Comments/Responses on PRL draft 1.05
for collaboration review
"B -> D** mu nu X"

Thank you everybody for the comments - they are definetely improving the  quality of the paper!

Before we reply to the comments one by one we'd like to explain one issue
which has been mentioned almost in all mails : as it appeared in our draft we had a 2.4 sigma disagreement between PDG and our result on Br(B->D^0_1 mu nu X) which is derived at page 7.
It turns out we did not interprete correctly what was used in PDG as this Br and did not scale our result up by the additional isospin factor 3/2.

This Br(B->D^0_1 mu nu X) was put in the last day in response to the previous comments and apparently it was not checked enough - sorry. The credit goes  to Leo Bellantoni who noticed this.

to explain more : 
PDG quotes Br (B -> bar{D_1) \ell^+ \nu X)
D0 measures Br (B -> bar{D^0_1} \mu^+ \nu X) Br (bar{D^0_1} -> D^*- pi+) and assumes that D_1 decays only to D*pi.  after we stated this we erroneously replaced the second Br by 1.
however, the isospin symmetry (which as we know is strictly conserved in strong decays) gives Br (D1->D* pi^+-)  = 2/3.  the remaining 33% of Br should have the neutral pion which was not used in our analysis so we need to correct for it. this is exactly the procedure which was used by ALEPH to derive the numbers used by PDG. Note that the value of R is not changed because of this argument since it applies both to D1 and D*2.

we will modify the text accordingly and the discrepancy with the PDG will be gone (PDG 0.74 +- 0.16 vs. D0 0.48 +- 0.09). 2.4 sigma becomes 1.4 sigma.


5/26/2005 from Iain Bertram

General comment: In the introduction you state that the ratio is not very model dependent. Unfortunately I do not see any comment/comparison between the measured and predicted value. Did I miss something here.

"not very model dependent" still has a spread from 0.4 to 1.6.. the latest theoretical papers prefered somewhat lower numbers around 0.4-0.7 but  it looks like they were motivated to explain the "experimental" value of ~0.4. this value 0.4 is an (unpublished) result of the averaging of several preliminary results on Br(B->D*2 munuX) which later were superceeded by upper limits in the corresponding publications.  the average "experimental" value was never updated after the publications but is referenced sometimes by the theorists. we don't agree with the calculation of R using the upper limits and we also think that the analysis of the systematics was not adequate.

all this was said to show that there is some deal of controversy in this value R. this has been discussed multiple times with EB and the conclusion was that we should ignore unpublished results and avoid this comparison otherwise it will take too much space and will be controversial. in any case we'll be the first published result which directly measures R including the dedicated analysis of its systematic errors and we hope it will attract some attention from theory.

some additional info on this topic discussed with EB can be found on the top of this page
http://d0server1.fnal.gov/users/nomerot/Run2A/Dstst/prl_comments.html

Abstract:

Not sure that the use of luminosity is correct: Suggest Using a data sample with 460 pb^-1
replaced "includes approximately 460pb-1" -> "corresponds approximately to 460 pb-1"

the product of branching fractions of the
ok

Page 5: column 2: para 1, line 16-17:
a better sentence may be:

The MC resolution was scaled up by 20% to account for the difference between the data and the MC.
ok

5/26/05 from Greg Landsberg

First of all, let me congratulate you on an impressive measurement with nearly 0.5 fb^-1 of data, which improved on our knowledge of the B -> D** branching fraction accuracy by a whole factor of three and lowered its central value by a factor of two!
thank you, please see the comment on the top of the page re the factor of 2.

The paper is well-written, and I mostly have some style suggestions, along with a handful of physics comments. I’ll use the following notations: P – page, c – column, p – paragraph, fp – full paragraph, and l – line.

<>GENERAL COMMENTS:  <>1)     My main general physics comment on the paper is that it’s not clear at all from the text whether you used just the Dbar^* \mu^+ final states, or their charge conjugates as well. Most of the paper text suggests the former; however, Fig. 2 refers explicitly to D^* states decaying into D*^+ pi^-, which are charge-conjugates of the Dbar^* states the paper mostly refers to. Please, state it clearly whether you combined the two or not for the determination of th branching fractions. If you haven’t than first of all why (your statistical error is considerable and reducing it by 40% would’ve helped) and second of all, what Fig. 2 refers to?
we have ref.[10] that sais the standard thing about change conjugate states - of course, we do use them. we think the reference appears in the first appropriate place but we'll double check it.

2)     A couple general style comments: according to the PRL style, (stat) and (syst) have to be typeset in \rm: \mbox{\rm (stat)}. Please, change this through the text. Also, the more modern notation for branching fraction is {\cal B}, not {\it Br}, and I believe we’ve used it in a number of published papers. I suggest switching to this notation through the text, which would also save you a few lines and make formulas shorter.
ok

<>3)     Since your measurement improves on the world average so much, it would be very useful to add the new world average to this paper, which should be very close to your number. Since the new average represents a two sigma shift down from the current one, it’s worth stating explicitly in conclusions. 
we don't want to compete with  HFAG in doing the new Br averages. the discrepancy is much reduced now, see the top of the page.

