Regardless of which model we invert (magnitude or phase), we must
acknowledge that there are limitations on the frequency ranges in
which either model is valid. These ranges are functions of the
accuracy desired and of the true value of
. We already
have noted in previous sections that as
becomes very
small, the current model is not valid. Thus, none of the
constraints on
for backsolving should cause alarm in this
context. We present more detail below.