Directional microphones have become a common feature in hearing aids to help hearing aid users understand speech in noisy situations. Recent technological enhancements to directional devices provide consumers with a better overall listening experience in noise.
In 2000, the author published an article titled “Signal-to-noise Ratio Loss and Directional-microphone Hearing Aids” in Seminars in Hearing.1 This article was written at a time when directional-microphone hearing aids were experiencing a revival of sorts in the hearing aid industry. Although directionality had been available in hearing aids before this time (the first directional-microphone hearing aids entered the market in the early 1970s), the hearing aids incorporating directionality in 2000 had higher directivity indices (DIs) and the ability to switch between omnidirectional and directional settings. These improvements ultimately provided greater overall end-user benefit and led to the acceptance of directionality as an important feature in hearing aids.
Eleven years later, directionality has become standard in even lower-cost hearing aids. While directionality is available to more individuals who wear hearing aids, technological enhancements and new developments have made directionality even more beneficial today. The improvements to directionality in the past 11 years will be the topic of this paper.
Laurel A. Christensen, PhD, is chief audiology officer at GN ReSound Group, Glenview, Ill.
Directional Benefit and Limitations
SNR tests and limiting conditions. Individuals with hearing impairment not only have hearing loss—or loss of audibility, which is compensated for through amplification in hearing aids—but also have increased difficulty hearing in noise. This difficulty hearing in noise can be quantified by measuring a listener’s signal-to-noise ratio (SNR) loss.2 SNR loss is the increase in SNR (in dB) required by someone with a hearing loss to understand speech in noise, relative to the average SNR required for listeners with normal hearing.
Two tests are commercially available to measure SNR loss: the Speech-In-Noise test (SIN)3,4 and the Hearing-in-Noise Test (HINT).5 There are large individual differences among hearing-impaired listeners on measures of SNR loss. A general trend is that SNR loss increases with hearing loss, but the variance is quite large and can range from no loss (normal-hearing performance in noise) to greater than 20 dB of SNR loss.
To provide better hearing in noise, hearing aids incorporating directionality improve the SNR for the users.6-8 A hearing aid with directionality can improve the SNR approximately 3 to 8.5 dB,6,7 but this varies with different environments and the location of the speaker. Walden and colleagues9 found that directional microphones in hearing aids are most effective when the signal of interest is in front of the listener, within 10 feet, and the background noise is spatially separated from the signal of interest location.
These conditions limit directional benefit, but the provided improvement in SNR can help those with SNR losses ranging from 4 to 8 dB in many environments. For those with greater SNR losses, a directional-microphone hearing aid can still provide the benefits of listening comfort and can help with understanding of speech in noise as long as contextual and speech cues are available. Listeners with SNR losses greater than 8 or 10 dB will need additional help to understand in noise. Later in this paper, some new technology is discussed that can effectively help these listeners understand speech in noise.
Subjective benefit from directionality. Kochkin’s MarkeTrak VI10 article in the October 2002 HR about satisfaction with hearing aids pointed out that only 10% of the hearing aids in the survey that year were directional. He compared MarkeTrak satisfaction to a sample of nearly 500 DSP hearing aids of the same brand from one European manufacturer, segmented as to whether they were omnidirectional (n=200) or directional (n=296). The study demonstrated that directional DSP technology was decidedly superior to omnidirectional DSP technology. Multiple-microphone (directional) digital hearing aids were rated 17% points higher on consumer satisfaction than the typical hearing aid in the market, and were rated higher on 18 of 39 factors measured in the study. Multiple-microphone digital hearing aids were rated higher than single-microphone product by 14 percentage points on overall customer satisfaction, on 14 of 39 MarkeTrak variables including six of the top-10 factors impacting overall customer satisfaction, as well as 5 of 13 listening situations measured in the study. Compared to the typical hearing aid in the market (MarkeTrak),11 directional digital hearing aids showed double-digit customer satisfaction improvements on 14 customer satisfaction variables compared to 4 for omnidirectional digital hearing aids.
In 2010, although satisfaction for all hearing aid users (those with hearing aids both with and without directionality) in MarkeTrak VIII12 improved by an average of 7 percentage points in noisy situations since 2004 (likely because so many instruments in this survey were equipped with directionality), 25% of consumers still gave “Use in noisy situations” a negative ranking, and only 37% were “satisfied” or “very satisfied” when using hearing aids in noise.
This is likely due to multiple factors, including the large variability in SNR losses among hearing-impaired listeners, giving some listeners limited benefit in real-world situations even when using hearing aids with directionality. Other factors limiting benefit include some problems that have traditionally accompanied hearing instruments with directionality, such as poorer sound quality when using this feature, switching algorithms that are not entirely accurate, and over-amplification of near-field signals (eg, wind noise).
Recent developments and technological enhancements in directional-microphone hearing aids have tried to address some of these shortfalls and further improve user benefit. The following will focus on these developments, including the emergence of wireless technology, which provides the opportunity for even greater help when listening in noise for the hearing aid user.
Split-band Directionality for Better Sound Quality
While the benefit of using directionality in noisy environments has been demonstrated in both the real world and laboratory as outlined above, until more recently, listening in the directional setting has often come at the cost of sound quality.
Inherent in the design of directional hearing aids is a low frequency roll-off, which occurs because low frequency sounds have similar phase relationships between the front and rear microphones. To accommodate for the decrease in audibility caused by this roll-off in the directional setting, a boost in low frequency amplification is usually applied—called equalization.
Equalization causes the internal noise floor of the hearing aid to increase and ultimately can detract from the benefit of the directional setting.13 However, not compensating for this low frequency roll-off has the undesired result of a tinny sound quality in the directional setting. Thus, traditional designs have the trade-off of either being too noisy or too tinny rather than sounding natural.