4)     You use “uncertainty” and “error” interchangingly in the paper. Please, switch to a common style: “uncertainty” (see line-by-line comments).
ok

5)     The statistical uncertainty on the number of mu+D0bar events you quote: 216870 +- 1280 is three times larger than the square root of the number of events. Since the background contribution is presumably rather small (and well determined!), why is the uncertainty so large? Note that the other counts: N(D*) = 55450 +- 280 and 31160 +- 230 have their statistical uncertainties close to the square root of the number of events.
the bkg for the muD0 signal is much higher than for the muD* signal. (the plot is available in the analysis note 4324). this explains the higher uncertainty - it indeed comes from the bkg. the muD* signal is almost bkg free and there the uncertainty is close to the sqrt(N).
 
6)     In Fig. 2 you show the D** mass fit with the D* mass fixed at its nominal value. Does your measured D* mass agree with the PDG value indeed? If so, presumably you have applied track momentum correction, but it’s not mentioned in the text. I think a sentence to that effect should be added; probably in the same paragraph where you describe the momentum smearing fudge factor. Also, when you do the fit, are the positions and/or the widths of the two BW resonances fixed at their w.a. and the detector resolution, respectively? Again, of so, you should say that, and if not, you should quote the mass/width obtained as a result of the fit, with their uncertainties.
we plot m(D**) = m(D**)_reco - m(D*)_reco + m(D*)_PDG - this is the exact meaning of D* mass constraint mentioned in the text. we do not apply the track momentum correction explicitely. the relative position of the two BW (mass difference between them) and their widths are fixed to the PDG values. the mass resolutions are taken from MC. we do say this in the text. the mass of D1 is a floating fit parameter which comes out from the fit as 2415  MeV (PDG 2420 Mev). the 5 MeV difference is indeed due to the wrong momentum scale in the tracking. this difference does not affect the measurement and it is not mentioned in the text. the effect of the momentum scale on the mass difference between the D1 and D*2 which is equal to 37 Mev is negligible. The effect of the momentum scale on the BW parametrization has been studied and was found negligible as well.
 
7)     Finally, at the end of P6, where you quote the main result, it may be worth mentioning that the numbers of events in the two D** resonances obtained from the fit are strongly anticorrelated; consequently, the precision on the branching into either of the two states is higher than that for the exclusive ones.
ok
 
LINE-BY-LINE COMMENTS:
 
1)     ABSTRACT: A few comments here. First, I’d add “in 2002-2004” after the “DØ detector”; second you should remove “product” in the first line: first, it should’ve been “product of,” but more importantly the sentence now reads as if you determined the product of branching fractions in two D** mesons, which is not the case. Your formulas already show explicitly what you’ve measured, so no need to confuse the reader with the word “product.” Finally, the abstract now is way too long. (I am not even sure that not-in-line formulas are allowed by PRL.) I suggest removing the first branching fraction from the abstract (it’s the sum of the second and third, anyway) and typeset the remaining three formulas in-line. For the first two it’s trivial; for the last one use B1/B2 instead of \frac{B1}{B2}. That would keep the abstract clean and under the PRL limit.
ok, all good suggestions 

2)     P4,c1,p1,l4: states -> mesons.
ok
 
3)     P4,c1,p1,ll8-11: technically, L and j are not angular momenta but the corresponding quantum numbers; particularly the sentence about j being a sum of the spin and orbital angular momentum is a bit misleading. However, I don’t really have a good suggestion on how to phrase what you are saying without making the sentence too awkward, so I’d leave it up to you to see if you want to phrase it a bit more precisely.
i see your point, will think what we can do..
 
4)     P4,c1,p1,l11: I’d add a sentence here: “Collectively these states are referred to as D** mesons.” Otherwise the next sentence is confusing, as notation D** is not present in any of the members of the two doublets.
ok
 
5)     P4,c1,p1,l14,16 and further in text: standard notations for the s- and d-waves are lower case “s” and “d.” Also, since you denote the total angular momentum as “j,” you should use “l” instead of “L” for orbital momentum further in the text.
ok
 
6)     P4,c1,p1,last two lines:  to have large decay widths -> to be broad.
wording changed
 
7)     P4,c1,p2,ll2-3 below Eq.(1): for the orbitally excited D mesons -> for these states.
ok
Next line: in the infinite CHARM QUARK mass limit.
ok
Next line should read “O(1/m_c),” rather than “1/m_c.”
ok
Finally, I’d rephrase the last sentence as: “Together with the MEASUREMENT of … DETERMINATION of $R$ will … and factorization HYPOTHESIS FOR the non-leptonic …”
ok
 
8)     P4,c1,p3,ll7-8: rephrase as: “… DELPHI [7] experiments, with only the latter, preliminary result providing a measurement of ${\cal B}(B \to D^*_2\ell\bar\nu)$, while others only setting upper limits …”
i actually like it better the way it is - will consult the EB
 
9)     P4,c1,p4,ll1-2: includes approximately -> corresponds to $\approx$.
ok
 
10)  P4,c1,p4,ll9-11: The whole sentence “The pseudorapidity…” can be made a reference, if pressed for space. Also, “An outer” -> “The outer”.
ok
 
