The effect of linguistic entropy on speech perception in noise in young and elderly listeners.

The rationale for a method to quantify the information content of linguistic stimuli, i.e., the linguistic entropy, is developed. The method is an adapted version of the letter-guessing procedure originally devised by Shannon [Bell Syst. Tech. J. 30, 50-64 (1951)]. It is applied to sentences included in a widely used test to measure speech-reception thresholds and originally selected to be approximately equally redundant. Results of a first experiment reveal that this method enables one to detect subtle differences between sentences and sentence lists with respect to linguistic entropy. Results of a second experiment show that (1) in young listeners and with the sentences employed, manipulating linguistic entropy can result in an effect on SRT of approximately 4 dB in terms of signal-to-noise ratio; (2) the range of this effect is approximately the same in elderly listeners.     

Psychometric functions for discrimination of two-component complex tones in listeners with normal hearing and listeners with hearing loss

This study compared the ability of 5 listeners with normal hearing and 12 listeners with moderate to moderately severe sensorineural hearing loss to discriminate complementary two-component complex tones (TCCTs). The TCCTs consist of two pure tone components (f1 and f2) which differ in frequency by delta f (Hz) and in level by delta L (dB). In one of the complementary tones, the level of the component f1 is greater than the level of component f2 by the increment delta L; in the other tone, the level of component f2 exceeds that of component f1 by delta L. Five stimulus conditions were included in this study: fc = 1000 Hz, delta L = 3 dB; fc = 1000 Hz, delta L = 1 dB; fc = 2000 Hz, delta L = 3 dB; fc = 2000 Hz, delta L = 1 dB; and fc = 4000 Hz, delta L = 3 dB. In listeners with normal hearing, discrimination of complementary TCCTs (with a fixed delta L and a variable delta f) is described by an inverted U-shaped psychometric function in which discrimination improves as delta f increases, is (nearly) perfect for a range of delta f's, and then decreases again as delta f increases. In contrast, group psychometric functions for listeners with hearing loss are shifted to the right such that above chance performance occurs at larger values of delta f than in listeners with normal hearing. Group psychometric functions for listeners with hearing loss do not show a decrease in performance at the largest values of delta f included in this study. Decreased TCCT discrimination is evident when listeners with hearing loss are compared to listeners with normal hearing at both equal SPLs and at equal sensation levels. In both groups of listeners, TCCT discrimination is significantly worse at high center frequencies. Results from normal-hearing listeners are generally consistent with a temporal model of TCCT discrimination. Listeners with hearing loss may have deficits in using phase locking in the TCCT discrimination task and so may rely more on place cues in TCCT discrimination.     

Individual differences in auditory discrimination of spectral shape and speech-identification performance among elderly listeners

Speech-understanding difficulties observed in elderly hearing-impaired listeners are predominantly errors in the recognition of consonants, particularly within consonants that share the same manner of articulation. Spectral shape is an important acoustic cue that serves to distinguish such consonants. The present study examined whether individual differences in speech understanding among elderly hearing-impaired listeners could be explained by individual differences in spectral-shape discrimination ability. This study included a group of 20 elderly hearing-impaired listeners, as well as a group of young normal-hearing adults for comparison purposes. All subjects were tested on speech-identification tasks, with natural and computer-synthesized speech stimuli, and on a series of spectral-shape discrimination tasks. As expected, the young normal-hearing adults performed better than the elderly listeners on many of the identification tasks and on all but two discrimination tasks. Regression analyses of the data from the elderly listeners revealed moderate predictive relationships between some of the spectral-shape discrimination thresholds and speech-identification performance. The results indicated that when all stimuli were at least minimally audible, some of the individual differences in the identification of natural and synthetic speech tokens by elderly hearing-impaired listeners were associated with corresponding differences in their spectral-shape discrimination abilities for similar sounds.     

Quantifying the intelligibility of speech in noise for non-native listeners

When listening to languages learned at a later age, speech intelligibility is generally lower than when listening to one's native language. The main purpose of this study is to quantify speech intelligibility in noise for specific populations of non-native listeners, only broadly addressing the underlying perceptual and linguistic processing. An easy method is sought to extend these quantitative findings to other listener populations. Dutch subjects listening to Germans and English speech, ranging from reasonable to excellent proficiency in these languages, were found to require a 1-7 dB better speech-to-noise ratio to obtain 50% sentence intelligibility than native listeners. Also, the psychometric function for sentence recognition in noise was found to be shallower for non-native than for native listeners (worst-case slope around the 50% point of 7.5%/dB, compared to 12.6%/dB for native listeners). Differences between native and non-native speech intelligibility are largely predicted by linguistic entropy estimates as derived from a letter guessing task. Less effective use of context effects (especially semantic redundancy) explains the reduced speech intelligibility for non-native listeners. While measuring speech intelligibility for many different populations of listeners (languages, linguistic experience) may be prohibitively time consuming, obtaining predictions of non-native intelligibility from linguistic entropy may help to extend the results of this study to other listener populations.     

