1: \begin{abstract}

2: We reevaluate the $B\to\rho\,l^+\nu_l$ decay width as a full $B\to\pi\pi\,l^+\nu_l$

3: four-particle decay, 

4:  in which 

5: the two final pions are produced via an intermediate $\rho$ meson. The decay width can be written

6: as a convolution of the $B\to\rho\,l^+\nu_l$ decay width, for

7: an off-shell $\rho$, with the $\rho\to\pi\pi$  line shape.

8: This allows to  fully incorporate the effects of the finite $\rho$

9:  meson width. As shown, consideration of the $\rho$ meson 

10: width effects  increase the $|V_{ub}|$ value by some 8\%, rendering it in

11: better agreement with the determination based in the

12:   $B\to\pi$ decay. We take the $q^2$ dependence of the $B\to \rho$

13: semileptonic form factors  from a dispersive Omn\`es representation. 

14: The Omn\`es subtraction constants

15: and the overall normalization parameter $|V_{ub}|$ are fitted to  light cone sum rules and 

16: lattice QCD  theoretical form-factor calculations, in the low and high $q^2$ regions

17: respectively, together to the  CLEO, BaBar and Belle  experimental

18: partial branching fraction distributions. The extracted value from

19: this global fit is

20: $|V_{ub}|=(3.12\pm 0.13)\times 10^{-3}$, in agreement  with the average

21: $B\to \pi$ exclusive value $|V_{ub}|=(3.23\pm 0.31)\times 10^{-3}$

22:  quoted by the Particle Data Group. The extracted value increases to 

23: $|V_{ub}|=(3.51\pm 0.16)\times 10^{-3}$ if only the most recent 

24: Belle Collaboration data

25: is used. This latter value is in agreement with different theoretical

26: determinations based in the $B\to\pi$ semileptonic decay and the values obtained

27: by the CKMfitter and UTfit groups. In any case a clear tension with the 

28: $|V_{ub}|$ value extracted from inclusive

29: semileptonic $b\to u$ decays still persists.

30: \end{abstract}

31: