Inverse FFT

[Heavenly]

I am trying to do a fft and a ifft on a earthquake record but I am having trouble with the ifft part of the spreadsheet. I used Spec:=fft(v) and then Spec2:=ifft(Spec) but I am not sure if that is correct. I have attached my MathCAD spreadsheet. I am also wondering when you use the fft vs. the FFT function you get the same frequency content but different magnitudes because of the how the vector is normalized, will you get the same magnitude in the time domain based on either approach after the taking the inverse?

[Heavenly]

I figured out why the graph wasnt working. It was because I had the wrong Y axis, it needed to be ifft(Spec).

[TomGutman]

fft and ifft should be exact inverses (up to rounding errors).

Your sheet fails on any system other than your own, as nobody else has the data file that it relies on. When posting a sheet it is normally best to copy the data into the sheet to be posted (usually as a data table). That allows anybody to immediately see and review the calculations.
__________________
� � � � Tom Gutman

[ptc-1368288]

On 2/5/2009 6:42:49 PM, Heavenly130 wrote:
...
>I am also
>wondering when you use the fft
>vs. the FFT function you get
>the same frequency content but
>different magnitudes because
>of the how the vector is
>normalized, will you get the
>same magnitude in the time
>domain based on either
>approach after the taking the
>inverse?
______________________________

Scaling is a matter of convention. For a "scaling" to be representative of "real amplitude, you have to calibrate on a dummy signal like a step function. You can do easier by scaling 0..1 your vector.

jmG

[TheodoreM.Bones]

Ready made Fourier Transform postings are available. You have to provide the calibration data yourself. Here is a blue sjy data set with complicated functional values and many points, so it is slow running. Never mind trying to vectorize anything. You don't need it here.

[ptc-1368288]

"Ready made Fourier Transform postings are available"
_________________________

You can't be more right, Theodore !

There are 1000's of postings under that name.
The chance of successful help is within < 4 hrs in this collab. This time laps has elapsed, therefore gone with the wind.

Sunny day, cold, typical of february up there.

Salut, Theodore.

Jean