The contour test of loudness perception

Comparing loudness normalization (IHAFF) with speech intelligibility maximization (NAL-NL1) when implemented in a two-channel device

OBJECTIVE

At least two rationales are available for fitting wide dynamic range compression hearing aids. The goal of one rationale is to normalize loudness, and the goal of the second rationale is to maximize speech intelligibility. Neither rationale has been validated against other fitting rationales for the range of input levels common to the hearing aid user in the real world. The goal of the study was to compare the two rationales when implemented in a 2-channel compression hearing aid.

DESIGN

Loudness normalization and speech intelligibility maximization were implemented using the Independent Hearing Aid Fitting Forum (IHAFF) and the National Acoustic Laboratories' Nonlinear (NAL-NL1) prescriptive formulas. Twenty-four subjects (eight for each of three groups of mild flat, moderate/severe flat, and steeply sloping hearing loss) participated in the study. Each subject completed an initial laboratory test, field test, and final laboratory test. The laboratory test consisted of a paired-comparison judgment for each prescriptive formula using four stimuli under both quiet and noisy listening conditions and a sentence recognition test using Bamford-Kowal-Bench sentences. In the field test, subjects evaluated the two rationales in individually selected everyday listening conditions for 4 wk. A digital simulation of the fitting rationales implemented in two channels was used for laboratory testing and a digital 2-memory, 2-channel device was used for field testing. Subjects adjusted the overall gain of each response to their preferred listening level in both the laboratory and in the field.

RESULTS

Data collected in the laboratory before and after the field test showed no indication of significant learning or acclimatization effects. For each stimulus presented in the paired-comparison test more subjects preferred NAL-NL1 than preferred IHAFF. For the sentence recognition test, subjects performed significantly better with NAL-NL1 than IHAFF in a low-frequency weighted background noise. Sixteen out of 22 subjects who completed the field test reported a preference for the NAL-NL1 response. The remaining six subjects preferred IHAFF. The paired-comparison test and field test revealed that while the achieved root-mean-square (rms) difference between fittings for an input level of 65 dB SPL was small, the preference for either rationale was small. As the rms difference between fittings increased, the score in favor of NAL-NL1 increased. The correlation between the differences in satisfaction score obtained in the field test and the rms differences between the responses fitted was statistically significant.

CONCLUSIONS

When the two fitting rationales prescribed substantially different responses for a 65 dB SPL input and these differences were achieved in the fitting, then the subjects preferred NAL-NL1. Even when the difference between fittings was small, the subjects preferred and performed better with NAL-NL1 when listening in a low-frequency weighted background noise.

The selection and validation of output sound pressure level in multichannel hearing aids

OBJECTIVE

To validate the Australian National Acoustic Laboratories' (NAL) procedure for prescribing output sound pressure level (OSPL) for multichannel hearing aids (Dillon & Storey, 1998)

DESIGN

The NAL OSPL prescriptive procedure for multichannel hearing aids was used to calculate Predicted OSPL, Predicted Maximum Acceptable OSPL and Predicted Minimum Acceptable OSPL for 20 subjects with sensorineural hearing loss fitted with a 2-channel linear hearing aid. Subjects rated the speech clarity and quality of average (65 dBA) and loud (80 dBA) speech, in quiet and in noise, with the hearing aid set to a number of OSPL settings. These data were used to evaluate the validity of the Predicted OSPL. Frequency-specific loudness discomfort levels (LDLs) were measured to determine whether use of measured LDLs would improve the accuracy of the prediction.

RESULTS

The Predicted Minimum Acceptable OSPL was in good agreement with the measured minimum acceptable OSPL for both the low- and high-frequency channels. The Predicted Maximum Acceptable OSPL was in good agreement with the measured maximum acceptable OSPL for the low-frequency channel, but was only a fair predictor for the high-frequency channel. The use of measured LDLs rather than predicted LDLs did little to improve the accuracy of the fitting. A direct comparison between the NAL single-channel and multichannel prescribed OSPL settings showed that most listeners rated speech clarity higher for the multichannel settings.