11)  P4,c2,ll1-2: D** MESONS PRODUCED in … manifest THEMSELVES as …
ok
 
12)  P4,c2,p3,l3: and THE charge …
ok
 
13)  P5,c1,p1,ll6-8: error -> uncertainty; also “error on the vertex” is meaningless. You should say “uncertainty on the vertex position” and specify the direction – is it in z or in the XY-plane? Also, “B-vertex fit” (need a hyphen).
ok, good point
You require the chi^2 of <25 for the Nd.o.f. = 3, which is very generous. Why is the cut so high?
this cut was designed as a clean up cut to get rid of the tails and to have efficiency close to 100%
Finally, here, and three lines down, spell “3” out as “three” (as small numeral).
ok

14)  P5,c1,p1,l13: error -> uncertainty.
ok
 
15)  P5,c1,fp1,l9: … to the primary and secondary ($D^*$) vertices …
ok

16)  P5,c1,fp1,ll11-12: error -> uncertainty. Next sentence should be rephrased as: “To select tracks belonging to the $B$ vertex, the ratio of IP significances … greater than four.”
ok
 
17)  P5,c1,fp1,last line: 1 -> one.
ok, Jerry's suggestion was "unity"
 
18)  P5,c1,fp2, second sentence: remove it entirely – it’s a straight quotation from the figure caption.
ok
 
19)  P5,c1,fp3,l3: with THE numbers…
ok
 
20)  P5,c1,Eq. (2) and further in the text: “Res” should be typeset as {\rm Res}.
ok
 
21)  P5,c2,p1,l1 and later in the text: I believe that “formulas” is a more conventional plural in the American English.
Susan did not correct it - will double check with her
 
22)  P5,c2,p1,l10: parametErization.
ok
 
23)  P5,c2,p2,ll2-3: same-sign combinations (a hyphen). Also, it has not been defined what you mean by the same-sign combinations. I’d say explicitly that you require the muon and the pion to have the same (negative) sign. Calling them “wrong sign combinations” is probably better too.
ok
 
24)  P6,c2,fp1,ll4-5: errors -> uncertainties.
ok
 
25)  P7,l3: remove “this branching ratio”.
ok
 
26)  P7, Ref. [1]: add {\bf B} to the last two Phys. Lett. citations, which don’t have it.
ok


5/26/2005 from Darien Wood
Hi Guennadi, Sergey and Andrei,
 
    Very nice job on the D** paper.  I have just a few comments:
 
Main comment:  You never tell the reader how you triggered on these
events.  It seems strange to omit this information.
ok, added a sentence

Abstract: I wonder if it acceptable to use the more compact notation
$Br(\bar{b}\to B \to (D...)\mu..)$ in place of $Br(\bar{b}\to
B)\times Br(B \to (D...)\mu...)$, at least here in the abstract?
the notation that you propose is not commonly used. a more compact notation could be Br(b \to D ....) which is used by PDG and is actually used in the paper when we quote PDG. however it is less informative and we'll need to introduce the one we are using now anyway when we calculate Br(B \to D**..) in the discussion section at the end. we'll check with EB on this.

we also replaced {\it Br} with {\cal B} which makes the formulas more compact.

p 4., col 1, last paragraph:
    "1.8 T toroids" -> "1.8 T iron toroids"
ok
 
p 5., fig 2: In the text, when you refer to definite charges, you
seem to consistently start from the D*- and keep all charges
consistent with this origin.  But in the figure legend, you use D*+.
I assume that you are always combining with the charge conjugate
state, but it might be easier to follow if you also label the figure
with the D*- convention.
ok, good point

p. 7:  When you give the result for R, it might be interesting to
remind the reader of the prediction from HQET and to say that the
measurement in in agreement.
ok


5/26/2005 from Saclay group
             Dear EB,

             On behalf of the Saclay group, please find below our comments .

             This Letter definetely deserves publication, since it shows clearly that the D0 collaboration
     has a tracking system which allows that sort of precise measurements with semi-inclusive final states.
             It is a clear demonstration of the good quality of our data, as far the SMT and CFT are concerned,
     together with our muon tagging, and the publication will be an encouragement to all involved in spectroscopy
     of B mesons.
             Since none of us is an expert in spectroscopy, we just say that the whole thing seems OK : easy and
     precise reading are a strong point in favor of this letter.

             Cheers.                                                           Pierre
thank you
 
5/24/2005
from Terry Wyatt
I had the following comments on the paper:

Global comments on the terms used throughout the paper:
Resonances can be narrow or broad.  Widths can be small or
large.  I think we should not employ the mixed terms "narrow" and
"large" when refering to "widths".
I don't think the term "axial plane" is appropriate.  We should
probably use "plane transverse to the beam direction" or some such.
ok, "narrow" replaced by "small" specifying the magnitude (O(10 MeV/c^2) for small and O(100 MeV/c^2 for large).
"Axial" is defined now as a reference.

