Identificación de algunos parámetros espectrales que determinanla calidad de la voz

The main objective of this paper is to contribute in the process of classification of voice quality. This contribution is achieved by means of the analysis of a set of proposed parameters (PMR, SNR, SNRL, SNRM, SNRH), obtained at the frequency domain. These parameters were obtained from the frequency domain after transforming the signal using the DFT. As a result of this work twenty four voices sampling of pathological and good voices were analysed, several parameters were proposed and the results were statistically tested, which yielded three statistically significant parameters PMR, SNRM and SNRH. These parameters have not been found in the reports of investigations performed by others researchers in spite of the great number of articles in this field.