CONCLUSIONS

In two channel hearing aids, the NAL Predicted Minimum Acceptable OSPL and Predicted Maximum Acceptable OSPL are reasonable predictors of minimum and maximum OSPL levels measured using sound clarity and quality ratings. The results of this study support the use of the NAL prescriptive formula for setting OSPL in multichannel hearing aids. Such settings should be verified by having the listener rate the loudness of an intense speech signal. If tolerance problems are evident, the OSPL in the high-frequency channel(s) should be reduced first.

Verifying loudness perception after hearing aid fitting

During the verification phase of a hearing aid fitting, clinicians often want to assess the extent to which loudness perceptions for amplified sounds are similar to those typical of normal hearers. This type of verification calls for a criterion for "normal" loudness perception of sounds presented in a sound field. This research sought to answer several questions about the parameters of a valid "normal" criterion for a verification procedure using the Contour test of loudness perception. Loudness data were obtained from 30 listeners with normal hearing. Results indicated that a loudness growth function obtained with earphone listening is not an appropriate normative reference for hearing aid fitting verification. Instead, the normative data should be based on sound field listening. Results also indicated that the same normative function could be used to assess both unilateral and bilateral fittings. Further, it is likely that the same normative function can be used for most frequency responses that are likely to be used in feasible fittings. Finally, it was found that a previously published normative function obtained using an automated test procedure was not faithfully replicated using a carefully executed fully manual test procedure. We concluded that, until a replicable normative function is established, practitioners will need to generate their own local norms to perform postfitting verification of loudness normalization.

The preferred number of channels (one, two, or four) in NAL-NL1 prescribed wide dynamic range compression (WDRC) devices

OBJECTIVE

The recently introduced NAL-NL1 rationale for fitting WDRC devices prescribes a relatively high compression threshold and prescribes compression ratios lower than those prescribed by loudness normalization rationales. The aim of this study was to investigate whether the compression characteristic prescribed by NAL-NL1 is most effective in a single-channel scheme or in a multi-channel scheme.

DESIGN

Twenty-four subjects with flat or steeply sloping hearing loss participated in the study. One, two, and four channels were implemented digitally in the laboratory and evaluated on the basis of a paired-comparison test and a speech recognition test. The test stimuli consisted of speech and noise presented at average input levels, and speech and noise alternating every 3 sec among different input levels. The single-channel and 2-channel NAL-NL1 prescriptions were also evaluated in individually selected everyday situations in the field using a digital 2-memory device.

RESULTS

The three compression schemes produced no significant difference in speech recognition scores. Most subjects showed no preference for either scheme in the paired-comparison test. Those who did mainly selected the single-channel scheme. These preferences can be explained on the basis on audibility and quality. In the field all subjects with a steeply sloping loss, but one, preferred the 2-channel scheme. Among the subjects with a flat loss more preferred the single-channel scheme than preferred the 2-channel scheme. Statistical analyses showed that those who preferred the 2-channel scheme were fitted with significantly greater differences in the compression ratio in the high frequencies, and those who preferred the single-channel scheme were fitted with significantly greater differences in the high-frequency gain for a 65 dB input.

CONCLUSIONS

Multi-channel compression prescribed according to NAL-NL1 in up to four channels showed no adverse effects on speech recognition relative to a single-channel scheme. The paired-comparison test showed a small, but explainable preference for the single-channel scheme. The field test revealed a preference for the 2-channel scheme by subjects with steeply sloping loss. When using the NAL-NL1 rationale it is recommended to use at

An examination of several characteristics that affect the prediction of OSPL90 in hearing aids

OBJECTIVE

Investigators at the National Acoustic Laboratories have provided a theoretical derivation and experimental validation of a formula for setting the maximum output of hearing aids (Dillon & Storey, 1998; Storey, Dillon, Yeend, & Wigney, 1998). Given that measurement of discomfort levels for setting maximum output can be both time-consuming and of questionable reliability, the use of a prescriptive formula warrants consideration. In this article, an extensive data base was considered and issues of normal hearing, clinical protocol, age and gender were investigated in an effort to further determine optimal maximum output settings.

