MATLAB Hanning & Hamming Windows in MATLAB: Why the Discrepancy?

[truva]

Hi all

As far as I know Hanning window is defined as the following:
.5+.5*cos(2*pi*n/N) and n is between [-(N-1)/2,(N-1)/2] here N is odd number.

MATLAB functions "hanning" with symetric or periodic options or "hann" do not match with the results of the above formula. There is always a slight annoying difference !

Do you know the problem?

(Same thing for Hamming window)

 

[Pythagorean]

I used both in my DSP course without a problem. Maybe this will help:

http://www.mathworks.com/help/toolbox/signal/ref/hann.html

 

[truva]

As I understand the book is wrong (Notes On Digital Signal Processing by C. Britton Rorabaugh). If I replace the N by (N-1) in the denumerator of the formulation which is 5+.5*cos(2*pi*n/(N-1)), the results coincide exactly with that of MATLAB.

Thank you for the link.