|Francis Kuk, PhD, is director of audiology at the Widex Office of Research in Clinical Amplification (ORCA), Lisle, Ill, and Lars Baekgaard, MS, is a research engineer at Widex A/S, Vaerloese, Denmark.|
One of the key ingredients to having a hearing aid with high-fidelity sound is to have as broad a frequency output response as possible. While many present-day hearing aids have bandwidths that reach 8000 Hz, some newer hearing aids have a high-frequency response that extends beyond 10,000 Hz. For example, the Widex Passion hearing aid, a receiver-in-the-canal (RIC) style aid, has a bandwidth that extends to more than 10,000 Hz (as measured in an ear simulator). In addition, the use of dual-receiver (or two-way) technology in a micro-size BTE (Widex mind440 m4-m-CB) provides a moderately high output, as well as a broad bandwidth (>10,000 Hz).
While these types of developments mark a technological breakthrough for hearing aid and component manufacturers, it may cause confusion to some dispensing professionals regarding verification and validation procedures. This paper provides background on extending hearing aid bandwidths and considers some of the related issues.
Factors Affecting the Bandwidth of a Hearing Aid
The bandwidth (or frequency response) reflects the range of output signals offered by a hearing aid. In a previous paper,1 we indicated that three stages—the input stage, the processing stage, and the output stage—affect the output bandwidth of the processed signal and the naturalness of sounds.
|FIGURE 1. Real-ear aided response (REAR) of the mind440 m4-m-CB hearing aid when connected to a foam earmold with a Libby horn (red, 4 mm internal diameter), a #13 tube (blue, 1.9 mm), and a thin tube (green, 0.9 mm) for the BTE plumbing.|
Input stage variables. The input stage refers to the microphone and the sampling rate of the analog-to-digital converter (ADC). To have a broad bandwidth, the microphone must have a broad flat response, and the sampling rate of the ADC must be as high as possible.
Many microphones today are sensitive to a broad range of frequencies and are not the limiting factor of the output frequency range. The only exception may be the step microphone (or filtered microphone), which purposely filters off the low frequencies to avoid potential saturation from overdrive.2 Another possibility is filtering off the high frequencies to avoid anti-aliasing artifacts due to insufficient sampling3 or to minimize interference caused by ultrasonic signals that are increasingly used in commercial sensors and security systems.
Of more relevance is the sampling rate. The Nyquist theorem requires that the sampling frequency must be at least twice as high as the highest frequency in the input in order to correctly capture the nuances of the input signal. Thus, a sampling rate of 20 kHz is required to capture inputs as high as 10 kHz. If the input has energy at 16 kHz, the information between 10 kHz and 16 kHz would not be captured correctly. This means that the usable signal from the hearing aid would not exceed 10 kHz even though the bandwidth may be stated as much higher. That is, a bandwidth of 12 kHz only makes sense if the input stage of the hearing aid is able to handle frequencies up to 12 kHz.
Output receivers and coupling. The output stage of a hearing aid is primarily affected by the receivers used. Receivers transduce the electrical output of the hearing aid into an acoustic form. The types of receivers (size, materials, etc) determine its electrical output characteristics (eg, maximum output levels, resonance peaks, and bandwidths).
However, the acoustic spectrum is further shaped by the presence of any earhook/tubing connected to the receivers. All receivers have some lengths of tubing connected; however, the length and the diameter of the tubing connected to a receiver vary depending on whether it is a BTE or an ITE product:
BTEs. A short tubing is used in a BTE to direct the acoustic output from the receiver to the earhook. A longer piece of tubing from the earmold connects the earhook to the earmold, and directs the acoustic output to the wearer’s ear canal. These tubings have natural resonances that further shape the receiver frequency response.
A typical BTE tubing system acts like a tube that is closed on one end (receiver end) but open on the other (eardrum end). This gives the opportunity for ¼-wavelength resonances. That is, frequencies that have a wavelength 4 times the length of the tube would produce resonance peaks at odd multiples of themselves. For a #13 tube (internal diameter of about 1.9 mm) with a typical length of 75 mm (or 3 inches), the tubing resonance occurs at 1000 Hz, 3000 Hz, and 5000 Hz. The “bumps” around 2000 Hz and 4000 Hz are, respectively, the mechanical resonance of the receiver and the Helmholtz resonance formed by the air within the tube system and the air inside the receiver.4
Today, thin tubes of internal diameters around 0.9 mm have gained popularity. The use of this type of tubing shifts the tubing resonance downwards in frequencies, as well as reduces high-frequency output by 5 dB to 10 dB when compared to standard #13 tube fittings.5 Thus, the use of thin-tube fittings narrows hearing aid bandwidth (relative to #13 tubing).
