Ciao Riccardo. I have few questions about your study. 1) Why is delta_e of the B-reco correlated with the M_nu resolution ? 2) You do not seem to have statistical problems. Why not using higher purity sample ? 3) The structure of the m_nu distribution seems to be more complicated than a double gaussian: you maybe need to describe it better befor you can draw conclusions on the resolution. Cheers Franco