Filter size check is wrong

[paulmueller]

The filter size check in filter.py is done with integer values of the image size. But it should be with Fourier space coordinates.

[PinkShnack]

I don't think the filter size check is wrong in this case (this issue is v. old so perhaps it's out of date anyway).

Because users do not use this function directly (they usually use it through OAH), I would only change the docstring slightly to be a bit more clear.

"""
    filter_size: float
        Size of the filter in Fourier space.

        In this function, ``filter_size`` is interpreted in the same
        *normalized frequency units* as returned by :func:`numpy.fft.fftfreq`
        (and typically used together with :func:`numpy.fft.fftshift`), i.e.
        values are in the range ``[-0.5, 0.5)`` along each axis.

        Implications for each ``filter_name``:

        - For the radial/square filters (e.g. "disk", "gauss", "square", ...),
          values up to about ``0.5`` correspond to radii within the Fourier
          domain. Values larger than ``0.5`` (e.g. ``1``) effectively yield an
          all-pass filter (no filtering).
        - The "tukey" filter is different and that scales with
          the array size. For "tukey", ``filter_size=1`` can exceed the array bounds.

"""