Broader Bandwidth Improves Sound Quality for Hearing-Impaired Listeners

This study investigates if wider bandwidth improves sound quality for hearing aid users. Results show preference for wider bandwidths for speech and music, with listeners reporting that a 10,000 Hz bandwidth hearing aid sounded clearer with more nuances audible than an 8000 Hz bandwidth hearing aid.

M. Lisa Sjolander, AuD, is a clinical audiologist and Marcus Holmberg, PhD, is a technical audiologist at Oticon A/S, Smørum, Denmark. Correspondence can be addressed to Lisa Sjolander at or Marcus Holmberg at .

Modern hearing aids are smaller, sleeker, and have broader bandwidths than older-generation devices. Being smaller and sleeker definitely matters. But what about the broader bandwidth?

Bandwidth impacts sound quality, which may in turn affect hearing aid users’ preferences with hearing aids.¹ Perceptions of fuller sound and greater clarity can be expected with broader bandwidths.

Previous studies have shown that some hearing-impaired individuals benefit from and prefer wider bandwidths.^1,2 Karlsen, Flynn, and Eneroth² investigated the effects of 6000 Hz versus 8000 Hz bandwidths on listening effort and consonant recognition, finding significant improvements in consonant recognition with hearing aids that have an 8000 Hz bandwidth. In addition, participants in the study also reported significantly less listening effort with the 8000 Hz bandwidth devices.

Simpson, McDermott, and Dowell³ also investigated the effects of bandwidths on consonant identification for hearing-impaired individuals with moderate-to-severe high frequency hearing losses. Testing was conducted with four bandwidths: 1400, 2000, 2800, and 5600 Hz. They showed that increasing the audibility of high frequency information increased consonant identification scores for 9 out of 10 subjects.

FIGURE 1. Average audiogram of test participants. Bars indicate maximum and minimum hearing losses of test subjects.

Studies indicate that the benefits of broader bandwidths are more pronounced in noise than in quiet for both hearing-impaired and normal-hearing listeners.^5,6 Turner and Henry⁶ suggest that broader bandwidths result in greater accuracy for consonant recognition because higher frequency regions are used to identify consonants when less information is available at lower frequencies. To test the effects of high frequency information on hearing-impaired listeners in quiet and noise, they tested syllable identification in quiet and in noise with seven different bandwidth settings from 350 to 9000 Hz. All 13 hearing-impaired listeners in the study had an increase in correct syllable identification with broader bandwidths, and the improvements with broader bandwidths were greater in noise than in quiet.

Horwitz, Ahlstrom, and Dubno⁵ also show that high frequency information has particular value for hearing-impaired listeners in background noise. They found that consonant recognition in noise increases with increasing bandwidths (the highest bandwidth tested was 5600 Hz). They hypothesize that greater improvements with broader bandwidths in noise are seen because high frequency information is particularly important in noise where fewer redundant speech cues are available.

Ricketts, Dittberner, and Johnson¹ studied another aspect of high frequency listening; they looked at individual preferences for listening with broader and narrower bandwidths for normal-hearing and hearing-impaired listeners. Specifically, they investigated listening preferences for music with 5500 Hz and 9500 Hz bandwidths. Results showed that both hearing-impaired listeners and normal-hearing listeners generally prefer wider bandwidths.

Additionally, Ricketts et al¹ speculate that the low bandwidths available in many hearing aids may contribute to the limited use of hearing aids by hearing-impaired people—particularly those with mild-to-moderate hearing losses who do not like the poor sound quality of hearing aids with narrow bandwidths. Killion⁷ argues that hearing aid users can hear the differences between wider and narrower bandwidths, and that they generally prefer the broader bandwidth due to improved sound quality.

The present study investigates if wider bandwidth improves sound quality for hearing aid users. Participants in the study compared the sound quality of the same hearing aid with 8000 Hz and 10,000 Hz bandwidth.

FIGURE 2. Percentage of the time that 10,000 and 8000 Hz bandwidths were preferred. Error bars indicate 95% confidence intervals; dashed line indicates chance performance.

Methods

Test subjects. A total of 15 people (12 men and three women) with mild-to-moderate hearing loss participated in this study (Figure 1). Participant ages ranged from 32 to 73 years, with an average of 62 years.

Hearing aids. Twelve participants wore their own Dual XW hearing aids with fine tuning. The other three participants were fit with Dual XW hearing aids the day of the test. The automatic features of My Voice, directionality, and noise reduction were turned off during testing to avoid possible interference with the results.

