Dear all,
I only get this as a digest so many apologies if I am out of sync,
with any discusions.
Terry Steen many thanks for your suggestion, but I'm guilty of not
making myself too clear I would require 2 possibly 3 decimal places
of precision.  Categorising a 16 bit number into 11/12 steps would be
too course.
Any help would be most appreciated, even an algorithm on how to take
natural log of binary number in binary form.
Cheers
Andy Farrar