Looking at it another way, tubes with a larger internal diameter (than #13 tubing) enhance the magnitude of the high-frequency output. An example is the use of Libby horn tubing, which gradually increases its internal diameter from 1.9 mm to 3 or 4 mm at its medial end. Figure 1 shows the in situ output of the mind440 m-CB hearing aid connected to a Libby horn, a #13 tube, and a thin tube. Despite the fact that the receiver used in this example is one with the extended bandwidth (to 10,000 Hz), the use of the Libby horn has further increased the output in the high frequencies by 4-5 dB. In contrast, the use of thin tubing decreased the output above 1000 Hz by as much as 10 dB at some frequencies.
While some may question the need for acoustic plumbing in today’s digital hearing aids that offer flexibility in frequency shaping, it can still be very useful in cases where the degree of hearing loss pushes the output toward the maximum power output (MPO) of the hearing aid.
ITEs. For an ITE/ITC/CIC, the tube connecting the receiver to the sound outlet is typically not long enough to provide significant peaks. Thus, the typical peaks around 2000 Hz and 3000 Hz are the mechanical resonance of the receivers. It is typically higher in frequency than the mechanical resonance of a BTE receiver because of its smaller size and lighter, stiffer diaphragm.
Receiver dimensions. The output SPL of a receiver and its frequency response is dependent on the physical dimensions of the receiver. In general, the larger the receiver, the greater the air volume for the diaphragm to vibrate. This results in a higher output before peak clipping saturation occurs.
The bandwidth of a receiver is also limited by the size of the receiver. As indicated earlier, a receiver works by moving the diaphragm in synchrony with the frequency of the changing electrical field. The higher the frequency (or the broader the bandwidth), the more rapidly the diaphragm needs to move. This is possible with a smaller receiver (thus a smaller and stiffer diaphragm) that needs to drive a small volume of air. It becomes more difficult in the higher frequencies with a larger receiver without compromising the low-frequency sensitivity of the receiver and increasing the current drain of the hearing aid dramatically.
Thus, the bandwidth of most hearing aid receivers is a compromise of current drain, size, and the desired frequency region where special attention is needed. That is, if one needs to have an extended bandwidth in the high frequencies, one may need to sacrifice the low-frequency sensitivity of the hearing aid and vice versa. Figure 2 shows the physical dimensions of several receivers used in the Widex hearing aids and the associated frequency response and MPOs. The trend of a higher MPO, but narrower bandwidth, as the receiver gets larger is clear.
One solution to achieve an extended high-frequency bandwidth without too much compromise is the use of two-way or dual receivers. In simple terms, this is a two-receiver system where one is optimized for low frequencies and the other is optimized for high frequencies. The output from both receivers is summed before reaching the eardrum. The advantage of such an approach is that an extended high-frequency bandwidth can be achieved in a moderately high output receiver with little compromise on the low-frequency response. Furthermore, it economizes on the current drain and minimizes on the potential of saturation distortion. The micro-sized, Widex m4-m-CB (ClearBand) model uses such a receiver system to reach an in situ bandwidth that extends from 100 Hz to 10,000 Hz.
|TABLE 1. Comparison chart showing the direct correspondence between the physical dimensions of the receivers and the frequency response and maximum power output of hearing aids.|
Broader Bandwidth: Who Benefits?
The literature on the benefits of an extended bandwidth appears mixed, depending on the hearing status of the subjects. The majority of studies conducted in normal-hearing listeners and those under headphones concluded that a broader bandwidth was associated with better sound quality and better speech intelligibility. The benefit was reported in quiet as well as in noise for both normal-hearing and hearing-impaired subjects with less than a moderate degree of hearing loss.6-9
In studies with hearing-impaired children, it has become generally accepted that an extended bandwidth is necessary for optimal speech and language development. For example, Stelmachowicz et al10,11 showed significant improvement in identification of /s/ with increasing bandwidth. Some studies also suggested that high frequency is necessary for optimal localization in the vertical plane.12
Despite these endorsements for a broad (or extended) bandwidth, several studies have questioned the limited benefit of extended bandwidth for people with a sloping high-frequency hearing losses exceeding 55 dBHL.13,14 Indeed, these studies showed a decrease in speech recognition with increasing bandwidth. Additionally, the concept of “dead” regions15 raises additional questions on the need for, and the appropriateness of, amplification for high-frequency hearing loss.
