The Effect of Spatially Separated Sound Sources on Speech Intelligibility

Abstract

A series of five experiments was conducted to investigate the effects of spatial separation of speakers on the intelligibility of spondaic and PB words in noise and the identification of synthetic sentences in noise and competing message. Conditions in which the spatial location of the speakers produced interaural time differences ranked highest in intelligibility. The rank order of other conditions was dependent on the S/N ratio at the monaural near ear. Separations of only 10° between the speech and noise sources resulted in measurable changes in intelligibility. The binaural intelligibility scores were enhanced substantially over the monaural near ear results during conditions where an interaural time difference was present. This result was observed more effectively when spondaic words or sentences were used rather than PB words. The implications of this result were related to the interaural time difference and the frequency range of the critical information in the primary message. Although the initial experiments were facilitated by recording through an artificial head, almost identical results were obtained in the final experiment when subjects were tested in the sound field.

Spatial unmasking and attention related to the cocktail party problem

“…(2) Masking release using artificial head Hirsh (1950)[27] 2 syllables broad band noise 2.5 dB threshold of intelligibility 90˚separation 2 syllables broad band noise 4.5 dB threshold of intelligibility 90˚separation (2 mike vs monaural) Dirks and Wilson (1969) [31] spondee words broad band noise 7 dB intelligibility 90˚separation (2 mike vs monaural) spondee words broad band noise 9 dB intelligibility 90˚separation (dummy head) Bronkhorst and Plomp (1988) [30] sentence (female talker) speech spectrum noise 9.4 dB SRT 90˚separation (KEMAR) Koehnke and Besing (1996) [16] monosyllabic words speech spectrum noise 13 dB 50% intelligibility 90˚separation (KEMAR) Yost et al (1996) [28] key words (male talker) words (male talker) 30% identification binaural vs monaural Peissig and Kollmeier (1997) [35] speech (male talker) speech spectrum noise or multi-talker speech 10 dB SRT 105˚separation (HRTF) Hawley et al (1999) [29] key words (male talker) competing sentence 30% error rate 30˚separation(KEMAR) Drullman and Bronkhorst (2000) [36] words and sentence (male talker) words (male and female talker) 40∼50% intelligibility binaural vs monaural Freyman et al (2001) [18] key words (female talker) sentence(female talker) 35% word score 60˚separation (KEMAR) Shinn-Cunningham et al (2001) [37] key words(male talker) speech spectrum noise 6 dB SRT 45˚separation (HRTF) (3) Masking release due to interaural time difference Schubert (1956) [44] words noise 3 dB intelligibility ∆t = 470 µs Schubert and Schultz (1962) [41] words sentences 4 dB intelligibility ∆t = 0.5 ms Levitt and Rabiner (1967) [9] words broad band noise 3 dB 50% intelligibility ∆t = 0.5 ∼ 10 ms speech broad band noise 13 dB threshold ∆t = 0.5 ∼ 10 ms speech broad band noise 3 dB intelligibility ∆t = 0.5 ∼ 10 ms Carhart et al (1967) [12] spondee words white noise 3 dB intelligibility ∆t = 0.8 ms spondee words broad band noise 6 dB threshold ∆t = 0.8 ms spondee words broad band noise 4 dB intelligibility ∆t = 0.8 ms Carhart et al (1968) [45] spondee words broad band noise 1.5 dB intelligibility ∆t = 0.8 ms monosyllable broad...…”

“…Hirsh [27] obtained a 4.5 dB advantage for the binaural hearing condition over the monaural one when the directional difference between the target and interference sources was 90 using an artificial head. Dirks and Wilson [31] compared the hearing conditions with 90 separation of source direction and obtained about a 7 dB improvement for the binaural condition. They also reported that the advantage due to 90 separation in binaural hearing was 6 to 9 dB.…”

How Does Learning to Read Affect Speech Perception?

“…Because each discharge of the stimulator produces a click, participants always wore an ear plug in their left ear (the one nearest the stimulating coil) and auditory stimuli were delivered directly to their right ear via an earphone (Sennheiser, CX-300). This facilitated hearing the stimuli by reducing the TMS noise and by introducing a spatial separation between the noise and speech signal (Dirks and Wilson, 1969;Zurek, 1993).…”

Effect of Dual Sensory Loss on Auditory Localization: Implications for Intervention

“…The improvement in speech intelligibility is because of two factors, head shadow or head diffraction effects and binaural auditory processing. [62][63][64] The head shadow effect is purely acoustic and arises when the speech and interference or noise sources are in different locations. Binaural auditory processing is the ability of the binaural system to make use of the interaural difference cues (IID and ITD) in the received sounds, which enhances the separation of signal from interference (noise).…”