Characteristics of hoarse, rough and normal voices: A comparative acoustic spectrographic analysis

Introduction: Vocal quality can be assessed by auditory perceptual and acoustic methods. The terms used to describe vocal quality are not standardized. However, two of them deserve special attention regarding the parameters that distinguish roughness and hoarseness, both usually referred in the literature as hoarseness. Aim: To evaluate comparatively and to define basic differences among hoarse, rough and normal voices using acoustic spectrography. Study Design: Prospective clinical randomized. Material and Method: Six hoarse, six rough and six normal voices sorted by auditory perceptual evaluation, and analyzed by spectrography under five parameters: fundamental frequency (f 0), extension of harmonics, definition of harmonics, presence of noise between the harmonics and above them. Results: Hoarse voices presented lower f0, whereas rough voices showed higher f0 in comparison to normal voices. Considering the spectrogram from 0 to 5.4 kHz, the mean value of the uppermost harmonic frequency reached 4,555 Hz for normal voices, 2,125 Hz for rough voices and 1,147 Hz for hoarse voices. Harmonic tracing in hoarse voices showed worse definition and greater amount of noise among and above the harmonics. Conclusions: There are striking spectrographic differences among hoarse, rough and normal voices. Harmonics are present in greater amount and show better definition in normal voices, and decrease in amount in rough and hoarse voices, respectively. Inversely, noise appears in great amount and in a diffuse way in hoarse voices and in smaller amount in rough and normal voices.