Elekta Neuromag NM21283A, Руководство пользователя

Страница: 55 / 85

Basic measurements 8
NM21283A-A 49
8.7 FFT measurement
The job “FFT” is for measuring power spectra of the channels. This can
be used to examine the external disturbances and noise characteristics in
more detail compared to the “Noise” measurement. The spectrum estima-
tion uses a Hanning window and non-overlapping averaging.
The parameters affecting the FFT measurements are shown in table below.
The Time buffer size defines the length of the time trace that is actually
measured. From the beginning of each time trace, FFTsize samples are
taken and transformed into Spectral noise density per unit band (1/Hz).
The length of the time buffer should be at least the length of the Fourier
transform. In order to get better statistics, several spectra can be averaged.
The number of averages is controlled by parameter FFTaverages. It is pos-
sible to apply SSP noise reduction to the data before calculating the FFT
in order to diminish the contribution of external disturbances.
The last four parameters in Table 8.3 define two bands, one for numerical
estimation of the low frequency noise and another for estimating the white
noise level. The noise power in these bands can be shown as a bar dia-
gram.
Parameter Type Description
TimeBufSize integer Time buffer length. [samples]
FFTsize integer FFT transformation length. [samples]
FFThpf float High pass filter setting. [Hz]
FFTlpf float Low pass filter setting. [Hz]
FFTsfreq float Sampling frequency. [Hz]
FFTaverages integer Number of averages.
FFTssp 0 or 1 Is SSP used?
FFTlowStart float Start of low freq. estimation range. [Hz]
FFTlowEnd float End of low freq. estimation range. [Hz]
FFTwhiteStart float Start of white noise estimation range.[Hz]
FFTwhiteEnd float End of white noise estimation range. [Hz]
Table 8.3 Parameters affecting FFT measurements