Inverse Fourier Transform

[cobayatron]

Hi,

I have a basic doubt when using the IFT.vi

I'm analysing some data I do the FFT of some signals (accelerometers) and then I pick a particular peak and perform the IFT to obtain a time decay signal. I've tried both IFT Real and IFT Complex and they do produce different results but I'm not sure which one I should use. Anyone could clarify when to use one and the other?

Thanks!

[Henrik_Volkers]

I can't open your vi.. however ..

If you use the basic FFT and IFFT vis, you should make yourself familar how they proceed with data.

I asume that you have a real input of n points (x of a x,t time signal, in comparison to a IQ complex signal ) . FFT give an output of n complex values ,where usually only the first half is used, since the other half is the 'negative mirror'. (In the picture, the input hat 2000 values (sweep) , the first 1000 values of the FFT represents the spectrum. (not scaled ... !  I didn#T rename the x scale .. schould be n for number of value ;))

If you want to cut out some bins (and get back a real value signal) , you have to set the rest to zero and choose 1D real for the IFFT.

A good test: If you don't filter and use no (rectangular) window, you should get back the same signal.

Correct scaling to samplerate and frequency bins is your homework now... to get familar with the FFT and IFFT 😄

Hint: The LabVIEW help provide some details.. 

[cobayatron]

Thanks Henrik Volkers

I think my question is more basic though, the difference between the 1D Real/1D Complex option in de IFT.vi is whether the incoming signal is Real or Complex? Everytime I perform a FFT on accelerometer's data that I want to reverse that will be Real. So when do you actually use 1D Complex option. Thnaks again

[Henrik_Volkers · Accepted solution]

RTFMH 😄  (read the fantastic help 😉 )

The input of the IFFT is always complex  and you choose the 1D complex option, if you want an complex IFFT output.

[cobayatron]

Thank you very much!