Path: utzoo!attcan!uunet!decwrl!sdd.hp.com!uakari.primate.wisc.edu!crdgw1!ge-dab!sunwhere!harrison
From: harrison@sunwhere.DAB.GE.COM (Gregory Harrison)
Newsgroups: comp.dsp
Subject: Re: Practical DSP problem
Message-ID: <6145@ge-dab.GE.COM>
Date: 9 Oct 90 15:57:32 GMT
References: <3575@syma.sussex.ac.uk> <9360012@hplsla.HP.COM>
Sender: news@ge-dab.GE.COM
Reply-To: harrison@sunwhere.UUCP (Gregory Harrison)
Organization: GE Simulation & Control Systems Dept., Daytona Beach, FL
Lines: 18

In article <9360012@hplsla.HP.COM> bobc@hplsla.HP.COM (Bob Cutler) writes:
>>
>>I have an application where I need phase, rather than amplitude, but
>>the problem is similar. In that case, I compute:
>>
>>  for i = 0 through a bunch
>>   I = sum(sinwt*sample[i])
>>   Q = sum(coswt*sample[i])

I believe that this is called the Geotzel method, and should be documented	in good DSP texts.  It is much more costly to compute an entire spectrum in 
this manner, but I would guess that it works just fine with single line 
spectra.  I wonder if you could use a window on it.  I wonder if you could
process a wigner kernel in this manner.  I believe that using this method
you create a tuned filter.  I wonder how to handle non-stationary cases.

Score:  3 wonders, 3 tidbits of info.

Greg