“…The argument quickly shifted from introducing µ in order to avoid the potential catastrophic errors in the original equations of Section 4.3.1 to considering an optimal value. For several decades, based originally on the works of Kress [395][396][397][398] and the further works of Amini [399], there has been a strong recommendation, with supporting research, that µ = i k is reasonably close to the 'optimal' choice, as least for the simple boundaries, like spheres. Obviously i k is unsuitable for low wavenumbers, as the parameter would be large, violating the condition |µ| ∞, and the second equation would dominate and a cap on its value is recommended, for example,…”