From another viewpoint, this ML procedure can be interpreted as one of a clustering formulation under a harmonic constraint between Gaussian mixture components where spectral density function is assumed to be a statistical distribution of micro-energies along frequency axis. As we regard as cluster centroid of cluster , the a posteriori probability in equation (6) as a membership degree of each micro-energy to every cluster and the log-likelihood as a distance function (square Euclidean distance, for instance) between centroid and a micro-energy, thus the function in equation (3) turns out to be the general objective function for fuzzy clustering. We call this concept ``Harmonic Clustering.''