Recently, Hornsby16 reported on the ability of 63 people to identify sentences in noise in a wideband (178 Hz to 9000 Hz) and a low-pass (3534 Hz) condition. While all subjects showed some benefits with the broadened bandwidth, less benefit was seen with the subjects with more severe losses. In the same presentation, Ricketts17 showed that subjects who preferred a wider bandwidth exhibited audiogram slopes of 7.5 dB/octave or less, whereas those who showed a consistent preference for a narrower bandwidth exhibited audiogram slopes of 12.5/octave or more. The presenters concluded that those who had a steeper audiogram slope (ie, more severe high-frequency loss) preferred a narrower bandwidth.
|FIGURE 2. Comparison of coupler gain of the mind4-m-CB hearing aid on a 711 coupler (blue) and a HA-2 coupler (red).||FIGURE 3. REAR of the mind440-m-ClearBand hearing aid (above 3000 Hz) measured at three distances from the eardrum. Note the similarity in output below 4000 Hz with the three distances.|
So, How Should I Look at Extended Bandwidth, in Practice?
To summarize, the available studies seem to suggest that normal-hearing people and people with a mild hearing loss would most likely realize the sound quality and speech intelligibility benefit of the extended bandwidth. As hearing loss increases, it is likely that the noted benefits decrease.
Given today’s transducer technology, it is unlikely that hearing aids with an extended bandwidth will benefit those with a severe-to-profound degree of hearing loss in the high frequencies. There are at least two reasons for this.
- Current hearing aids with an extended bandwidth have only a moderate MPO (peak <125 dBSPL). This means that, even though the bandwidth may be broader, there is still insufficient high-frequency output to meet the demands of a severe-to-profound hearing loss. Indeed, as Figure 2 shows, hearing aids intended for a severe-to-profound loss typically have a bandwidth that is significantly narrower than those for a milder degree of hearing loss. In current hearing aids, bandwidth behaves in the opposite manner as MPO.
- The likelihood of a dead region increases as the degree of hearing loss increases.15 This is especially the case for a precipitous hearing loss. The implication is that, even though sufficient output may be available, the wearer with a severe-to-profound loss in the high frequencies may not be able to utilize the available high-frequency output. This possibility is seen in the studies, which showed a lack of benefit from the extended bandwidth in patients with even a moderately severe hearing loss. These patients may be better served with the use of frequency transposition.18
Verification of Extended Bandwidth
The reported bandwidth of a hearing aid is affected by the coupler that is used for the measurements. Typically, the output measured on standard 2cc couplers (HA-1 and HA-2) only reflects the hearing aid output for the specific coupler. It does not reflect the actual output in the ear canal. Thus, standard 2cc coupler responses can be used only for quality assurance purposes.
A more accurate way to examine the in situ output of the hearing aid is to perform real-ear measurements. Alternatively, if one is interested in the hearing aid output in the average adult ear, one may use a 711-coupler or a Zwislocki coupler in which the impedance of the coupler matches more closely that of the average adult ear. That is why some manufacturers also report hearing aid output-responses on the 711-coupler. In general, the output from a 711-coupler is higher in magnitude and broader in bandwidth than that measured with a standard 2cc coupler. The difference increases as the frequency increases.
Figure 2 shows the output difference between a 711-coupler and a 2cc HA-2 coupler. One can see differences as much as 15 dB in the high frequencies. While the 711-coupler is more reflective of real-ear output, it is more costly and is not readily available for clinical use.
One would need to be even more careful when verifying the real-ear output of hearing aids with an extended bandwidth. The closeness of the probe-tube microphone to the eardrum could affect the accuracy of the measurement in the high frequencies. The closer the probe-mic is to the eardrum, the lower the risk of the probe-mic encountering standing wave minima and yielding an inaccurate measurement. On the other hand, a probe-mic placed close to the eardrum causes discomfort to the patient. As a compromise, it has often been recommended that the probe tube be placed around 5 mm beyond the tip of the earmold or within 6 mm from the eardrum. This would allow one to reliably measure the high-frequency output of the hearing aid to within 2 dB through 6000 Hz.19
However, the insertion depth that may be appropriate for measuring output at 4000 Hz may not be adequate for measuring output above that frequency. Figure 3 shows the real-ear output of the mind m4-m-CB hearing aid (with the extended high-frequency output) measured almost at the eardrum (12 mm from earmold tip), and at 3 mm (or 9 mm from earmold tip) and 6 mm from the eardrum (or 6 mm from earmold tip). The difference in output at 4000 Hz was within 1 dB among the three insertion depths. However, up to a 2 dB difference was noted between the 0 mm and 3 mm distance, and 15 dB at the 6 mm distance at 6000 Hz.
