Start with more precision, your end result will be more precise. Think of it this way: You're going to do a series of steps on a number, all using floating-point math of very-high detail. Would you rather start with your initial sample only accurate to one decimal point (4.5) or four (4.4511). With a series of steps, the final result could be very different in the two cases. Whether the difference is really audible or not is a totally different question!

[This message has been edited by bigmac (edited June 11, 2002).]