Dave Underwood [12 December 2009]

I like version 2.1.  It is a major improvement.
The only big problem I see is that as far as I can tell, there is no definition
of MT but it is in the discussion extensively after line 418.

Thanks for catching this.

We have changed the sentence (now in line 427)

"The $m_T$-dependence of the radii in all cases is quite similar."
-->
"The dependence of the radii on $m_T\equiv\sqrt{k_T^2+m^2}$ is quite similar in all cases."


Should it be obvious in Eqn 5 why coulomb is multiplicative rather than addative?
I havn't thought about this or gotten to the references yet.

It is a fair question.  In principle, in using the Koonin-Pratt equation, one
convolutes with the square of the wavefunction in relative coordinates.  In this
case it should be the symmetrized Coulomb wavefunction, or better said, Coulomb-distorted
symmetrized plane wavefunction.  It turns out that, for sources smaller than the Bohr radius (good for heavy
ion collisions and very good for pp collisions), one can instead use symmetrized plane waves
times unsymmetrized Coulomb.  The reason to make this approximation is the convenient
fitting function that results, for Gaussian sources.

This is discussed for example in our review paper (Reference 11); see around Equation 10 in
that paper.

While it is a good question, my own feeling is not to discuss it explicitly in our
paper, as it really is a settled detail in the HBT procedure and not one we study in our
paper.


  The conclusions don't come across as strong as the introduction.

Yes, I agree with this statement.  I think what we have here is a presentation of
the data.  The introduction tried to motivate that this data is generically important

1.  must understand "reference" pp in order to understand AA,
2.  HBT probes dynamically-generated geometric substructure in AA which is THE crucial feature of AA
   
3.  despite massive collection of HBT data in AA and similarly massive collection in hadron-hadron, no direct connection has been possible before.
   

but drawing conclusions at this point is difficult.

I make no secret of MY conclusion from this data (that pp collisions generate strong
flow), but it is likewise no secret that STAR is actively hostile to this conclusion.
However I choose to respect the right of STAR to subscribe to a view,
so am willing to leave these important results open to speculation (in the "discussion"
section).  Maybe a later paper will go more into the physics.  (OK, not "maybe"...)

  There are a few typos
line 28 systmem
line 105 explicity
line 268 eqn needs a j


Thanks, they are fixed.  Good catch on the "j."

 

 Michal Sumbera [13 December 2009]

This version, how could it be the other way after so long elapsed form the
previous one, represent substantial improvement.

Here are my, somehow scattered, comments on version 2.1:

1. Reading fresh a new I found too many "correlations" in it. In some
sentences it apperas 4 times. I suggest from time to time use the other
word like effects e.g. in abstract
"... the presence of strong non-femtoscopic effects."


Yes, I see what you are saying.  The word "correlation" shows up
a LOT.  It is somewhat unavoidable in a paper like this.  The word itself
appears 126 times in version 2.1!!!

Using a thesaurus for "correlation" yields the following:
connection, association, link, tie-in, tie-up, relation, relationship,
interrelationship, interdependence, interaction, interconnection; correspondence, parallel.
Not really promising.  Your suggestion of "effect" can be used in some places
(like the one you suggest), but that word is really much more vague.

Firstly, we have to accept that this is a paper about correlation functions,
that we discuss femtoscopic and non-femtoscopic correlations, not to be confused
with space-momentum correlations that are very different but *probed by* two-pion
momentum correlations!

But okay I agree.  I have gone through the paper in detail, and tried to avoid
saying "correlation" too soon after using the word previously.  I think you should
find it better now.  Section II.B. is probably the worst offender, but it is better
than it was.

"Correlation" is now mentioned "only" 85 times in version 2.2, and it usually
is not repeated too close together, like it was in version 2.1.  Anyway, it should
be better, take a look.


2. line 119: It has been suggested [25-27] to express..." I'm not quite
sure if we construct the 3d-correlation function this way.

But, we do.  Probably I don't understand your point.  This will be a good thing to discuss
by evo meeting.
 

3. equations (7), (8) and tables I -IV: Consider to change the subscript
of the radii R_G etc from capitals to lower case i.e. R_G -> R_g. The
reason being that you are using (o,s,l) also in the lower case.

Well, now that you bring it up, we are using "O,S,L" often in *upper* case in
Table headings, and (as you mention in #5 below) sometimes "out" instead of "o"
as a subscript.

I agree that we need to standardize.  "o,s,l" (lower case) are the industry standard,
so I will change all - in text - to that.  R_G (not R_g) is more the standard, so I will
leave "G" in upper case.

We will need to standardize the figures as well as the text.  Since we have this
update of the figures pending (so that all symbols are distinguishable in black-and-white),
this standardization will occur simultaneously with that.  This I am asking Zibi to do,
so will do that after tomorrow's meeting.
 

4. Eq.(14): would'n it improve the readability to include in the first
line (or replace on the secon one) expression 1+\beta Y_{2,0}+\ksi
Y_{2,2}. It's a long time ago when some people heard about the wave
function of hydrogen atom...

Okay, we have done this, thanks.
 

5.  Fig.1: Change C(q_{out}) -> C(q_{o}) as well as q_side -> q_s and
q_long -> q_l.

Yes, thanks.  We will do this simultaneously with the figure upgrade.
 

6. Line 279-280:
 "...and consider whether their values make physical sense."  ->
"...and consider whether their values are mutually compatible and physical."

Good, thanks.
 

7.Line 385: "... is based on the desire to represent" -> "...is based on
determination to represent..."

okay, changed.
 

8. lines 433-436:

"Although three-dimensional correlation functions encode
more information about the homogeneity region than do
one-dimensional correlation functions, most previous particle
physics experiments have constructed and analyzed the latter."

->

"Since three-dimensional correlation functions encode
more information about the homogeneity region than do
one-dimensional correlation functions, they are also more statistics
hungry. So most of the previous particle physics experiments have
constructed and analyzed the latter."

Umm, okay, I like your wording better, so it is changed.  However, as we both know, 3D correlation functions
are not really that much more statistics hungry than 1D ones, a bit counter-intiutively.
 

9. When discussing energy-momentum conservation in pp I have not found in
the article any word about inelasticity of the collision. Its know for
long that pp (especially) does not provide the best fireball. For the
later antiproton-proton annihilation where inelasticity is 1 are
excellent. When scanning the literature, ahve you guys run into ANY
corresponding HBT data?

We have had a few more email exchanges on this, after you wrote up these comments.
I think we should discuss this more on the evo meeting.
 

10.Lines 708-9: I do not think that conservation law-driven effects are
problematic. What they may be is that they may show up less. So I suggest
wording:

"At the Large Hadron Collider, similar comparisons will be possible, and
the much higher energies available will render conservation law-driven
effects less important."

That sounds good, thanks.

11. Last but not least I still don't like your  Fig.9 with too few points
to deserve their presentation in thsi way. What about turning Figs. 8 and
8 into a single figure with different left and right scales?

Well, I don't like the left and right scales things-- too many different
radii and colors.
How about a compromise?  How about we combine Figures 8 and 9, such
that they share a common x-axis (dN/deta), but there is an upper panel (currently
Figure 8) and a lower one (currently Figure 9).  Does this make sense?
Let's talk about that at our evo meeting.  I won't change it for now until we talk.