Abstract:
I think we need to mention that the luminosity of 460 pb-1 was collected 
in
proton-antiproton collisions at the Fermilab Tevatron collider at
centre-of-mass energy 1.96 TeV.
ok, this was also suggested by others

p4, c1, para 3:
I think it is inappropriate in a publication to refer to a 5-year old
unpublished experimental result (ref [7]).
this is the only existing result which has a measurement of Br(B->D*2 l nu), all others have a limit. 
we discussed this issue multiple times with EB and this is what came out of these discussions.
we could bring it up again with EB to decide what to do. 

p4, c1, para 4:
If I have understood correctly all tracks used in the analysis are
required to have at least 6 CFT hits?  I assume that can be any mix of 
axial or stereo hits?
these are the standard 3D hits used by tracking. the p14 tracking (and i suspect p17 tracking as well) is not using purely 2D hits though it could..

This requirement gives an implicit restriction on the |eta| of tracks
that can be used in the analysis.  I'd suggest giving the track cuts
in the detector section and giving the |eta| that this corresponds to,
rather than the nominal |eta| acceptances given at the moment.
as it's described now - this is exactly what has been done. the N_CFT>6 will indeed create a cutoff
at eta =~ 1.6 but because of the z position smearing this will be a poorely defined cutoff. 
we prefer to leave it as is.   

(The duplication of the track quality cuts in p5, c1 para 1 and 2
should definitely be avoided.)
track quality cuts are not explicitely defined in the paper. we could define them and use this 
definition later for pi** but this will increase the total amount of text. 

p4, c2, para 2, first sentence:
Should be "\mu^+Dbar0"
              ^^
the sign of the muon was not specified deliberately since the sample includes both sign combinations 
(correct and wrong). replaced \mu -> \mu^\pm.

Is it really necessary to give the |eta|<2?  In the detector section
we have stated already that muons are identified within this range.
(Strictly speaking there may be a distinction between muon detector
eta and "physics" eta, but that is a detail that is probably
unnecessary to worry about in a letter.)
we may have chosen to use a restricted coverage of the muon system for the analysis so we think it's 
appropriate to specify it here. 

p4, c2, para 2, last but one sentence:
Looks like there might be a space missing between "Dbar0" and
"exceeds"?
ok

p5, c1, para 1:
I guess the explanation of how the proper decay length is calculated
is given in ref [11].  I think a reference at "proper decay length" is
probably needed?
ok

Is the level of combinatorial background in the delta_M plot any less
after the cuts on proper decay length have been applied?  If so, for 
publicity purposes perhaps we could show fig 1 after having applied these 
cuts?
the signal looks pretty much the same - the internal note has this plot. we decided to show the plot 
which effectively demostrates our yield per pb-1 for D*mu.
 
p5, c1, para 2:
I'd suggest:
"The ratio of IP significances .... greater than 1."
should become:
"In order to select tracks belonging to the B vertex, the ratio of IP
significances of the track with respect to the primary and D$^*$
vertices was required to be greater than four and the IP significance
with respect to the primary vertex was required to be greater than 1."
ok

p5, c1, para 4:
"in each of them" -> "respectively"?
ok

p5, c2, para 2:
What exactly does "the widths of the D_1 and D_2 were fixed in the
fit" mean?  Was a constraint (with the appropriate uncertainty) 
added to the fit or were the widths
actually fixed?  Does this explain the large systematic coming from
the widths in Table 1?  Since we have an order of magnitude more data
than previous experiments, are we not able to determine these widths
more precisely?  If not, then what is limiting our ability to do so:
intrinsically poor momentum resolution (about which we can do little) 
or poor understanding of our momentum resolution (which we can in
principal do something about)?  Since the uncertainty on the widths
currently dominate the overall uncertainty, what are the prospects 
for future improvements in these D** branching ratio measurements?  
(A mischievous referee is going to ask us why we didn't fix the 
D** masses too?  ;-) 
1) widths and delta m : the widths and dm were fixed in the fit. we varied them within their uncertainties one by one to estimate the corresponding systematic errors. the same procedure was used by the LEP experiments and we wanted to be consistent with them so the results can be compared. The widths and dm are better measured in charm production experiments : their PDG values come from there and we cannot compete with them (remember that we are studying the B decays to D** munu). This was the motivation to have them fixed. It gives a better statistical precision for our fit.  
2) Re the prospects : CDF measured recently the masses and widths of narrow D**in the charm production. this is a conference result now and they plan to publish it later this year. we were considering to use their numbers and had some discussions with them a few months ago. their result improves accuracy on the width of D1 by a factor of ~2.  another thing is that CDF together with Belle measured a larger width for D*2 compared to the one from PDG. The wider D*2 would fit better our data. since the width of D1 is the main contributor to our systematic error it will make sense to have an update of this D0 result after the CDF result is published and PDG updates the D** parameters. as a cross check we made fits with the CDF/Belle parameters and the central value was within the current systematic errors so we should not have any surprises.
3) re the momentum scale : i paste here my response to a similer question by Greg
the mass of D1 is a floating fit
parameter which comes out from the fit as 2415  MeV (PDG 2420 Mev). the
5 MeV difference is indeed due to the wrong momentum scale in the
tracking. this difference does not affect the measurement and it is not
mentioned in the text. the effect of the momentum scale on the mass
difference between the D1 and D*2 which is equal to 37 Mev is
negligible. The effect of the momentum scale on the BW parametrization
has been studied and was found negligible as well.
4) our mass resolution is not critical for this analysis - the resonances are wide compared to the resolution. note that the contribution to the systematic error from the resolution is much smaller than from the widths.
 
