MATLAB: How to find the bandwidth of a signal

sampling ratewvd

i have a audio signal ,I want to know the bandwidth because its specgram has artefacts i think it may be due to the sample rate.

Best Answer

For all but the simplest cases the term bandwidth is ambiguous. You probably want to look at the long-time Fourier transform instead of the short-timer (i.e., spectrogram). Assuming your audio signal is single channel and called x

plot(20*log10(abs(fft(x))))

From there you might be able to define the lower and upper frequency bounds of your signal.

Related Solutions

MATLAB: How to find the spectral three dominant frequency components at 100Hz,500Hz&1000Hz, from a time domain signal generated by applying 1000 clicks(auditory stimulus) sampled at 10kHz which was band passed between 100Hz and 3000Hz during recording stag

Since you only need to know the spectrum at 3 particular frequencies, I would say do a numerical approx of the Fourier transform integral at those frequencies, e.g.,

 f=100; %100 Hz
 sampinterval=1/10e3; %one over 10 kHz
 fourierCoeffs=exp(-2*pi*f*j*(-T:sampinterval:T)); %Fourier basis coefficients
 trapz(yoursignal.*fourierCoeffs));

MATLAB: Do I see frequencies which do not exist in the original signal in the output of the FFT function

The behavior observed is due to a phenomenon called aliasing. Aliasing refers to the distortion that occurs when a continuous time signal has frequencies larger than half of the sampling rate. The process of aliasing describes the phenomenon in which components of the signal at high frequencies are mistaken for components at lower frequencies.

To avoid aliasing, ensure that the sampling rate is high enough to avoid any spectral overlap, or use an anti-aliasing filter. In the code, for example, consider "third_frequency" to be 600. Since the sampling frequency, Fs, is 1000, you will notice a component at 1000 – 600 = 400 Hz. If "third_frequency" is 800, you will see a component at 1000 – 800 = 200 Hz.

Best Answer

Related Solutions

MATLAB: How to find the spectral three dominant frequency components at 100Hz,500Hz&1000Hz, from a time domain signal generated by applying 1000 clicks(auditory stimulus) sampled at 10kHz which was band passed between 100Hz and 3000Hz during recording stag

MATLAB: Do I see frequencies which do not exist in the original signal in the output of the FFT function

Related Question