Bimodal perception of spectrum compressed speech
Larry D. Paarmann, Michael K. Wynne, Carmen M. Pugh, Jun Bi · The Journal of the Acoustical Society of America · 1994
In this study, ten normal-hearing adult listeners were asked to identify CUNY nonsense syllables presented auditorily and bimodally (audition and vision) via video tape in two conditions: Low-pass filtered and spectrum compressed by means of nonlinear sampling in the frequency domain [L. D. Paarmann, M. D. Guiher, and M. K. Wynne, J. Comp. Users Speech and Hear. 7, 257–281 (1991)]. The data collected were analyzed using a 2×2 ANOVA to determine if significant differences exist across the two main effects and if the interaction between the variables is significant. Although this study does not directly address the effectiveness of spectrum compression of the speech signal to improve speech understanding for hearing-impaired listeners, the results do strongly support the hypothesis that spectrum compressed speech is more intelligible to normal-hearing listeners than low-pass filtered speech with the same bandwidth, and it suggests that, especially in a bimodal listening situation, hearing impaired listeners’ speech intelligibility could be improved by such a processing method.