Notes on - Windowed Sinc Filters
          ----------------------------------

                                         Original Rev.  21-Dec-2021
                                          Current Rev.   2-Jan-2022


The seed for the kernel of a Windowed Sinc Filter is the  sin(x)/x
sinc function.  To make things practical only a finite length of
the sinc function can be used in a filter.  The filters described
here will always use a sinc function of length  M+1  samples where
M is always even.  This resulting M+1 samples is symmetric about
the center lobe with M/2 samples on either side of the center lobe.
For use in the filters the whole set of M+1 samples is shifted
to the right so that the array index runs from  0 through M,
i.e.  M+1 samples.

If you just cutoff the sinc function that will cause serious
ripple in the filter output.  The way to control this problem is
to "window" the sinc function.  The standard windowing function
to use is the  Blackman  function which is:

  W(n)  =  0.42  -  0.5 cos( 2 pi n / M)  +  0.08 cos( 4 pi n / M)


The shift of the windowed sinc filter kernel index by  M/2  places
to the right  causes a shift in the filter output by the same
amount to the right  (typically later in time).