DESIGN

Data were gathered from five previous investigations, for a total of 433 subjects (total ears = 710). Threshold of discomfort (TD) measures were obtained using one of two adaptations of the Ascending Method of Limits, one with category anchors and one without.

RESULTS

Subjects with normal hearing had significantly lower TDs than subjects with hearing loss. A different regression line for measured TDs as a function of hearing level was noted for subjects whose hearing threshold levels fell between 20 and 60 dB HL and those with thresholds above 60 dB HL. When all effects (hearing level, method, age and gender) were considered in a single predictive model for the two threshold groups, only method and threshold were significant predictors of TD. However, for the subjects with thresholds between 20 and 60 dB HL, less than 4% of the variance in TD measures could be accounted for by those factors. For subjects with threshold above 60 dB HL, 22% of the variance in TD measures could be accounted for by those variables.

CONCLUSIONS

For both groups of subjects (20 to 60 dB HL and above 60 dB HL) methodology and hearing thresholds are significant predictors of discomfort levels. Age and gender are not. Given the small variance accounted for by any of these factors, measures of discomfort using standardized methodologies seem warranted.

The impact of head angle on monaural and binaural performance with directional and omnidirectional hearing aids

OBJECTIVE

To evaluate the impact of head turn and monaural and binaural fittings on the sentence reception thresholds of hearing-impaired listeners wearing directional and omnidirectional hearing aids.

DESIGN

Sentence reception thresholds were measured for 20 listeners fit monaurally and binaurally with behind-the-ear hearing aids set in both directional and omnidirectional modes. All listeners exhibited symmetrical, sloping, sensorineural hearing loss. The aided performance across these four fittings was evaluated for three different head and body angles. The three angles reflected body turns of 0 degrees, 15 degrees, and 30 degrees as measured relative to the primary sound source, with 0 degrees denoting the listener directly facing the sound source. Listeners were instructed to keep their heads in a fixed horizontal position and turn their heads and bodies to face visual targets at the three test angles. Sentences from the Hearing in Noise Test presented with a background of five, spatially separated, uncorrelated samples of cafeteria noise served as test material. All testing was performed in a moderately reverberant (Rt = 631 msec) "living room" environment.

RESULTS

Participants generally performed significantly better when fit with directional versus omnidirectional hearing aids, and when fit binaurally versus monaurally across test conditions. The measured "binaural advantage" was reduced with increasing head angle. Participants performed significantly better with a 30 degree head angle than when directly facing the primary speaker. This "head turn advantage" was most prominent for monaural (versus binaural) conditions. Binaural and head turn advantages were not significantly different across directional and omnidirectional modes.

CONCLUSIONS

These data provide additional support for the use of directional hearing aids and binaural amplification to improve speech intelligibility in noisy environments. The magnitude of these advantages was similar to that reported in previous investigations. The data also showed that hearing aid wearers achieved significantly better speech intelligibility in noise by turning their heads and bodies to a position in which they were not directly facing the sound source. This head turn advantage was in good agreement with the increase in Directivity Index with head turn and reflected the fact that hearing aids are generally most sensitive to sounds arriving from angles other than directly in front of the hearing aid wearer. Although these data suggest that many monaural hearing aid wearers may significantly improve speech intelligibility in noise through the use of head turn, the interaction between this advantage and the potential loss of visual cues with head turn is unknown.