p5, c2, last line:
Insert the word "respectively" before the full-stop?
ok

p6, c1, para 2:
Is the world average Br(b->D*l\nuX) from e+e-?  Are we certain it is
appropriate to use this proton-antiproton?  Is there direct evidence
that the mix of b hadrons is the same in e+e- and hadron colliders?
(This is more for my education than as a comment on the paper.)
here the question is whether it is appropriate to use Br(b->B)=39.7% which is indeed based on the e+e- results. when PDG averages the Br it makes this assumption that the Br are the same at LEP and at the Tevatron. this is motivated by the assumption that the fragmentation of the quarks after they are produced is the same and the coherence of the two produced b quarks in the e+e- processes does not affect the fragmentation. the only measurement of Br(b->B) at the Tevatron is the CDF RunI result which is in agreement with the PDG number. note that all our and CDF previous analyses made this assumption as well and the issue has been discussed in depth by the same EB for the Bs->mumu publication. 

p6, c2, para 1:
"Contributions from B_s ..... neglected."
Have upper limits on B_s and Lambda_b -> D** been measured, or is this
the result of a back of the envelope calculation?  Is any more
explanation needed in the paper?
this is a result of the sample composition estimates which we've done for the lifetime ratio analysis. the Bs contribution was ~1% and the Lb contribution was not worth mentioning. we probably should add this reference after the sentence.
 
p6, c2, para 2, 1st line:
"of" -> "on"
ok

p6, c2, para 3:
I think the phrase "fitting the narrow D** .... described above, and"
could all be deleted without any bad effect.
ok

p7, c1, para 1:
Shouldn't we replace the sentence
"The PDG does not .... D*_2 meson."
with
"We present a first determination of
the B->D_2^* branching ratio = XXXX, under the assumption that the
branching ratio for D*_2 -> D*pi decays is 30\pm6\% [12]"?
ok, in fact PDG gives a limit on this Br therefore indeed your suggestion is more appropriate.

Shouldn't we add a comment in the conclusions that the ratio is
consistent with the expectations?  
please see our response to the same question by Iain. 

Ref 10:
I think the information that "B refers to B0 and B^+" is sufficiently
important to move into the text.  I found myself worrying about this
throughout my first reading of the paper and only came across it much
later.
ok

A purely nerdy question:  How was the .pdf file produced?  I
couldn't copy text using Acrobat into an ascii file, as I normally
do with draft papers.  This made typing in my comments rather more
tedious than usual!
the .ps was created by dvips at my clued0o node. that's what i do always...

5/20/2005 from Jianming
Dear all,
 
Congratulations for a nice paper! I have a few comments, mostly about clarifications.
 
1) notation: for people not familiar with this kind of physics, it will be really helpful to use consistent notations. For example, the paper uses both D_1 and D_1^0 for the same particle I assume. If so, I'd suggest to use one notation. Same is true for D_2^* and D_2^{*0}.
there exist both the charged D** and the neutral D** - all agrumens in the introduction are applicable to both. however for our signal we use the D**0 only. may be indeed this transition could be made more explicit in the text.

2) page 4, left column, 1st paragraph: it is implicitly said that D** refers to D_1, D_2^*, D_0^*, D_1^'. Can it be more explicit?
ok, we had the same comment from Greg (and a good suggestion how to fix this)

3) page 4, Fig. 1: the x-axis label uses M_{Kpipi} while the caption uses M(Kpipi). Can you make them consistent? Same for M_{Kpi} vs M(Kpi). Depending on what change you make, you may also need to change the "\Delta M" definition in the 3rd paragraph, right column, page 4. 
ok

Same comment for Fig. 2.
ok

4) I don't quite understand the meaning of Br(b->B). Is it the fraction of a b-quark fragments into a B meson? If so, I'd say defined it explicitly when it is first referred in the text. If not, still define it.
yes, Br(bar{b} -> B) is the fraction of anti-b-quark fragmenting to B0 or B+. ref.[10] explains that B refers to either B0 or B+. we'll add explanations to the text on the exact meaning of Br(bar{b} -> B).

5) page 6, left column, the 1st formula: I'd think the branching fraction Br(D*->D0 pi) should be cancelled in the ratio since both D* and D** are reconstructed through D*->D0 pi. Why the formula has this there?
the formula is correct. indeed the D* is reconstructed through D* -> D0pi+ while it also decays through neutral modes that we don't use. this Br takes it into account.
 
6) page 6, the results: I don't understand why all three fractions have statistical uncertainties of 0.011. In particular, I thought the first one is just the sum of the next two (0.129+0.051=0.180). Why the statistical uncertainty on the sum is the same as the two components? Why the statistical uncertainties are the same for D_1 and D_2^*?
the sum has a better accuracy because the errors on D1 and D*2 are anti-correlated. an easy way to understand it : the total number of events in the two peaks (which really looks like one peak) depends weakly on the parameters of the resonances - it's just the number of events above the bkg. the same errors in the three cases is just a coincidence.

