Something not covered above was the effect of using the infinite
limits approximation directly, while still using the rectangular
window. Herein, we rigorously show why this is not a valid
approach. We begin our pursuit of this doomed idea by stating the
assumption:
This approximation is a much stronger one than the large limits
approximation, and allows a greatly simplified expression for the
real and imaginary parts of
. Here, we consider why this
is not valid for the rectangle window case. Though our result is
discouraging, we recall that the infinite limits approximation may
in fact be simulated by using a different windowing function, as
has been covered earlier.
To explore the validity of the infinite limits approximation for
the rectangle window case, we will compare this approximation to
that made above (the large limits approximation). Since we know
the large limits approximation is very accurate (to within 2.3%
for
) we will consider the error introduced by the infinite
limits approximation relative to the large limits
approximation, and judge its validity accordingly.
Presently, we compare the large limits approximation
![]() |
![]() |
(148) | |
![]() |
![]() |
(149) |
and the infinite limits approximation