I want to ask what's the difference between computing IFFT from a hermitian-symmetrical complex matrix, when
1st matrix (lets call it X) is created by making a matrix [0, x, 0, fliplr(conj(x)] and then simply -> ifft(X)
2nd matrix (also X) is just computed by adding 'symmetric' flag to ifft -> ifft(X,'symmetric')
The 1st output is longer (twice + 2 zeroes) and I have no idea how to process the signal further.
Which method performs better?