Zoom FFT Analysis

The zoom FFT analysis tools in the NI Sound and Vibration Measurement Suite acquire multiple blocks of data, modulates, and downsamples to obtain a lower sampling frequency. The block size is decoupled from the achievable frequency resolution because the zoom FFT analysis tools accumulate the decimated data until you acquire the required number of points. Because the transform operates on a decimated set of data, you need to compute only a relatively small spectrum. The data is accumulated, so do not think of the acquisition time as the time required to acquire one block of samples. Instead, the acquisition time is the time required to accumulate the required set of decimated samples.

The zoom FFT tools complete the following steps to process the sampled data:

1. Modulate the acquired data to center the analysis band at 0 frequency.
2. Filter the modulated data in the time domain to isolate the analysis band and prevent aliasing when the data is resampled at a lower sampling frequency.
3. Decimate the filtered data to reduce the effective sampling frequency.
4. Accumulate the decimated data until sufficient samples are available to compute the spectrum.
5. Use the Discrete Zak Transform to efficiently compute the desired spectral lines.
6. Demodulate, or shift, the computed spectrum.

The following equation describes the frequency resolution of the zoom spectrum:

The exact frequency resolution depends upon parameters that are coerced for optimum computation. The exact frequency resolution is returned in the spectrum.

Use zoom FFT when your application requires decoupling of the frequency resolution and the input block size.