Relative loudness perception of low and high frequency sounds in the open and occluded ear

A comparison of published equal loudness contours indicates that different shapes are obtained at a comfortable level when the measurements are done in an occluded ear than when they are done in an open ear, even though all measurements are expressed as dB SPL at the eardrum. This paper presents the result from a loudness balancing test which confirms this observation. Eleven normal-hearing listeners balanced the level of a 500- and a 3000-Hz octave band babble-noise to the level of a 1500-Hz octave band babble-noise. The balancing test was completed in open and occluded ears using a loudspeaker and a hearing aid receiver, respectively. A probe tube microphone was used to measure the actual levels presented near the individual's eardrum. The results show that an average of 10 dB higher level was selected for the 500-Hz octave band when listening with the occluded ear than when listening with the open ear. A large range of factors is discussed, but no physical explanation for the discrepancy was found. The findings could have consequences for psychoacoustic experiments and for the use of loudness measurements for hearing aid prescription.

Impact of noise source configuration on directional hearing aid benefit and performance

OBJECTIVE

To evaluate the impact of the position of noise source(s) and reverberation on the directional benefit and performance of three commercially available directional hearing aids.

DESIGN

Directional benefit and performance were measured for four different configurations of competing noise source(s) in two different reverberant rooms. Three pairs of hearing aids representing three commercial models were selected based on electroacoustic evaluation of directivity. Directional benefit and performance of 25 subjects with symmetrical, sloping, sensorineural hearing loss were measured in all test environments using a modified version of the Hearing in Noise Test.

RESULTS

Both reverberation and configuration of the competing noise source(s) significantly affected directional benefit and performance. There was no significant correlation between directional benefit and directional performance. The order of benefit and performance across hearing aid brands (from best to worst) varied depending on the noise source configuration.

CONCLUSIONS

Data revealed increasing reverberation significantly decreased directional benefit and performance. The absolute and relative (rank ordering) directional benefit and performance varied across hearing aid brand, with noise source configuration. These results suggest that data collected in traditional test environments (e.g., a single competing noise placed at 180 degrees azimuth) cannot be used to accurately predict directional benefit or performance in the majority of other test and real-world environments. The impact of reverberation and noise source configuration on directional benefit/performance can be explained fairly well by the interaction between the spatial properties of the noise source(s) and the polar directivity patterns of the hearing aids.

Monaural and binaural loudness measures in cochlear implant users with contralateral residual hearing

OBJECTIVE

The aim was to measure the loudness of monaural and binaural stimuli in a group of cochlear implant users who had residual hearing in the nonimplanted ear, and to consider the implications of these measures for a binaural fitting consisting of a hearing aid and an implant in opposite ears. Three independent hypotheses were addressed: that the shapes of the electric and acoustic loudness growth functions would be similar, although the dynamic ranges would differ; that standard implant and hearing aid fittings would result in substantial loudness mismatches between the acoustic and electric signals; and that loudness summation would occur for binaural combinations of electric and acoustic signals.

DESIGN

A modified version of the "Loudness Growth in 1/2-Octave Bands" method (Allen, Hall, & Jeng, 1990) was used to measure loudness growth for each ear of nine subjects. At the time of the experiment, the subject group included all implant users in Melbourne and Denver who were available for research and who also had sufficient residual hearing to use a hearing aid in the nonimplanted ear. Five acoustic frequencies and five electrodes were measured for each subject. The same subjects also estimated the loudness of a set of stimuli including monaural and binaural signals chosen to cover the loudness range from very soft to loud.

RESULTS

The shapes of the averaged loudness growth functions were similar in impaired and electrically stimulated ears, although the shapes of iso-loudness curves were quite different in the two ears, and dynamic ranges varied considerably. Calculations based on the psychophysical data demonstrated that standard fitting procedures for cochlear implants and hearing aids lead to a complex pattern of loudness differences between the ears. A substantial amount of loudness summation was observed for the binaural stimuli, with most summation occurring when the acoustic and electric components were of equal loudness. This is consistent with observations for subjects with normal hearing and subjects with bilaterally impaired hearing.

CONCLUSIONS

These experiments provide data on which criteria and methods for the binaural fitting of cochlear implants and hearing aids may be based. It is unlikely that standard monaural fitting methods for cochlear implants and hearing aids will result in balanced loudness between the two ears across a reasonably broad range of frequencies and levels. It is also likely that output levels of both devices will need to be reduced relative to a monaural fitting to compensate for the binaural summation of loudness in some listeners.