<>7) page 6, last line: is it expected that D_1 decays only to D* pi? If so, should there be a reference?
it's parity selection rules which are (kind of) mentioned in the introduction. we'll see if this can be made explicit somewhere.
  <>8) page 7, left column, 1st paragraph: Shall there a statement/explanation on the difference between our measured value and the PDG value of Br(B->D_1 l nu X)? Why not quote our measured Br(B->D_2^* l nu X) here?
please see the top of the page re this item - the difference is much reduced now.

9) page 7, references:
  Ref.[8]: please drop the 2nd half of the reference (the paper by Diehl and LeCompte), and add  the title "The Upgraded D\O\ Detector" to the 1st half of the reference.
ok
  Ref.[11] has been published, please update.
ok

10) please make sure to update with the latest author list.
ok
 

5/20/2005  from Ron Madaras
Guennadi, Sergey and Andrei,

1) You did not use the latest author list or the latest list acknowledgment paragraph at the time of your 5/11/05 draft 1.05 (new ones were posted on 5/10/05). Even newer ones are now available and must be used.
ok
Please check for the latest versions of the author list,
list of visitor addresses, and the acknowledgment paragraph
before you distribute another draft and before you submit
the paper for publication.
(see http://www-d0.fnal.gov/~madaras/authorlist.html)

Do not change or update the author list after the paper is
submitted; it should stay frozen for your paper at the time
of submittal.

Please let me know if there are any changes in the author list or
list of institutions in the proofs that you receive back from the
publisher.

2) In the short detector description, you should at least mention that there is a liquid argon calorimeter. You don't have to give any details of it, since you say that only the muon and tracking systems were used in the analysis.
ok

3) Table 1: why are some of the errors for the ratio (last column) much greater than the preceding 2 columns? If some errors cancel in the ratio, I would expect the numbers in the last column to be smaller than the preceding 2 columns.
some errors indeed cancel out (like the ones coming from Br which cancel exactly) but some contributions are more independent for D1 and D*2 so the total error on their ratio is increased as it should. note that we fit directly the ratio so these systematic errors take into account correlations between N(D1) and N(D*2).

4) p. 7: you get 0.32 ± 0.06, with PDG = 0.74 ± 0.16. This is almost 3 sigma different isn't it? I think you need to say something about this large difference. The reader might think that since D0 is so "wrong" with this measurement that they can't trust the other results in the paper. (I know that this isn't a proper conclusion; I'm just afraid some readers might think this way). Is there anything you can say about the almost 3 sigma difference?
it was 2.4 sigma discrepancy. please see the top of the page re this discrepancy - it's much decreased now.

5) Not being very familiar with D** mesons and some of the notation you used, I found that the paper was hard to read. I was constantly trying to figure out what particles/states you meant. I think you need to spell things out a bit more. Some specific suggestions/problems:
a) Buried on p.6 you have the sentence "The D** notation stands for D_10 or D_2^*0 or both of them, (D_10,D_2^*0)." You should put that sentence in the very first paragraph, the very first time you use the notation D** (13th line of 1st paragraph).
ok, also suggested by others

b) The very first sentence in the right column on p.4 talks about D** production in B-->...., and says that it manifests itself as D^*-pi^+, but there are no pi's in the B-->... formula. That is confusing.
good point. in principle X stands for the pion in the initial decay chain. may be we should add explanation like " ... where X stands for one or more pions in the final state."

c) Fig 1 at the top says B-->D^*,mu,nu,X; But shouldn't the D^* be D^*-?? In the paragraph that refers to Fig 1 you say explicitly D^*- candidates.
Fig.1 shows both sign combinations (as stated in the legend) so the signs were omitted from the decay chain deliberately. we'll change the text to explain it better.

d) Five lines from the bottom of the right column on p.4 you have D^*. But again, do you mean D^*-?? You start this paragraph talking about D^*- candidates.
ok, "D*" -> "D^{*\pm}"

e) First line of first paragraph on left column, p. 5, you have D^*. But again, do you mean D^*-??
ok, "D*" -> "D^{*\pm}"

f) Fig 2: In the equation at the top, you have D^*. Previously you have written D^*-. But is this D^*+ now, since for data you have D^*+pi^-? This is confusing.
ok, signs will be changed

5/18/2005  from Leo Bellantoni
Hi,
    I have only had a brief chance to look at the paper, alas!

    Q1: At the top of pg 7, it appears that the number 0.32% was obtained by dividing 0.129% by 0.397.  That would not take into account the isospin factor of 2/3 which was actually mentioned on pg 6 as Br(D*-  ->  D0bar pi-) = 0.677.  If I've got this right, then the number 0.32% should actually be something like 0.48 +- 0.15%, thereby eliminating a 2.4 sigma discrepancy with the PDG, which was not actually commented upon.
Leo, we really appreciate your finding, please see the top of the page re this issue.

    Q2: Is this somebody's thesis?
unfortunately, no
                               

5/17/2005  from Dmitri Denisov
Gena, Sergei and Andrei!