NoSo and NoS pi detection as a function of masker bandwidth in normal-hearing and cochlear-impaired listeners

NoSo and NoS pi detection thresholds for a 500-Hz pure-tone signal were measured as a function of masking noise bandwidth in normal-hearing and cochlear hearing-impaired subjects. NoSo and NoS pi critical bands were derived from the bandlimited noise functions. A notched noise measure of the monaural critical band was also obtained for each ear. One hypothesis tested was that an asymmetrical monaural critical band would result in a relatively steep improvement of the NoS pi detection threshold as a function of decreasing masker bandwidth and would, therefore, be associated with a wider binaural critical band. This was hypothesized because the outputs of the left and right auditory filters would be more decorrelated the greater the interaural difference in the monaural critical band. However, as the noise bandwidth was narrowed, the decorrelation would lessen, resulting in a relatively steep improvement in NoS pi detection. Results indicated that the masking level difference (MLD) was smaller and that the monaural critical bands were generally wider in cochlear-impaired listeners. NoSo and NoS pi critical bands were somewhat larger in the cochlear hearing-impaired listeners having relatively wide monaural critical bands. There was a significant correlation between monaural critical band asymmetry and the NoS pi critical band; however, this correlation was insignificant when a control was employed for the critical band in the worse ear. Therefore, the present results did not support a strong association between monaural critical band asymmetry and the width of the NoS pi critical band.     

Contribution of consonant versus vowel information to sentence intelligibility for young normal-hearing and elderly hearing-impaired listeners

The purpose of this study was to examine the contribution of information provided by vowels versus consonants to sentence intelligibility in young normal-hearing (YNH) and typical elderly hearing-impaired (EHI) listeners. Sentences were presented in three conditions, unaltered or with either the vowels or the consonants replaced with speech shaped noise. Sentences from male and female talkers in the TIMIT database were selected. Baseline performance was established at a 70 dB SPL level using YNH listeners. Subsequently EHI and YNH participants listened at 95 dB SPL. Participants listened to each sentence twice and were asked to repeat the entire sentence after each presentation. Words were scored correct if identified exactly. Average performance for unaltered sentences was greater than 94%. Overall, EHI listeners performed more poorly than YNH listeners. However, vowel-only sentences were always significantly more intelligible than consonant-only sentences, usually by a ratio of 2:1 across groups. In contrast to written English or words spoken in isolation, these results demonstrated that for spoken sentences, vowels carry more information about sentence intelligibility than consonants for both young normal-hearing and elderly hearing-impaired listeners.     

Upward spread of masking, hearing loss, and speech recognition in young and elderly listeners

Upward spreading of masking, measured in terms of absolute masked threshold, is greater in hearing-impaired listeners than in listeners with normal hearing. The purpose of this study was to make further observations on upward-masked thresholds and speech recognition in noise in elderly listeners. Two age groups were used: One group consisted of listeners who were more than 60 years old, and the second group consisted of listeners who were less than 36 years old. Both groups had listeners with normal hearing as well as listeners with mild to moderate sensorineural loss. The masking paradigm consisted of a continuous low-pass-filtered (1000-Hz) noise, which was mixed with the output of a self-tracking, sweep-frequency Bekesy audiometer. Thresholds were measured in quiet and with maskers at 70 and 90 dB SPL. The upward-masked thresholds were similar for young and elderly hearing-impaired listeners. A few elderly listeners had lower upward-masked thresholds compared with the young control group; however, their on-frequency masked thresholds were nearly identical to the control group. A significant correlation was found between upward-masked thresholds and the Speech Perception in Noise (SPIN) test in elderly listeners.     

