On Mon, 19 Feb 2007 14:08:54 -0800, Richard Clark
wrote:

On Mon, 19 Feb 2007 15:16:29 -0600, "Henry Kolesnik"
wrote:

There's a good example of a 3D plot he
http://www.spectraplus.com/screenshots.html#3d
Has anyone used this software?

Hi Hank,

Yes, I have. It is exceedingly expensive, but you can probably get 30
days of free use if you have a one-time knock off project. Of course,
it may take 30 days to figure it out.

FFT analyzers are a dime a dozen, but few know how to use them
accurately - or are even aware of what can be done with them. Bob's
complaint, notwithstanding, a 1024 bin FFT employing the proper mixing
inputs can resolve any note on a guitar to within hundredths of a
cycle. I am sure this is of no interest to you, however.

Yes, you can easily resolve to 100ths of a cycle, but not from
a native FFT. (See my other response.) You can do it from
the waveform by measuring the time between cycles. That's
one of the methods I'm working on. The other is to do it from
the FFT via peak interpolation. The waveform method will
be very accurate with simple static waveforms, but I expect it
will have trouble with the 2nd harmonic of a plucked string,
which is not an integer multiple so "rolls" through the waveform.
The FFT interpolator will not have the same raw accuracy, and must
have absolutely separated spectral peaks since it assumes the
one it's interpolating is the only one. I'm not hopeful it can be
made good enough to tune a guitar, at least not in the bottom
octave where the harmonics land on adjacent spectral lines.

Best regards,


Bob Masta

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Scope, Spectrum, Spectrogram, Signal Generator
Science with your sound card!