|Vub | from the Spectrum of B → πeν Flavor Physics and CP Violation Conference, Bled, 2007 1

dΓ dq2 (B̄ → πlν̄l) = GF |Vub| 192πmB λ(q)|f+(q)| , (2) where λ(q) = (mB + m 2 π − q) − 4mBmπ is the phase-space factor. The calculation of f+ has been the subject of numerous papers; the current state-ofthe-art methods are unquenched lattice simulations [1, 2] and QCD sum rules on the light-cone (LCSRs) [3, 4]. A particular challenge for any theoretical calculation is the prediction of the shape of f+(q ) for all physical q: LCSRs work best for small q; lattice calculations, on the other hand, are to date most reliable for large q. Hence, until very recently, the prediction of the B → πlν decay rate necessarily involved an extrapolation of the form factor, either to large or to small q. If, on the other hand, the q spectrum were known from experiment, the shape of f+ could be constrained, allowing an extension of the LCSR and lattice predictions beyond their region of validity. A first study of the impact of the measurement, in 2005, of the q spectrum in 5 bins in q by the BaBar collaboration [5] on the shape of f+ was presented in Ref. [6]. The situation has improved dramatically in 2006 with the publication of high-precision data of the q spectrum [7], with 12 bins in q and full statistical and systematic error correlation matrices. These data allow one to fit the form factor to various parametrisations and determine the value of |Vub|f+(0) [8]. As it turns out, the results from all but the simplest parametrisation agree up to tiny differences which suggests that the resulting value of |Vub|f+(0) is truly model-independent. In these proceedings we report the results for |Vub|f+(0) and |Vub|, obtained in Ref. [7, 8]. There are four parametrisations of f+ which are frequently used in the literature. All of them include the essential feature that f+ has a pole at q = mB∗ ; as B (1) is a narrow resonance with mB∗ = 5.325GeV < mB+mπ, it is expected to have a distinctive impact on the form factor. The parametrisations are: