Hi, in my attempt to adopt the track and neutral selection from the vub measurement for the mass moments analysis I realized that there has been a significant broadening of the missing mass distribution taken place between old vub selection (middle of last year) and the new one. This broadening can be seen by comparing missing mass plots in BAD 347 (v4 page 45) and BAD 540 (v8 page 57). Using the b->u enh. distribution one finds a broadening of roughly 50% off the new missing mass distribution with respect to the old one (it should also be noted that there is virtually no difference concerning the quality of the data MC comparison for this plots?!) This large increase in MM2 will certainly have a severe impact on the mass resolution of Mx and will probably not be acceptable for the new direct measurement approach (slops of the calibration curves will get significantly worse due to the loss of information caused by the new vub track and neutral selection). Does someone know the reason for this broadening ... it almost looks like that there are cuts at work which throw away useful information in the event ... or do I misunderstand something here? Thanks for your help, Oliver