SignalScope FFT Analyzer for Setting Azimuth

Has anyone used the FFT Analyzer function on the SignalScope program and tracks 2 & 3 on side 1 of the Analogue Productions Ultimate Test Record to set azimuth on their cartridge? Which measurement do you use for measuring the crosstalk? Is it Standard Deviation, Varience (Unbiased) or something else?
From my perspective I don't see why you would need to go to the frequency domain for azimuth adjustment.  The cross talk wave forms tend to be the same frequency as the main waveform and you are interested in minimizing amplitude anyway.  I have found that a better test record for Azimuth is "Azimuth Optimizer Test Record" LPT Azimuth 080108.  It has 2 sides and both sides alternate 1 kHz recordings on left then right channel.  Therefore you can see your cross talk across the entire record.  I use a analog oscilloscope Tektronix 475 and adjust azimuth to minimize, and equalize crosstalk.  For example:  When the right channel has a signal, you look at the amplitude of the waveform in the left channel (crosstalk).  When the left channel has a signal, you look at the amplitude of the waveform in the right channel.  Adjust azimuth to try and make those 2 amplitudes equal and minimal.  So as small as possible and as equal as possible.  
You will notice that amplitudes do change slightly as the needle moves across the entire record.  What I do is optimize as close as possible to the Lofgren tangent points. 
I use a now ancient Signet Cartridge Analyzer and the Shure test LP made to work with the Signet.  I agree with Captain Winters; I don't see why you'd need to use FFT, but I guess your question is related to the fact that you already own the computer program of which it is a part. Therefore, you wouldn't have to spend any more money. The Signet has a 'scope output, optional. I do own a 'scope but I've never connected it to the Signet. I just use the built-in db meter.
Thanks captain_winters and lewm.  I used the Sound Level Meter and Meter Bridge in the SignalScope app, along with the Oscilloppoi app to fine tune the azimuth.  I know you can't get it absolutely 100% even for crosstalk levels, but I was able to get it very, very very close.
Great success with Fozgometer
Pitbull, There is disagreement even among the extreme cognoscenti about whether the goal should be equal crosstalk in both channels or best possible crosstalk in EACH channel, while disregarding any difference between the two channels.  One thing is for sure, you usually cannot have it both ways; the setting that gives equal crosstalk, L into R and R into L, will generally not be the same as the setting that gives the best possible numbers, in terms of db. That's my personal experience, and it fits with what has been written. You can find essays on this subject at Vinyl Asylum and possibly on Vinyl Engine.