Fig. 11 and 12 it is mentioned in the axis
that it is the average value for Fig. 9 it is not.
Okay you are right; we make it consistent now by labeling Figure 9 to say average value, too.
11. 706
the “lonngitudinal” direction
-->
the “longitudinal” direction
fixed, thanks.
- Printer-friendly version
- Login or register to post comments