The acoustic reflex threshold: not predictive for loudness perception in normally-hearing listeners

The working hypothesis of an ongoing study is that the quick and reliable procedure of acoustic reflex threshold (ART) determination in conjunction with measurements of HTL may yield accurate estimates of loudness. The aim of this study was to investigate whether differences in loudness in normally-hearing subjects are reflected in the ARTs and to collect normal material with respect to pure-tone elicited ART and loudness categories. Categorical loudness scaling (CLS) and ART measurements were performed at frequencies of 0.5, 1, 2 and 4 kHz in 60 normally-hearing subjects (HTL<20 dB HL, 26 males, 34 females, aged 21-63 years) with no history or sequelae of middle ear disease. Subjects reporting disturbing tinnitus were excluded. The results show that the ART is not a predictor of individual loudness perception for normally-hearing subjects. Using a numerical scale (HTL=0, 'very soft'=5, 'soft'=15, 'OK'=25, 'loud'=35, 'very loud'=45 and 'too loud'=50) loudness for pure tones grows almost linearly at approximately 0.4 arbitrary loudness units per dB below the 'loud' category. Above the 'loud' category the slope is around 1 unit per dB. The median ART was 85 dB HL at frequencies of 0.5, 1, 2 and 4 kHz. No differences in loudness perception across frequencies were found.

The relationship between the acoustic reflex threshold and levels of loudness categories in hearing-impaired listeners

When applied as a tool for hearing aid fitting, categorical loudness scaling (CLS) is time consuming and not feasible in all subjects. It is therefore desirable to use objective measures for accurate prediction of loudness categories among hearing-impaired individuals. The present study aimed at exploring whether loudness perception at the ART is constant with varying hearing threshold. Seventy-five subjects with various degrees of hearing impairment, measurable acoustic reflex and normal middle ear function participated. The HTLs, ARTs and the levels of six loudness categories at frequencies 0.5, 1, 2 and 4 kHz were determined for all subjects. Loudness at the ART was found to be correlated with the amount of hearing loss. On the basis of these results, it is concluded that the ART cannot be used for accurate estimation of loudness in hearing-impaired subjects.

Inter-relationship between different psychoacoustic measures assumed to be related to the cochlear active mechanism

The active mechanism in the cochlea is thought to depend on the integrity of the outer hair cells (OHCs). Cochlear hearing loss is usually associated with damage to both inner hair cells (IHCs) and OHCs, with the latter resulting in a reduction in or complete loss of the function of the active mechanism. It is believed that the active mechanism contributes to the sharpness of tuning on the basilar membrane (BM) and is also responsible for compressive input-output functions on the BM. Hence, one would expect a close relationship between measures of sharpness of tuning and measures of compression. This idea was tested by comparing three different measures of the status of the active mechanism, at center frequencies of 2, 4, and 6 kHz, using subjects with normal hearing, with unilateral or highly asymmetric cochlear hearing loss, and with bilateral loss. The first measure, HLOHC, was an indirect measure of the amount of the hearing loss attributable to OHC damage; this was based on loudness matches between the two ears of subjects with unilateral hearing loss and was derived using a loudness model. The second measure was the equivalent rectangular bandwidth (ERB) of the auditory filter, which was estimated using the notched-noise method. The third measure was based on the slopes of growth-of-masking functions obtained in forward masking. The ratio of slopes for a masker centered well below the signal frequency and a masker centered at the signal frequency gives a measure of BM compression at the place corresponding to the signal frequency; a ratio close to 1 indicates little or no compression, while ratios less than 1 indicate that compression is occurring at the signal place. Generally, the results showed the expected pattern. The ERB tended to increase with increasing HLOHC. The ratio of the forward-masking slopes increased from about 0.3 to about 1 as HLOHC increased from 0 to 55 dB. The ratio of the slopes was highly correlated with the ERB (r = 0.92), indicating that the sharpness of the auditory filter decreases as the compression on the BM decreases.