The relation between loudness and intensity difference limens for tones in quiet and noise backgrounds

Recent studies of the relation between loudness and intensity difference limens (DLs) suggest that, if two tones of the same frequency are equally loud, they will have equal relative DLs [R. S. Schlauch and C.C. Wier, J. Speech Hear. Res. 30, 13-20 (1987); J.J. Zwislocki and H.N. Jordan, J. Acoust. Soc. Am. 79, 772-780 (1986)]. To test this hypothesis, loudness matches and intensity DLs for a 1000-Hz pure tone in quiet and in a 40-dB SPL spectrum level broadband noise were obtained for four subjects with normal hearing. The DLs were obtained in both gated- and continuous-pedestal conditions. Contrary to previous reports, equally loud tones do not yield equal relative DLs at several midintensities in the gated condition and at many intensities in the continuous condition. While the equal-loudness, equal-relative-DL hypothesis is not supported by the data, the relation between loudness and intensity discrimination appears to be well described by a model reported by Houtsma et al. [J. Acoust. Soc. Am. 68, 807-813 (1980)].     

Detection and F0 discrimination of harmonic complex tones in the presence of competing tones or noise

Normal-hearing listeners' ability to "hear out" the pitch of a target harmonic complex tone (HCT) was tested with simultaneous HCT or noise maskers, all bandpass-filtered into the same spectral region (1200-3600 Hz). Target-to-masker ratios (TMRs) necessary to discriminate fixed fundamental-frequency (F0) differences were measured for target F0s between 100 and 400 Hz. At high F0s (400 Hz), asynchronous gating of masker and signal, presenting the masker in a different F0 range, and reducing the F0 rove of the masker, all resulted in improved performance. At the low F0s (100 Hz), none of these manipulations improved performance significantly. The findings are generally consistent with the idea that the ability to segregate sounds based on cues such as F0 differences and onset/offset asynchronies can be strongly limited by peripheral harmonic resolvability. However, some cases were observed where perceptual segregation appeared possible, even when no peripherally resolved harmonics were present in the mixture of target and masker. A final experiment, comparing TMRs necessary for detection and F0 discrimination, showed that F0 discrimination of the target was possible with noise maskers at only a few decibels above detection threshold, whereas similar performance with HCT maskers was only possible 15-25 dB above detection threshold.     

Factors affecting masking release for speech in modulated noise for normal-hearing and hearing-impaired listeners

The Speech Reception Threshold for sentences in stationary noise and in several amplitude-modulated noises was measured for 8 normal-hearing listeners, 29 sensorineural hearing-impaired listeners, and 16 normal-hearing listeners with simulated hearing loss. This approach makes it possible to determine whether the reduced benefit from masker modulations, as often observed for hearing-impaired listeners, is due to a loss of signal audibility, or due to suprathreshold deficits, such as reduced spectral and temporal resolution, which were measured in four separate psychophysical tasks. Results show that the reduced masking release can only partly be accounted for by reduced audibility, and that, when considering suprathreshold deficits, the normal effects associated with a raised presentation level should be taken into account. In this perspective, reduced spectral resolution does not appear to qualify as an actual suprathreshold deficit, while reduced temporal resolution does. Temporal resolution and age are shown to be the main factors governing masking release for speech in modulated noise, accounting for more than half of the intersubject variance. Their influence appears to be related to the processing of mainly the higher stimulus frequencies. Results based on calculations of the Speech Intelligibility Index in modulated noise confirm these conclusions.     

Rate coding model for discrimination of simple tones in the presence of noise

The predictions of a rate coding model for frequency and amplitude jnd's in the presence of noise are presented for a 1-kHz, 100-ms tone. The model for the neural response incorporates physiological data on dynamic range distribution and rate suppression. A central processor is assumed to estimate the tone frequency, or amplitude, from the tone-evoked rate increment profile. This central processor acts like an ideal detector with respect to the neural noise. The effects of the neural noise as well as the signal variability on the discrimination performance level are evaluated, and the signal variability is found to be significant. The combined effect of threshold distribution, rate suppression, and signal variability make the jnd's practically invariant with noise level, in accordance with published psychophysical data. The values of the frequency jnd at high signal-to-noise ratio, however, are borderline in their consistency with the data. A more obvious discrepancy exists between the model and the psychophysical data regarding the ratio of frequency to amplitude Weber fractions, which can be resolved only by modifying the model auditory filters to be five times sharper than those measured in cats.     

The detection of temporal gaps as a function of frequency region and absolute noise bandwidth.

