Comments on "Design of inherently stable two-dimensional recursive filter from one-dimensional filters"

In a recent paper, a method for designing a restricted class of stable two-dimensional recursive filters from one-dimensional filters was given. The class of filters considered are those with a second-order (in each variable) transfer function with some additional constraints. It was claimed that the method given can lead to all possible stable filters of the type considered there. The purpose of this correspondence is to prove through an example that the method given there cannot give all stable transfer functions even within the very restricted class of filters mentioned.