The effect of compression ratio on speech intelligibility and quality

Abstract

The present investigation examined the effect of different compression ratios (compression ratios of 2:1, 3:1, 8:1 and wide-dynamic-range compression) on speech intelligibility and quality in compression-limiting systems. Speech intelligibility and quality were evaluated for sentences in four-talker babble. Sentences were presented at a signal-to-noise ratio of 7.5 dB at six different input levels, and at two of these levels (80 and 100 dB SPL) four different signal-to-noise ratios. Speech intelligibility was evaluated in terms of percent correct words and the quality of the conditions was rated on a scale of 0 to 100%. Across subjects, across input levels, and across signal-to-noise ratios, different compression ratios did not differ in terms of speech intelligibility and quality. Individual subject analysis revealed that for three of the subjects 8:1 compression ratio conditions gave the best speech intelligibility and quality results. One of the subjects achieved his best speech intelligibility performance with the wide-dynamic-range compression condition. For four of the subjects, different compression ratios did not effect speech intelligibility and quality differently. Overall, results indicated that there were considerable variations in performance with different compression ratios among individuals with similar hearing sensitivity. [Work partly supported by Northwestern Univ. Doctoral Student Research Grant.]

The Effects of Digital Signal Processing Features on Children's Speech Recognition and Loudness Perception

“…Recall from Figure 4 that the DNR system of the hearing instruments in the current study demonstrated much more gain reduction with pink noise than with the Auditec four-talker babble used as the competing signal for the sentences in noise loudness perception task. This noise source was chosen in the interest of representing commonly encountered real-world listening situations (Fikret-Pasa, 1993). This choice is consistent with a goal of the current study, which was to identify a fitting strategy for use in the nonquiet environments experienced by children in their daily lives as described by Crukley et al (2011).…”

A Clinical Perspective on Cochlear Dead Regions: Intelligibility of Speech and Subjective Hearing Aid Benefit

“…To control for presentation level and configuration of hearing loss, an AI was calculated for each ear (Pavlovic, 1989). The AI calculations were based on subject thresholds, the speech spectrum of the QuickSIN speaker (Fikret-Pasa, 1993), the presentation level of the QuickSIN, and band importance functions for sentences (DePaolis et al, 1996). AI calculations should also take into account the level of the noise, but due to the variable SNR for the QuickSIN sentences, noise levels were not entered into the calculations.…”

Performance of Directional Microphones for Hearing Aids: Real-World versus Simulation

“…A sequence of sentence blocks modeled after the SIN (Speech in Noise) Test (Fikret-Pasa, 1993;Killion and Villchur, 1993 ) was recorded in various noisy real-world environments: a crowded street party (90-95 dBA), two restaurants (70-80 dBA and 60-65 dBA), a museum party (80-85 dBA), and a classroom party simulation (80-85 dBA). Because the methodology of the experimental design was not standardized, it is difficult to compare the study's results with past and future investigations of other directional hearing aids.…”