Re: FFT test with few kbits

From: Eric Backus (eric_backus_at_alum.mit.edu)
Date: 02/18/04 

Date: Wed, 18 Feb 2004 12:34:33 -0800

"Ernst Lippe" <ernstl-at-planet-dot-nl@ignore.this> wrote in message
news:40334b3d$0$1225$ba620dc5@nova.planet.nl...
> It is a bit complicated to compute the exact entropy in
> each Fourier component F_k, because in general it depends
> on the factorisation of k and N. But when N is a prime
> the average entropy in the Fourier components is close
> to 0.75 N bits.

Now that's a result I've never seen before. Can you give a general
explanation for where this comes from? 

-- 
Eric Backus
R&D Design Engineer
Agilent Technologies, Inc.
425-356-6010 Tel