Re: Large FFT -- NOT POWER of 2 ;{

 Not a answer , but you may know that efficiencies of the FFT
 come from symmetries based on the prime factors of the number
 of points . 44100 has the interesting set of factors :

   2 2 3 3 5 5 7 7

 Know any reason for this ?

--
On Thu, 01 Jan 2004 17:19:45 -0600, Richard Owlett wrote:
>=A0I wish to do 44100 point FFT's on some 16 bit data.
>=A0[ Yepp, looking at 1 sec samples of CD audio data { spoken=
 speech
>=A0actually) ]
>
>=A0I think integer versions would be fine.
>=A0I'm only interested in magnitude squared ( ie power ).
>
>=A0I know there are many sources for FFT for the number of samples=
 equal
>=A0to a power of 2. BUT, some other aspects of my project will be=
 much
>=A0simpler if I analyze 1 second long samples.
>
>=A0Can anyone point me to existing code?
>=A0Speed is not a MAJOR issue. Would like it to be faster than=
 Scilab ;]
--=A0
=A0Bob Armstrong -- http://CoSy.com -- 212-285-1864
Computing Environment : =A0http://CoSy.com/CoSy/
A WTC vision : http://CoSy.com/CoSy/ConicAllConnect/
Liberty : http://CoSy.com/Liberty.htm
----------------------------------------------------------------------
sftalk_at_forth.com      The SwiftForth programming discussion email list
To unsubscribe, send subject "unsubscribe" to sftalk-request_at_forth.com
For list command help, send subject "help" to sftalk-request_at_forth.com
Message archives are located at http://www.forth.com/archive/sftalk
----------------------------------------------------------------------
This list is a forum for SwiftForth users.  For product support and bug
reports, please send email to support_at_forth.com
----------------------------------------------------------------------