🚀 go-pugleaf

I asked about implementations/algorithms or papers/books 
disscussing Integer FFT's. Here is a summary:

_________________________________________________________________________
=========================================================================


I believe some integer FFT`s were found in Embedded Systems
Programming within the last year.



-- 
Mike Koopman                             phone: +1-814-269-2637
Concurrent Technologies Corporation    telefax: +1-814-269-2666
1450 Scalp Avenue                         ICBM:  40'15"N-78'50"W
Johnstown, PA  15904  USA               e-mail:  koopman@server1.ctc.com



_________________________________________________________________________
=========================================================================

> 	I am interested in integer FFT implementations/algorithms
> or papers/books disscussing them.  Do you have any pointers to this
> information?
> 
> 	Thanks in Advance,
> 
> 	mark shattuck (mds@phy.duke.edu)

I found very little in the literature about integer FFT, so I ended up
looking at the usual ones and then I applied standard integer methods to it.

I had numerous opportunity to write fixed point math, and usually had no
language support for it (most languages support integer and float, some
(like PL/I etc.) will handle fixed point) so I ended developing a personal
approach.

The basic idea is to be aware of the two limitations : magnitude and resolution,
Then one trades one for the other in such a way that the final results are within the desired range.

I have posted the code to comp.dsp a couple of days ago. If you cannot get it
then let me know and I will email to you. The fft.tar.Z.uue file is about 60K,
rather small.

-- 
Regards
	Eyal Lebedinsky		eyal@ise.canberra.edu.au


[I have this code now also, if someone wants it.]
-- 
	mark shattuck (mds@phy.duke.edu)