Sequential streaming due to manipulation of interaural time differences.

The effect of apparent spatial location on sequential streaming was investigated by manipulating interaural time differences (ITDs). The degree of obligatory stream segregation was inferred indirectly from the threshold for detecting a rhythmic irregularity in an otherwise isochronous sequence of interleaved "A" and "B" tones. Stimuli were bandpass-filtered harmonic complexes with a 100-Hz fundamental. The A and B tones had equal but opposite ITDs of 0, 0.25, 0.5, 1, or 2 ms and had the same or different passbands. The passband ranges were 1250-2500 Hz and 1768-3536 Hz in experiment 1, and 353-707 Hz and 500-1000 Hz in experiment 2. In both experiments, increases in ITD led to increases in threshold, mainly when the passbands of A and B were the same. The effects were largest for ITDs above 0.5 ms, for which rhythmic irregularities in the timing of the A or B tones alone may have disrupted performance. It is concluded that the differences in apparent spatial location produced by ITD have only weak effects on obligatory streaming.

Evaluation of a frequency transposition algorithm using wearable hearing aids.

Transposition of acoustic information from higher to lower frequencies may help people with severe or profound high-frequency hearing loss, especially when a 'dead region' is present. Previously, we (Robinson et al, 2007 ) evaluated the benefit of an FFT-based transposition algorithm in a laboratory study. Although results were promising, we hypothesized that further training and exposure would be needed to gain the full benefit. This was tested here by implementing the algorithm in wearable digital hearing aids. Five subjects with high-frequency dead regions used the aids for five weeks. Performance on the transposing and control conditions was compared objectively using speech tests (vowel-consonant-vowel, 's' detection, and speech in noise) and subjectively using questionnaires. Overall, the results showed no benefit with the transposition even after experience. Subjective preference was generally for the control condition.

The effects of hearing loss on growth-of-masking functions for sinusoidal and complex-tone maskers with differing phase spectra.

Growth-of-masking (GOM) functions in forward masking (0-ms masker-signal delay) were measured for normally hearing (NH) and hearing-impaired (HI) listeners using as maskers complex tones (harmonics 1-40, 100-Hz fundamental frequency) with components starting in cosine or random phase, and on-frequency sinusoids. The signal was a 20-ms sinusoid, usually with a frequency of 1 or 2 kHz. It is argued that differences in the slopes of the GOM functions for the random- and cosine-phase maskers provide a measure of the strength of compression in the cochlea. For the NH listeners and some of the HI listeners, the slopes were significantly greater for the random- than for the cosine-phase maskers, and for these listeners the slopes for the complex-tone maskers were less than for the sinusoidal maskers. For the remaining HI listeners, the slopes of the GOM functions were similar for all masker types. It is argued that these listeners had almost complete loss of cochlear compression. The GOM functions for the sinusoidal maskers had slopes between 0.45 and 0.78 and were typically in the range 0.6-0.7. The finding of slopes below one for listeners in whom cochlear compression was probably absent is not consistent with linear-integrator models of forward masking.

Temporal masking curves for hearing-impaired listeners.

The decay of forward masking was investigated for three subjects with moderate sensorineural hearing loss. For such subjects, compression on the basilar membrane (BM) is thought to be largely absent, enabling one to determine the decay of masking without the influence of compression. Temporal masking curves (TMCs), plots of the masker level at threshold against delay between masker offset and signal onset, were measured for delays of 0, 15, 30, 45, 60, and 75 ms, for signal frequencies, fs, of 500, 1000, 2000, 4000, and 6000 Hz. Masker frequencies were 0.5, 0.8, 1.0, 1.15, and 1.3 times fs. Most of the TMCs were well fitted with single-segment straight lines, which, except for high masker levels, were roughly parallel for each fs, supporting the belief that BM compression was largely absent in these subjects. However, the slopes of the TMCs were greater for fs = 500 and 1000 Hz than for higher frequencies, which may indicate that the decay of forward masking is not the same for all signal frequencies. The results suggest that it may not be valid to infer BM compression at low signal frequencies by using a reference TMC for a high fs.

The effect on speech intelligibility of varying compression time constants in a digital hearing aid.

