I have a bit complex time signal. When trying to plot its spectrum, I get some weird result. Denoting the time signal as x and time vector as t, whether I use the definition of the cepstrum by typing this formulae xcepstrum = real(ifft(log(abs(fft(x))))); or I use Matlab function cceps I get the same result. The problem is that the result doesn't make any sense. Here is a plot of the signal.

Signal in different domains

The values that appear as 0 in the cepstrum aren't really zero. Whren plotting using the semilogy function, I get this plot. enter image description here



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.