Temporal gap detection was measured as a function of absolute signal bandwidth at a low-, a mid-, and a high-frequency region in six listeners with normal hearing sensitivity. Gap detection threshold decreased monotonically with increasing stimulus bandwidth at each of the three frequency regions. Given conditions of equivalent absolute bandwidth, gap detection thresholds were not significantly different for upper cutoff frequencies ranging from 600 to 4400 Hz. A second experiment investigated gap detection thresholds at two pressure-spectrum levels, conditions typically resulting in substantially different estimates of frequency selectivity. Estimates of frequency selectivity were collected at the two levels using a notched-noise masker technique. The gap threshold-signal bandwidth functions were almost identical at pressure-spectrum levels of 70 dB and 40 dB for the two subjects in experiment II, while estimates of frequency selectivity showed poorer frequency selectivity at the 70-dB level than at 40 dB. Data from both experiments indicated that gap detection in bandlimited noise was inversely related to signal bandwidth and that gap detection did not vary significantly with changes in signal frequency over the range of 600 to 4400 Hz. Over the range of frequencies investigated, the results indicated no clear relation between gap detection for noise stimuli and peripheral auditory filtering.     

Prediction of speech intelligibility in spatial noise and reverberation for normal-hearing and hearing-impaired listeners

Binaural speech intelligibility of individual listeners under realistic conditions was predicted using a model consisting of a gammatone filter bank, an independent equalization-cancellation (EC) process in each frequency band, a gammatone resynthesis, and the speech intelligibility index (SII). Hearing loss was simulated by adding uncorrelated masking noises (according to the pure-tone audiogram) to the ear channels. Speech intelligibility measurements were carried out with 8 normal-hearing and 15 hearing-impaired listeners, collecting speech reception threshold (SRT) data for three different room acoustic conditions (anechoic, office room, cafeteria hall) and eight directions of a single noise source (speech in front). Artificial EC processing errors derived from binaural masking level difference data using pure tones were incorporated into the model. Except for an adjustment of the SII-to-intelligibility mapping function, no model parameter was fitted to the SRT data of this study. The overall correlation coefficient between predicted and observed SRTs was 0.95. The dependence of the SRT of an individual listener on the noise direction and on room acoustics was predicted with a median correlation coefficient of 0.91. The effect of individual hearing impairment was predicted with a median correlation coefficient of 0.95. However, for mild hearing losses the release from masking was overestimated.     

The detection of differences in the cues to distance by elderly hearing-impaired listeners

This experiment measured the capability of hearing-impaired individuals to discriminate differences in the cues to the distance of spoken sentences. The stimuli were generated synthetically, using a room-image procedure to calculate the direct sound and first 74 reflections for a source placed in a 7 x 9 m room, and then presenting each of those sounds individually through a circular array of 24 loudspeakers. Seventy-seven listeners participated, aged 22-83 years and with hearing levels from -5 to 59 dB HL. In conditions where a substantial change in overall level due to the inverse-square law was available as a cue, the elderly hearing-impaired listeners did not perform any different from control groups. In other conditions where that cue was unavailable (so leaving the direct-to-reverberant relationship as a cue), either because the reverberant field dominated the direct sound or because the overall level had been artificially equalized, hearing-impaired listeners performed worse than controls. There were significant correlations with listeners' self-reported distance capabilities as measured by the "Speech, Spatial, and Qualities of Hearing" questionnaire [S. Gatehouse and W. Noble, Int. J. Audiol. 43, 85-99 (2004)]. The results demonstrate that hearing-impaired listeners show deficits in the ability to use some of the cues which signal auditory distance.     

The interpretation of speech reception threshold data in normal-hearing and hearing-impaired listeners: steady-state noise

Speech-in-noise-measurements are important in clinical practice and have been the subject of research for a long time. The results of these measurements are often described in terms of the speech reception threshold (SRT) and SNR loss. Using the basic concepts that underlie several models of speech recognition in steady-state noise, the present study shows that these measures are ill-defined, most importantly because the slope of the speech recognition functions for hearing-impaired listeners always decreases with hearing loss. This slope can be determined from the slope of the normal-hearing speech recognition function when the SRT for the hearing-impaired listener is known. The SII-function (i.e., the speech intelligibility index (SII) against SNR) is important and provides insights into many potential pitfalls when interpreting SRT data. Standardized SNR loss, sSNR loss, is introduced as a universal measure of hearing loss for speech in steady-state noise. Experimental data demonstrates that, unlike the SRT or SNR loss, sSNR loss is invariant to the target point chosen, the scoring method or the type of speech material.     