Loudness perception is influenced by long-term hearing aid use

The aim of this study was to explore possible differences in the perception of loudness between long-term hearing aid full-time users and non-users. Categorical loudness scaling using pure-tone stimuli was carried out by hearing-impaired subjects. The mean levels of loudness categories at one frequency (hearing threshold: 50-75 dB HL) in a group of 18 hearing aid users (daily use < or = 15 hours per day) were compared with the corresponding levels found in 18 hearing-impaired non-users with the same distribution of hearing thresholds. The results show that, for hearing losses of 50-75 dB HL, the mean level rated as 'loud' by long-term full-time users of hearing aids is 4.5 dB above the mean level of the corresponding category rated by non-users. This difference is statistically significant (P<0.05). No significant differences were found for the lower categories. Among those subjects who had been wearing hearing aids for at least six months, no significant correlation was found between the levels of the 'loud' category and the length of time that hearing aids had been used.

Cross-modality matching: a tool for measuring loudness in sensorineural impairment

OBJECTIVE

The main goal of this study was to establish the viability of cross-modality matching (CMM) for the measurement of individual loudness functions in sensorineural-impaired hearing. To achieve this goal, CMM was tested rigorously to assess four measurement requirements: 1) internal consistency; 2) small relative variance across listeners; 3) test-retest reliability; and 4) data validity.

DESIGN

The measurements involved two sensory continua: perceived length and loudness. Sensation-magnitude functions were generated for all listeners from absolute magnitude estimation (AME) of perceived length, from CMM between loudness and perceived length, and from AME and absolute magnitude production (AMP) of loudness. A total of 211 listeners, 83 with normal hearing at the stimulus frequency and 128 with a diagnosis of cochlear impairment of long duration, performed all four magnitude-scaling tasks. Supplementary loudness matches also were obtained.

RESULTS

Based on the analysis of data, the following results were obtained. First, in accord with loudness measures in normal hearing, loudness measures in cochlear-impaired hearing showed that individuals with bilateral impairments can produce internally consistent loudness data. Second, over the stimulus range where cochlear impairment steepens the loudness function, in a log-log plot loudness slopes derived from CMM, like those obtained from AME and AMP of loudness, were larger in cochlear-impaired hearing than in normal hearing. However, the results of CMM were typically less variable than those obtained from AME and AMP of loudness, permitting a clear-cut distinction between loudness growth rates (slopes) in normal and cochlear-impaired hearing. Third, the results showed that within a cochlear-impaired population, much of the intersubject variability of the slope of the loudness function can be ascribed to the heterogeneity of individual thresholds. Consistent with loudness matching, the size of the slopes increased with the degree of hearing loss. The dependence of the size of the slopes on the degree of hearing loss was observed for hearing losses as large as 75 dB. Fourth, test-retest reliability data for 36 listeners showed that CMM can yield reliable and stable loudness-growth measures in cochlear-impaired hearing over the long term. Finally, equal-sensation matches obtained directly from loudness matching closely agreed with those obtained indirectly from magnitude scaling, indicating that CMM is a valid method for the measurement of loudness magnitudes.

CONCLUSIONS

Taken together, the results demonstrate that CMM can yield stable, accurate, and robust loudness growth measures in cochlear-impaired hearing. Given its apparent reliability, validity, and ease of application, CMM has the potential to become a powerful tool for assessing the growth of loudness in a clinical population. Loudness-level functions derived from CMM may well be important for determining the frequency-gain response of a hearing aid that most closely compensates for the distorted input-output function of the impaired auditory system.

