Just a point of clarification ... As Rogaway has pointed out several times on this list (in that quickly receding past), it is important to avoid any predictable correlation of the IV with the data. Hilarie