The temporal course of masking and the auditory filter shape

Recent experiments have shown that frequency selectivity measured in tone-on-tone simultaneous masking improves with increasing delay of a brief signal relative to the onset of a longer duration gated masker. To determine whether a similar improvement occurs for a notched-noise masker, threshold was measured for a 20-ms signal presented at the beginning, the temporal center, or the end of the 400-ms masker (simultaneous masking), or immediately following the masker (forward masking). The notch width was varied systematically and the notch was placed both symmetrically and asymmetrically about the 1-kHz signal frequency. Growth-of-masking functions were determined for each temporal condition, for a noise masker without a spectral notch. These functions were used to express the thresholds from the notched-noise experiment in terms of the level of a flat-spectrum noise which would produce the same threshold. In simultaneous masking the auditory filter shapes derived from the transformed data did not change significantly with signal delay, suggesting that the selectivity of the auditory filter does not develop over time. In forward masking the auditory filter shapes were sharper than those for simultaneous masking, particularly on the high-frequency side, which was attributed to suppression.     

Auditory filter shapes at low center frequencies in young and elderly hearing-impaired subjects.

Auditory filter shapes were measured for two groups of hearing-impaired subjects, young and elderly, matched for audiometric loss, for center frequencies (fc) of 100, 200, 400, and 800 Hz using a modified notched-noise method [B. R. Glasberg and B. C. J. Moore, Hear. Res. 47, 103-138 (1990)]. Two noise bands, each 0.4fc wide, were used; they were placed both symmetrically and asymmetrically about the signal frequency to allow the measurement of filter asymmetry. The overall noise level was either 77 or 87 dB SPL. Stimuli were delivered monaurally using Sennheiser HD424 earphones. Although auditory filters for the hearing-impaired subjects were generally broader than for normally hearing subjects [Moore et al., J. Acoust. Soc. Am. 87, 132-140 (1990)], some hearing-impaired subjects with mild losses had normal filters. The filters tended to broaden with increasing hearing loss. There were not any clear differences in filter characteristics between young and elderly hearing-impaired subjects. The signal-to-noise ratios at the outputs of the auditory filters required for threshold (K) tended to be lower than normal for the young hearing-impaired subjects, but were not significantly different from normal for the elderly hearing-impaired subjects. The lower K values for the young hearing-impaired subjects may occur because broadened auditory filters reduce the deleterious effects on signal detection of fluctuations in the noise.     

The importance of roof shape for road traffic noise shielding in the urban environment

The influence of roof shape on sound propagation in a densely build-up city centre is evaluated. Road traffic noise propagation from a street canyon to an adjacent, non-directly exposed building façade is numerically studied by the finite-difference time-domain method. A large number of common and less common roof shapes were analyzed in an idealized city canyon configuration. Roof shape can be responsible for differences in road traffic noise shielding exceeding 10 dBA, averaged over the shielded façade. Therefore, roof shape can be considered as an important means to limit sound pressure levels at a quiet side; various researchers indicated that the presence of silent zones and façades may limit city noise annoyance. With increasing vehicle speed, the choice of roof shape becomes more important. The roof top height was shown to be a bad predictor of shielding efficiency in the equal-building-volume approach followed in this study.     

Detection of tones in noise and the "severe departure" from Weber's law

Thresholds were measured for the detection of 20-ms sinusoids, with frequencies 500, 4000, or 6500 Hz, presented in bursts of bandpass noise of the same duration and centered around the signal frequency. A range of noise levels from 35 to 80 dB SPL was used. Noise at different center frequencies was equated in terms of the total noise power in an assumed auditory filter centered on the signal frequency. Thresholds were expressed as the signal levels, relative to these noise levels, necessary for subjects to achieve 71% correct. For 500-Hz signals, thresholds were about 5 dB regardless of noise level. For 6500-Hz signals, thresholds reached a maximum of 14 dB at intermediate noise levels of 55-65 dB SPL. For 4000-Hz signals, a maximum threshold of 10 dB was observed for noise levels of 45-55 dB SPL. When the bandpass noises were presented continuously, however, thresholds for 6500-Hz, 20-ms signals remained low (about 1 dB) and constant across level. These results are similar to those obtained for the intensity discrimination of brief tones in bandstop noise [R. P. Carlyon and B. C. J. Moore, J. Acoust. Soc. Am. 76, 1369-1376 (1984); R. P. Carlyon and B. C. J. Moore, J. Acoust. Soc. Am. 79, 453-460 (1986)].