zeros and poles of minimum phase filter

I used Hilbert transform to construct a minimum phase filter. The magnitude and phase response in frequency domain are correct. However, when I do iverse Fourier transform h = real(ifft(H)) to get it in time domain and use zplane(h,1) to draw the zero-pole plot, there are zeros out side the unit circle.Why? Any issues with my procedure? Thanks.

Answer 1

[Hao Yin]

Any answers? Thanks