The identification of nonsense syllables in quiet and in three types of background (babble, cafeteria and single female speaker) was measured using four hearing aid compression algorithms differing in attack and release time constants, and using linear amplification. The speech level was always 65 dB SPL. The compression algorithms, which were implemented in a Phonak Claro ITE hearing aid, were: (1) 'very fast'--the attack time was 8 ms and the release time was 32 ms, for all 20 channels; (2) 'slow-fast'--the attack and release times decreased from 500 ms for low frequencies to about lOOms for high frequencies; (3) 'fast-slow'-the attack and release times increased from about 50ms for low frequencies to 500 ms for high frequencies; and (4) 'slow + fast'-a very slow-acting gain control signal was combined with a fast-acting gain control signal, for each channel in a 10-channel system. Acoustical stimuli were presented monaurally via a circumaural headphone mounted over the hearing aid. The linear condition did not use the Claro aid; instead, the signal was digitally filtered to implement the Cambridge formula prior to delivery via the earphone. Five subjects with moderate sensorineural hearing loss were tested in a counter-balanced order across conditions. In quiet, performance was best for linear amplification and worst for the slow + fast algorithm. In the presence of background sounds, the highest scores were obtained with the linear-gain Cambridge formula implemented via headphones; a supplementary experiment suggested that this was due to the greater high-frequency gain resulting from the use of this formula. No significant differences were found between scores for the different compression algorithms. We conclude that the intelligibility of speech at a fixed level, presented in background sounds, is not markedly affected by rather substantial variations of the time constants in a multichannel compression system.

Effects of masker component phase on the forward masking produced by complex tones in normally hearing and hearing-impaired subjects.

For normally hearing subjects, harmonic complex tones that give "peaky" waveforms on the basilar membrane (Schroeder-positive phase, sine phase or cosine phase) lead to less forward masking than complex tones that give less peaky waveforms (Schroeder-negative phase or random phase), but have the same power spectrum. This difference has been attributed mainly to the combined effects of peripheral compression and suppression, both of which depend on the operation of the active mechanism in the cochlea. If this explanation is correct, the phase effect should be reduced or absent for subjects with moderate cochlear hearing loss. We measured growth-of-masking functions for forward maskers containing the first 40 harmonics of a 100-Hz fundamental, with components added either in cosine phase or random phase, using both normally hearing subjects and subjects with moderate cochlear hearing loss. The signal frequency was 1 or 2 kHz. For the normally hearing subjects, the mean slopes of the growth-of-masking functions at 1 and 2 kHz, respectively, were 0.53 and 0.44 for the random-phase masker and 0.31 and 0.26 for the cosine-phase masker. For high masker levels, the former produced considerably more masking than the latter. The phase effect was smaller for the hearing-impaired than for the normally hearing subjects, which is consistent with the idea that it is partly caused by peripheral compression and suppression. However, three of the five hearing-impaired subjects showed a significant effect of masker phase for at least one signal frequency. In one case, this occurred when the hearing loss at the signal frequency was 65 dB. The slopes of the growth-of-masking functions were consistently less than one for the hearing-impaired subjects. Further testing suggested that the efferent system was not involved in producing the phase effect.

Auditory streaming based on temporal structure in hearing-impaired listeners.

The influence of temporal cues on sequential stream segregation was investigated using five elderly hearing-impaired listeners. In experiment 1, an alternating pattern of A and B tones was used. Each tone was a harmonic complex with a 100-Hz fundamental, with one of three passbands (1250-2500, 1768-3636, or 2500-5000 Hz) and one of three component-phase relationships (cosine, alternating, or random). The complexes had an overall level of 96 dB SPL. The detection of a change in relative timing of the A and B tones was measured in a two-interval-forced-choice paradigm. The sequence in one interval remained isochronous while the sequence in the other started isochronously but became increasingly irregular with the addition of a cumulative delay between the A and B tones. Component phase relationship and passband difference both had significant effects on the minimum detectable delay, indicating that temporal structure produced obligatory stream segregation. In experiment 2, subjects continuously reported whether tones presented in a 30-s ABA-ABA- sequence were perceived as segregated or integrated. Differences in component phase between A and B significantly increased perceived segregation, but passband did not. In conclusion, stream segregation due to differences in temporal structure is robust in elderly subjects with cochlear hearing loss and comparable to that found previously in young normally hearing subjects.

Sequential streaming and effective level differences due to phase-spectrum manipulations.

Roberts et al. [J. Acoust. Soc. Am. 112, 2074-2085 (2002)] demonstrated that sequential stream segregation occurs with stimuli that differ only in phase spectrum. We investigated if this was partly due to differences in effective excitation level. Stimuli were harmonic complexes with a 100 Hz fundamental, 1250-2500 Hz passband, and cosine, alternating, or random component phase. In experiment 1, the complex tones were used as forward maskers of 20-ms probe tones at 1000, 1250, 1650, 2050, 2500, and 3000 Hz. While there was no significant difference in the masking produced by the cosine- and alternating-phase stimuli, the random-phase stimulus produced significantly greater masking, equivalent to a difference in overall effective excitation level of 12.6 dB. Experiments 2 and 3 used the asynchrony detection and subjective streaming tasks of Roberts et al. Successive stimuli had identical phase, but differed in level by 0, 1, 3, 5, 10, or 15 dB. Stream segregation increased once the level difference reached 5 dB. While some of the stream segregation observed by Roberts et al. may have been due to a difference in effective excitation level, this does not account for the stream segregation between cosine- and alternating-phase stimuli.