Testing. Participants performed a two-alternative forced-choice test comparing 8000 and 10,000 Hz bandwidth settings for three different signals: Rachmaninoff’s Piano Concerto No. 3, Dolly Parton’s Coat of Many Colors, and a female talker. Stimuli were played at 70 dBC from two speakers in front of the test subject.

Each trial started with exaggerated pairs, comparing 5000 Hz bandwidths to 8000 and 10,000 Hz. These trials were conducted to train the subjects to listen for the differences between the settings. Listeners were not informed what differences they were listening for; they were told only that they were listening for and selecting the setting with the best sound quality.

Participants used a touch screen to switch between settings in the hearing aids. Hearing aid settings were controlled using the Noahlink, which communicated with the hearing aids and the touch screen. Participants were instructed to listen to each of the settings and choose their preferred setting. Participants could switch between the settings as many times as they wanted. After preference selection was made on the touch screen, the next trial started automatically.

After each stimuli set, a brief informal interview was conducted with the participants. They were asked about what differences they heard between the settings, and how easy or difficult it was to hear the differences.

FIGURE 3. Percentage of the time that 10,000 and 8000 Hz bandwidths were preferred with outlier removed. Error bars indicate 95% confidence intervals; dashed line indicates chance performance.

Results and Discussion

Results showed a trend of preference for the 10,000 Hz bandwidth for all stimuli tested. For all results from combined stimuli listeners preferred 10,000 Hz over 8000 Hz 59% of the time. The 5000 Hz training stimuli was also compared to 8000 Hz and 10,000 Hz bandwidths. These comparisons also show a trend of preference for 10,000 Hz over 8000 Hz and 5000 Hz, and for 8000 Hz over 5000 Hz. Figure 2 shows each of the individual stimuli and the stimuli combined.

The results and the participants’ comments indicate that there is a lack of homogeneity among hearing-impaired listeners. Most listeners had a clear preference for higher bandwidth while other listeners showed a less clear trend of preference.

Only one person had a clear preference for the narrower bandwidth setting. This participant reported that she preferred the lower frequency setting because it was more comfortable and that, in general, she finds it to be more comfortable to have less high frequency gain. This participant was considered an outlier due to different preferences from all other participants.

When this person’s results were excluded in statistical analysis, results revealed a significant preference for 10,000 Hz over 8000 Hz (P<0.05, two-sided matched-pair sign test), as shown in Figure 3. These results indicate that most hearing-impaired listeners can be expected to prefer hearing aids with wider bandwidths.

Myths that Discourage Improvements in Hearing Aid Design, by Mead C. Killion, PhD, January 2004 HR.

The results from this study support the results previously obtained for sound quality preferences for hearing-impaired listeners. Ricketts et al¹ found a general preference of normal-hearing and hearing-impaired listeners for a 9000 Hz bandwidth over a 5500 Hz bandwidth. The present study expands these results by showing a preference for 10,000 Hz over 8000 Hz for open fit hearing aid users.

Conclusions

Results from this study show a preference for wider bandwidths for speech and music. Listeners reported that the hearing aid with a 10,000 Hz bandwidth sounded clearer with more nuances audible than a hearing aid with an 8000 Hz bandwidth. These results are important because they indicate that a wider bandwidth may increase hearing-impaired listeners’ overall satisfaction with the sound quality of their hearing aids.

Acknowledgment

The authors thank colleagues at Oticon A/S for their valuable input in this study.

References

1. Ricketts TA, Dittberner AB, Johnson EE. High frequency amplification and sound quality in listeners with normal though moderate hearing loss. J Speech Lang Hear Res. 2008;51:160-72.
2. Karlsen BL, Flynn MC, Eneroth K. The benefit of high-frequency bandwidth for hearing-impaired people under speech in noise conditions. Paper presented at: International Hearing Aid Research Conference (IHCON); 2006; Lake Tahoe, Calif.
3. Simpson A, McDermott HJ, Dowell RC. Benefits of audibility for listeners with severe high-frequency hearing loss. J Hear Res. 2005;210:42-52.
4. Best V, Carlile S. The role of high frequencies in speech localization. J Acoust Soc Am. 2005;118:353-363.
5. Horwitz AR, Ahlstrom JB, Dubno JR. Factors affecting the benefits of high-frequency amplification. J Speech Lang Hear Res. 2008;51:798-813.
6. Turner CW, Henry BA. Benefits of amplication for speech recognition in background noise. J Acoust Soc Am. 2002;112:4:1675-1680.
7. Killion MC. What special hearing aid properties do performing musicians require? Hearing Review. 2009;16(2):20-31.