Congratulations with interesting and well written article! Please, find below my questions/comments/recommendations:

1. Abstract:
a) providing integrated luminosity number in abstract (and in the text) is not informative. Triggers which pass the events for the analysis are most probably prescaled. The set of prescales in the "next" 460pb-1 will be different, so you can't even extrapolate number of events detected based on this number. I recommend to remove this number, rather refer to data set using time: like in 2002-2004 data set.
we agree with your arguments above. however we think that this number is relevant for the calculation of the average yields per pb-1 as it compares our "all integrated" performance to the competition and the comparison will be much in our favor. we'll add a sentence about the triggers used to collect the sample as suggested by Darien.

2. p.4:
a) I recommend to put ref. [10] as early as possible in the text and may be even in abstract. Keep in mind many people are reading abstracts only and specific mentioning of "mu+" might be confusing.
we don't think a reference can be put in an abstract - we'll give it a try to have it there explicitely. it is not unusual to have Br in this form in abstracts. otherwise we think we make this reference as early as possible.

b) par. 4: in the sentence about eta use "proton beam direction". Remove word "outer" in muon system description - there is no inner or outer muon system in D0. Reference [9] sounds like muon reference, while it is Run I D0 NIM reference - which is almost useless in this case (both for central tracker and muon system). Use standard Run II muon system reference from Ferbel's updated detector paragraph.
ok

3. Fig. 1
a) I suggest to use scales in MeV - will have much less "zeroes" in axis labels.
this may help to the vertical scale but  it will make things worse for the horizontal scale. we prefer to have the same units everywhere so we prefer not to change it.

4. p.4, par. 6:
a) here and in some other places of the article you use abbreviations of particles/antiparticles/charges rather "freely". I suggest you select single "format" and use it consistently through the paper. It will be much easier to understand. For example, in the first sentence of this paragraph you use mudbar0 while in the last mu+dbar0; in the second sentence you use d0 decaying into k+pi-, while it should be dbar0, etc.
of course, our goal was to be as clear as possible and to avoid any jargon or confusion. we'll address any specific problems that you or others will find. going in order:
- there was a reason why it was muDbar0 in the first case : this sample includes the both sign combinations. when we mentioned it later it refers to the signal with correct sign combination.
- D0->Kpi was a typo - thank you.
 
b) in the sentence before last pt of d meson is in GeV/c, not GeV/c**2;
ok

c) you select muons with pt > 2GeV. In order to penetrate muon iron (and calorimeter) muon needs at least 3 GeV of energy. Are you using A-layer only muon segments?
we are using only nseg=3 muons so the A- layer only muons are not used.. here is the muon pt distribution shown below (ptle) 
the color coding of histograms is: blue = data, red = MC before weighting, green = MC after weighting


(please see this link if you want more explanations about these distributions http://d0server1.fnal.gov/users/nomerot/Run2A/Dstst/prl_comments.html)
this distribution is a convolution of the penetrating muon pt distribution, their eta distribution and trigger/tracking efficiency and  indeed it peaks at ~ 4 Gev and we think it makes sense. cutting at 2 Gev we try to use as much good muons as possible.

5. p.4, par.7:
a) in the first sentence you say that we require "additional pion" which sounds like we are able to identify pions. We are actually just adding charged track assuming its mass is equal to pion mass. You might clarify this otherwise it could cause confusion for people outside of D0.
ok, good point. will revise the text

6. p.5, par.1:
a) at the begining (and in the other places in the paper) of this paragraph you provide details of the cuts as 2 SMT hits and 6 CFT hits. This is really hard to digest for the reader - are these "tight" cuts or what are you trying to say here? I recommend just dropping this information if this is not critical or adding something like 2 out of 4 hits or similar clarification.
ok, we should revise it indeed. Terry also proposed to define it as track quality cuts and use it later for pi**. we probably could add the total number of SMT and CFT layers to the detector description.

7. p.5, par.5:
a) when you describe k and k0 I would replace "magnitude" with "module" - you are talking about length of 3 vector;
what about omitting the word "magnitude" completely? since in the formulas is comes as (kz)^2 and z is just a number it should not cause any ambiguities.

b) when you describe mass resolution 2 Gaussians are used - why two?
this was requested by the EB and B group - indeed two gaussians describe the mass resolution better than one gaussian. a detailed discussion is available here
http://d0server1.fnal.gov/users/nomerot/Run2A/Dstst/prl_comments.html (dated 1/18/2005).

c) the sentence before last starts from "typically" which has vague meaning. Could you state that the mass resoltuion is in xx to yy range or in zz to ww range with 95% probability or similar.
since it's two gaussians with parameterizations different for D1 and D*2 it will need a couple fo sentences - we'll give it a try.

8. p.5, par.6:
a) you have references to [12] before, but now you are calling this reference PDG. Either name this reference PDG first time it is mentioned or just drop this naming of the reference.
ok, good point

9. p. 6, par. 2:
a) here again symbols started to be mixed. In some places you have "l" in another "mu" - it needs either explanation or (as you measure muon channel only) just keeping the same style;
this has a reason : PDG numbers have "l" since they were typically using both muons and electrons  and assumed universality. for our analysis we used muons only so we use "mu" for our numbers. we'll double check the consistency.

b) the last formula is fine, while first one (may be because of abbreviations use) is difficult to digest - do you have reference (Note)
where transformation is described?
this formula is described in a bit more detail in the analysis note 4324 v4.0. as suggested by Jianming we'll add explanation what Br(b->B) means.