Measurements of loudness growth in 1/2-octave bands for children and adults with normal hearing

The loudness growth in 1/2-octave bands (LGOB) procedure has been shown previously to provide valid estimates of loudness growth for adults with normal hearing and those with hearing loss (Allen, Hall, & Jeng, 1990), and it has been widely incorporated into fitting strategies for adult hearing aid users by a hearing aid manufacturer. Here, we applied a simple modification of LGOB to children and adults with normal hearing and then compared the loudness growth functions (as obtained from end-point data) between the two age groups. In addition, reliability data obtained within a single session and between test sessions were compared between the two groups. Large differences were observed in the means between the two groups for the lower boundary values, the upper boundary values, and the range between boundaries both within and across all frequencies. The data obtained from children also had greater variance than the adult data. In addition, there was more variability in the data across test sessions for children. Many test-retest differences for children exceeded 10 dB. Adult test-retest differences were generally less than 10 dB. Although the LGOB with the modifications used in this study may be used to measure loudness growth in children, its poor reliability with this age group may limit its clinical use for children with hearing loss. Additional work is needed to explore whether loudness growth measures can be adapted successfully to children and whether these measures contribute worthwhile information for fitting hearing aids to children.

An efficient, adaptive method of measuring loudness growth functions

This paper presents a new categorical loudness scaling procedure that differs from previously published loudness scaling procedures by (i) adaptively selecting a new set of levels for each new sequence, (ii) deriving levels that are equispaced on the loudness scale, and (iii) using a continuous scale with few labels. A major advantage of the adaptive procedure is that the individual dynamic range need not be measured prior to loudness testing. The adaptive procedure proved to be time efficient and to produce complete loudness functions from Not heard to Uncomfortably loud for normal hearing and hearing impaired subjects. The pattern of short-term and long-term reliability was similar to that reported for non-adaptive loudness scaling procedures. Three presentations produced a stable loudness function. Normative curves for one octave babble-noise at six test frequencies are presented and compared to normative data obtained with a selection of published categorical scaling procedures.

Accuracy of predicted ear canal speech levels using the VIOLA input/output-based fitting strategy

OBJECTIVE

The Visual Input/Output Locator Algorithm (VIOLA) is a software-assisted method for prescribing amplification targets and selecting a hearing aid to match the targets. Although the procedure calls for selection and fitting of hearing aids in terms of their pure-tone input/output functions in a coupler, it is assumed that a hearing aid that matches the coupler prescription targets will produce specific amplified speech levels in the patient's ear canal. This investigation evaluated the validity of that assumption.

DESIGN

Six hearing aids were evaluated. They were representative of linear and compression processing as well as single- and 2-channel designs. The "subject" was a KEMAR manikin with realistic assumed hearing loss and loudness perception characteristics. Each hearing aid was configured to match the subject's VIOLA prescription as closely as possible. Predicted ear canal speech levels were determined using the prescription rules and modified by the differences between coupler prescription targets and coupler performance of the actual hearing aids. With the subject wearing each hearing aid coupled to an unvented earmold, continuous speech was presented in the sound field and measured, after amplification, in the ear canal. The match between observed and predicted levels of amplified speech indicated the validity of the VIOLA assumptions under examination.

RESULTS

The match between predicted and observed levels was good for soft speech input levels. As speech input levels increased, the differences between observed and predicted levels also increased, with the largest differences seen for loud speech inputs. When differences were seen between observed and predicted levels, they were always in the direction of lower than predicted ear canal levels. The differences between observed and predicted levels were attributed to the effects of limiting, effects of compression ratio in wide range compression, the individual subject's field-to-microphone transfer function, and the subject's individual real-ear-to-coupler level difference.

CONCLUSIONS

Ear canal speech levels were reasonably close to predicted values, and the deviations from predicted levels were plausibly accounted for by consideration of hearing aid performance. Thus, the approach used by the VIOLA procedure holds considerable promise for extending clinical control over the complex and interactive parameters of nonlinear hearing aids. The results of this study indicate that selection and fitting of hearing aids using the current VIOLA procedure usually will result in the generation of lower than predicted ear canal speech levels, especially for loud speech inputs. However, the accuracy of the procedure could be improved substantially by modification of the software to account for the effects of limiting and those of the compression ratio in systems with compression thresholds lower than the level of unamplified loud speech.