'jumps' in phase response

[scdoro]

Hi all,

Function [H,W] is used to compute the samples of a DTFT of a finite length signal.
I have created its M-file and run the signal xn for n=128.

My questions are

1) Why are there 'jumps' in the phase response graph?
2) Where is the jumps in this phase response graph.
3) Is the jump the dip in the level from +ve value to -ve value?


P.S: Two attachments: Phase response of signal xn.
Full question


thanks

regards

 

[MSGHP]

it can be proved that where magnetude is zero the phase jumpes (pi)