10. p.6, par.6:
a) Replace "defined" with "presented" - some of these numbers are your
measurements, some are from PDG.
ok, though we don't see much difference between the two options.
2. What you actually present is not B to D** branching, but product of these branchings times (b to B) and (D** to D*). You should consider rephrasing sentence to represent what is actually measured. The same, by the way, is true about abstract.
ok

11. p.7, par. 2:
a) I suggest you add after R it's definition once more (from formula (1)). It is not fundamental abbreviation and it is not mentioned in the article except in the very begining, so reader could forget getting to the summary what R means (I did (:).
we think it will take too much space to define it twice. we can number it and reference the number.


5/16/2005 from Michael Sanders
Dear Authors, EB-members, others,

I had a look at the B -> D** decay paper draft. It's an impressive
piece of work, and again a proof that the D0 tracker is not so bad
after all. I have a couple of questions and comments, listed
below. Since I'm not really an expert in this sort of measurements, my
comments are mostly editorial.

Yours,
Michiel

========================================================================


Abstract:
 . I know it doesn't matter for this particular measurement, but
   quoting an integrated luminosity only makes sense if you mention what
   it is: $p\bar{p}$ colissions at $\sqrt{s}=1.96$ TeV. I don't think
   mentioning "accumulated with the D0 detector" is sufficient.
ok
 . Since R (eq (1)) seems to be an important number, shouldn't this be
   mentioned in the abstract too? Or maybe even instead of the long list
   of branching fractions?
abstract is very limited in space and we were already told that we have it too long. also we think that R is not something we measure directly but rather deduce from our measurement so it belongs to the discussion section.
 
 . You don't mention until much later, that charge conjugate states
   are always implied. Shouldn't that somehow be made clear when you
   quote numbers in the abstract?
ok, Dmitri has the same question. we'll try to do smth about it.

p 4, 1st column, 1st paragraph:
 . Can the definition of D** as the four j=1/2, j=3/2 D states be made
   more explicit? Maybe in the first sentence of the paragraph?
ok, also suggested by others

p 4, 2nd column, 2nd paragraph:
 . There is a space missing between "D0" and "exceeds" at the end of
   the paragraph.
ok

p 4, 2nd column, 3rd paragraph; Fig 1.
 . The quality of the fit to the peak doesn't seem to be too great
   (see the highest data point). Is this somehow taken into account in
   the uncertainty on the number of D* candidates (55450+-280)? Have you
   tried different functions to get a better description of the peak?
indeed we are underestimating the D* yield by ~100 events due to this (out of total 55k D*). for our Bd mixing studies we know that fitting the signal with 3 gaussians fixes it. for this analysis we neglected it since it's a 10-3 effect - negligible compared to the other sources of errors.
 
 . Can the y-axis in Fig 1. be labeled with "... / 0.3 MeV/c2" instead
   of "... / 0.0003 GeV/c2", or is that too confusing because of the
   x-axis which is in GeV/c2? I don't really like the long list of
   0's...
Dmitri had the same comment. we'd like to use the same units but may be since two people requested it it should be changed..

p 5, 1st column, 1st paragraph:
 . "the error on the B vertex < 0.5 mm": this must be the
   uncertainty (!) on the B vertex "position" or "decay length". An
   uncertainty on a "vertex" doesn't make sense.
ok

p 5, 1st column, 3rd paragraph:
 . What is meant by "the D* mass has been used as a mass constraint"? 
Greg had a similar question - here is the answer:
we plot m(D**) = m(D**)_reco -
m(D*)_reco + m(D*)_PDG - this is the exact meaning of D* mass
constraint
mentioned in the text. we do not apply the track momentum correction
explicitely. the relative position of the two BW (mass difference
between them) and their widths are fixed to the PDG values. the mass
resolutions are taken from MC. we do say this in the text. the mass of
D1 is a floating fit parameter which comes out from the fit as 2415 
MeV (PDG 2420 Mev). the 5 MeV difference is indeed due to the wrong
momentum scale in the tracking. this difference does not affect the
measurement and it is not mentioned in the text. the effect of the
momentum scale on the mass difference between the D1 and D*2 which is
equal to 37 Mev is negligible. The effect of the momentum scale on the
BW parametrization has been studied and was found negligible as well.

 
p 5, 1st column, last paragraph:
 . "the distribution was fit using ... functions": I think the correct
   English is: "functions were fit to the distribution". 
ok

p 6, 2nd column, 3rd paragraph:
 . "the systematic error" -> "the systematic uncertainty". An error is
   a mistake, and you shouldn't make mistakes. An uncertainty is
   something that you don't know.
ok

p 7, first sentence:
 . "0.32+-0.06" vs "0.74+-0.16": What am I supposed to conclude from
   this? That the PDG value is way-off, and this measurement is much
   more accurate than all older measurements combined? Or can I not
   directly compare these numbers?
please see the top of the page - the discrepancy is much reduced.
References:
 . [8]: Please mention the title of the NIM paper, unless it got
   submitted in the meantime...
 . [11]: as you know, this is PRL 94, 182001 (2005). By the way,
   congratulations on this one!
thank you!