Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!uunet!microsoft!brianw
From: brianw@microsoft.UUCP (Brian Willoughby)
Newsgroups: comp.dsp
Subject: Sliding FFT == freq. spectra in real time
Message-ID: <7879@microsoft.UUCP>
Date: 28 Sep 89 23:26:40 GMT
Reply-To: brianw@microsoft.UUCP (Brian Willoughby)
Distribution: na
Organization: Microsoft Corp., Redmond WA
Lines: 19

I looked up the EDN article that I mentioned in an earlier post:
Electronic Design News, Sept. 29, 1988, page 161, "Sliding FFT computes
frequency spectra in real time", by Tom Springer of Pinson Associates
Inc.

I don't have anything to try this out with (yet), but it appears to offer
more than a tenfold increase in speed when computing FFT's.  It updates
at the sample rate, provided that the system can compute N*4 multiply and
N*6 add/subtract operations N times per sample, where N is the number of
points.  I would be curious as to how fast the 56000 could execute this
algorithm.  BTW, it uses a pre-computed exponential phase shift table as
a constant factor, which eliminates the need for trig functions during
the real-time computations.

Brian Willoughby
UUCP:           ...!{tikal, sun, uunet, elwood}!microsoft!brianw
InterNet:       microsoft!brianw@uunet.UU.NET
  or:           microsoft!brianw@Sun.COM
Bitnet          brianw@microsoft.UUCP