Matlab. Filter FFT data to octave-band and fractional octave-band.

More information about generation of central frequencies on my repository: getANSIfrequencies

[Oct,fc,fl,fu] = FFT2octave(f,fftdata,b,varargin)

The first two inputs are the FFT, f is the frequency array and fftdata is the fft values.

The third input is the bandwidth:

• b=1 One-octave band.
• b=3 Third-octave band.
• b=2 Half-octave band.
• b=3/2 2/3-octave band.
• b=24 1/24-octave band.

Optional inputs

The fourth input indicates whether the FFT data is linear '' or logarithmic 'dB'.

The fifth input only affects when the value of b is 1 or 3. It indicates whether you want the normalized or calculated frequencies. By default they are calculated (false).

% To get in one-octave frequencies
[Oct,fc,fl,fu] = FFT2octave(f,fftdata,1)

% To get in one-octave frequencies if the data is in dB
[Oct,fc,fl,fu] = FFT2octave(f,fftdata,1,'dB')

% To get in one-octave frequencies normalized if the data is in dB
[Oct,fc,fl,fu] = FFT2octave(f,fftdata,1,'dB',true)

% To get in third-octave frequencies
[Oct,fc,fl,fu] = FFT2octave(f,fftdata,3)