Hi all, I found the reason for the bump at high generated Mx values. It is present only for the vub-generic MC. In the events with two semileptonic events, both the X systems were used to calculate the X mass. I put a fix in the code. If a b->ulnu and a b->clnu event are present the b->ulnu is taken. If two semilep. decays of the same kind (vub==2 || vcb ==2) are present, it takes the first one. I also verified the mxhadgen shapes for vub-breco and vub-generic and they look the same. Daniele