This suggests that, when validating the additional high-frequency output of a hearing aid with an extended bandwidth, it is important to ensure that the probe-mic is within 3 mm of the eardrum in order to accurately measure the high-frequency output above 4000 Hz. Otherwise, one may reach a false conclusion on the high-frequency output of the device. In his book, Dillon4 provides a summary chart (p 93) on the required distance of the probe-mic from the eardrum to ensure acceptable measurement errors.
In summary, the achievement of a bandwidth that extends beyond 10,000 Hz makes it possible for more people with a mild-to-moderate degree of hearing loss to enjoy even higher fidelity sounds than what is currently available. However, as a clinician, one needs to be realistic about the types of candidates for this technology, and use additional care when documenting the output of these hearing aids.
Citation for this article:
Kuk F, Baekgaard L. Considerations in fitting hearing aids with extended bandwidths. Hearing Review. 2009:16(11):32-38.
- Kuk F, Jessen A, Baekgaard L. Ensuring high fidelity in hearing aid sound processing. Hearing Review. 2009;16(3):34-43.
- Chasin M, Schmidt M. The use of a high frequency emphasis microphone for musicians. Hearing Review. 2009;16(2): 32-37.
- Kuk F. The effects of distortion in hearing aids. In: Valente M, ed. Hearing Aids: Regulations, Options, and Limitations. New York: Thieme Medical Publishing; 1996:327-367.
- Dillon H. Hearing Aids. Sydney: Boomerang Press; 2001.
- Kuk F, Baekgaard L. Hearing aid selection and BTE: choosing among various “open-ear” and “receiver-in-canal” options. Hearing Review. 2008;15(3):22-36.
- Hornsby B, Ricketts T. The effects of hearing loss on the contribution of high- and low-frequency speech information to speech understanding. J Acoust Soc Am. 2003;113:1706-1717.
- Hornsby B, Ricketts T. The effects of hearing loss on the contribution of high- and low-frequency speech information to speech understanding. II. Sloping hearing loss. J Acoust Soc Am. 2006;119(3):1752-63.
- Ricketts T, Dittberner A, Johnson E. High-frequency amplification and sound quality in listeners with normal through moderate hearing loss. J Sp Lang Hear Res. 2008;51:160-172.
- Stelmachowicz P, Pittman A, Hoover B, Lewis D. Effect of stimulus bandwidth on the perception of /s/ in normal- and hearing-impaired children and adults. J Acoust Soc Am. 2001;110(4): 2183-90.
- Stelmachowicz P, Pittman A, Hoover B, Lewis D. Aided perception of /s/ and /z/ by hearing- impaired children. Ear Hear. 2002;23:316-324.
- Byrne D, Noble W. Optimizing sound location with hearing aids. Trends Amplif. 1998;3(2):51-73.
- Ching T, Dillon H, Byrne D. Speech recognition of hearing-impaired listeners: predictions from audibility and the limited role of high-frequency amplification. J Acoust Soc Am. 1998;103:1128-1140.
- Hogan C, Turner C. High-frequency audibility: benefits for hearing-impaired listeners. J Acoust Soc Am. 1998;104:432-441.
- Moore B. Dead regions in the cochlea: diagnosis, perceptual consequences, and implications for the fitting of hearing aids. Trends Amplif. 2001;5(1):1-34.
- Hornsby B. Hearing aid bandwidth for speech understanding. Paper presented at: American Academy of Audiology convention; April 2007; Denver.
- Ricketts T. High frequency extension and sound quality. Paper presented at: American Academy of Audiology convention; April 2007; Denver.
- Kuk F, Keenan D, Peeters H, Korhonen P, Auriemmo J. Twelve lessons learned about linear frequency transposition. Hearing Review. 2008;15(11):32-41.
- Revit L. Real-ear measures. In: Valente M, ed. Strategies for Selecting and Verifying Hearing Aid Fittings. 2nd ed. New York: Thieme; 2002:66-124.