Hi all,
we've just finished re-evaluating the sys error coming from mb and
reweighting using the same Lambda bar and lambda1 central values as
before and using an error on Lambdabar of 122 MeV and  on lamda1 of
0.105 (CLEO moments  analysis).

We've run 400 trials extracting mb and a (using athe information on the
correlation of Lambdabar and lambda1: -0.8 -as from CLEO-) and looking at
the spread of the BRBR results.

The sys error that comes out is 17% (consistent with the 20% that we
quote before for an error on mb of 150MeV).

We'll not show the results for B0 and B+ separately including
systematic error : this work can be done in the PRL timescale
(uncorrelated error needs to be evaluated carefully).
But we could show the  results for B0 and B+ just with a
statistical error as in table 1 of the  conference paper in order to
provide a  test on the consistency of our measurement.

You can also find below few comments on points 3 and 4.

Cheers,
the authors.

On Fri, 28 Feb 2003, Stephane Willocq wrote:

> *** Discussion title: Review of Vub with B->Xulnu decays
> Email replies to [log in to unmask] must include:
>   In-Reply-To: <"/pubboard54/38/1"@babar-hn.slac.stanford.edu>
>   Subject: ...change this to be about your reply.
>
> Hi,
>
>   I read the Vub Breco physics note and have a few suggestions below.
> It is a great accomplishment and a very interesting result. Congratulations!
>
> 1) The title should probably specify that a branching fraction
> is measured
>
> * Sec.1:
>
> 2) In the presentation of the merits of this analysis, a lot of
> good points are made. I am not so sure about the point that the
> analysis avoids systematics related to determining lepton
> efficiencies. It is clear that you avoid this problem, but you
> enventually use BaBar's own measurement of the inclusive semileptonic
> BF which must figure this out. Therefore, your final Vub result
> is affected by this uncertainty so I wouldn't emphasize this point
> too strongly.
>
> 3) Concerning the perturbative corrections and uncertainties, the
> rate is said to have a first 1/mb**2 term of order of 5% but the
> 1/mb**3 term for Vub has an uncertainty as large as that, see Eq.(1).
> Does that mean the 1/mb**3 correction is big or that its uncertainty
> is huge?
>

We are using the pole mass in the generator and not a running mass. We
use the shape  function formalism at leading order  with the pole mass
to  extrapolate to  the full  BF.  For  this extrapolation  we  use an
uncertainty of 120  MeV  on LambdaBar which  is dominating the
theoretical  error  we quote.  The  error here  is  the  error on  the
measurement of CLEO.


> 4) In Eq.(1), mb is taken at 1 GeV. How does that relate to the
> value used in the generator? In other words, is the use of Eq.(1)
> consistent with the rest of the analysis?
>

The conversion of the BF to |Vub| is done at a higher order (1/m_b^2),
and here  a running mass  is used. This  is not inconsistent  with the
above. The error on the mass here  is 90 MeV, as described in PDG. The
error on  |Vub| due to this  uncertainty and the  neglection of higher
orders is  taken from PDG  as well, in  the literature the  errors are
smaller.  The increased  error  in PDG  is  probably a  result of  the
discussions at the CKM workshop of last year.



> 5) One picky little detail: I would think the basic notation for
> the decay mode would be B -> X l nu, rather than B -> X l nu-bar
> since the "B" meson contains a b-bar quark.
>
> * Sec.3:
>
> 6) It would be useful to give the maximum number of pions and kaons
> allowed by the Breco algorithm.
>
> 7) Same picky detail about using D(*)- and D0-bar if "B" (and B+ or B0)
> is used everywhere.
>
> 8) I would remove reference to "a priori" purity and just keep
> the word purity to avoid unnecessary confusion.
>
> 9) It would be interesting to quote the fraction of Breco B+ and B0
> with the correct charge assignment.
>
> 10) In the list of cuts, are you missing a minimum cut on Mmiss**2?
>
> * Sec. 4:
>
> 11) BGsl is not clearly defined, please add.
>
> 12) The function used to describe the Breco signal peak is not
> mentioned.
>
> 13) Is it correct to write "... results in Nsl meas = 32210 +/- 233"?
> that may help to make the meaning of all variables in Eq.(2) clear.
>
> 14) Along the same lines, the fraction of background is given. Should
> you add that BGsl = (6.8%) Nsl meas? (if that's what it is)
>
> 15) The next paragraph (lines 185-191) is not clear, could you try
> rephrasing it?
>
> 16) Same for the paragraph between lines 204 and 207. I suspect
> you mean "efficiency for selection B -> Xu l nu" rather than "efficiency
> for detecting...", right? Also, the word "tagged" refers to the lepton
> requirement?
>
> 17) I do not see a separate result for B0 and B+ with systematics. Do you intend to present full results for the publication?
>
> 18) In Fig.3, one cannot see any "background" component, is that
> right?
>
> 19) Since there is mention of the depleted sample in a number of places,
> should we include an MX distribution for that sample?
>
> * Sec. 6:
>
> 20) The BaBar semileptonic BF measurement seems to have some zeroes
> missing, the errors should be +/- 0.0018(stat) +/- 0.0030(syst).
>
> That's all for now.
>
> Cheers,                 Stephane.
>
>