A solution for this problem is to process the sound in the hearing aid the same way it is processed by a normal-hearing listener. This processing, called split-band directionality, approximates the unaided ear’s natural directional characteristics. Figure 1 illustrates how the open ear and split band directionality are similar. The KEMAR response for four frequencies is shown on the left in the polar plot. For the two lower frequencies, the response is essentially omnidirectional; for the higher frequencies, it is directional to the front.
Figure 1. Directional responses of the open ear on the left and split-band directionality on the right. Note that the two lower frequencies are more omnidirectional, while the two higher frequencies are heard more directional with attenuation to the sides and back.
In Figure 1, the panel on the right presents the same measurements performed on a hearing aid with split-band directionality. There is a good match between the split-band directional response and those of the open ear. Processing sound in a hearing aid this way results in more natural sound quality for the end user, but preserves the directional benefit that is present in traditional directional settings.
Research has shown a preference for listening in omnidirectional settings when compared to directional settings. In a study investigating the impact of visual cues on directional benefit, Wu and Bentler14 reported that many individuals fit with an equalized directional response experienced a “hissing sound.” In a subsequent field trial with the same participants and hearing instruments, Wu and Bentler15 found that loudness and internal noise were the most important predictors for preference of omnidirectional over directional microphone mode.
Other studies have also demonstrated strong preferences for omnidirectional microphone mode—even in situations where directional processing should provide more benefit.9,16 Split-band directionality provides a directional pattern closer to a person’s open ear, thereby striking a natural balance between environmental awareness and directional advantage. Groth and colleagues17 summarized the results of three studies investigating the effect of directional processing on sound quality. All three investigations used a double-blind design in which hearing-impaired listeners expressed a preference for the split-band directionality, omnidirectional processing, or a traditional directional response. Listeners indicated an overwhelming preference for the sound quality of omnidirectional processing over traditional directional processing, and preferred split-band directionality over traditional directionality more than twice as often.
An additional advantage of processing sound in a split-band manner is the spectral preservation of the low frequencies, allowing the listener to take advantage of the natural ear timing differences, which are important for sound localization. A recent study18 showed that interaural time differences (ITDs) are the most important cue to preserve for localizing sounds. In fact, the results indicated that interaural intensity differences (IIDs) could be mismatched up to 9 dB by compression in the hearing aid and not affect localization performance as long as some ITD cues were available.
These important ITDs are maintained in the split-band approach, as evidenced in a recent article from Groth and Laureyns.19 They reported on a study about the effect of different directional processing schemes on left/right and front/back localization performance of hearing-impaired listeners. The results showed that localization ability was maintained relative to the open ear using split-band directional processing.
Finally, one benefit of split-band directionality that deserves mention is that it also solves the problem of over-amplification and distortion of near-field signal, such as one’s own voice or wind noise in traditional directional settings. Again, equalization in a traditional hearing aid can make these signals distorted, and using omnidirectional settings in the low frequencies does not cause this over-amplification leading to distortion.
Some commercially available hearing aids with split-band directionality have the ability to set the frequency at which the processing of the input changes from omnidirectional to directional. This frequency (called the blending point) is set relative to the hearing loss of the patient. In general, if the average of a hearing aid user’s thresholds at 250 and 500 Hz is less than 40 dB, the frequency is set higher. This reduces the bass-boost-induced low frequency noise the user may experience with traditional directional processing. Conversely, if the average is greater than 40 dB, the frequency is set lower as low frequency noise would be less audible to this user.
In summary, split-band directionality is a method in which omnidirectional processing is applied to the low frequencies and directional processing is applied to the high frequencies, and then the signals are mixed. This type of processing allows for better sound quality while also preserving localization cues when listening in the directional hearing aid setting and still delivers good directional SNR improvement to the end users.
Asymmetrical Directionality
As stated earlier, directional microphones that were introduced in the early 2000s could be switched from omnidirectional to directional settings manually by the hearing aid user. In 2004, Cord and colleagues20 published a study that indicated that many users (30%) did not switch between the settings. The study stated that users often did not know when to switch and/or did not want to do this manual switching in everyday life.
To overcome this manual switching problem, hearing aids were introduced that automatically change from the omnidirectional to the directional setting, depending on the environment. These types of switching algorithms depend on environmental classification systems, which analyze the acoustic scene and make a decision about which microphone mode is most beneficial.
Thus, these systems are limited by the accuracy of the classification system and have no ability to determine the hearing aid user’s intent in complex listening situations. One field trial of automatic switching systems (Andrew Dittberner, PhD, unpublished data, 2011) showed that the switching systems were in the directional settings from 5% to 17% of the time. The results of Walden et al16 suggest that the average user is in an environment in which a directional-microphone setting can be beneficial approximately 33% of the time.
Thus, the switching algorithms currently used in hearing aids can be too conservative, with the end result being that the user is not in the directional setting when it could be beneficial. Additionally, the Cord et al20 study showed that, although many patients do not use their manual switching option, those who do use the option prefer the manual mode rather than rely on the decisions of automatic switching algorithms. The reason for this might be that the automatic switching algorithms are not switching effectively and/or appropriately.
The standard way to use directional processing in a bilateral hearing aid fitting has been to apply directionality simultaneously in both hearing aids of a binaural fitting. In other words, both hearing aids are in the directional setting in a noisy environment. Another way to use directional processing is to keep one hearing aid set to omnidirectional and the other hearing aid set to directional. This seemingly unconventional way to apply directional processing can provide a better listening experience for users of hearing aids and overcomes the limitations of directional systems discussed above. Specifically, an asymmetric fitting can overcome the lack of use of manual systems and the reliance on environmental classification systems. An additional benefit is that it does not cut a listener off from their environment as wearing two hearing aids in directional settings can do. The user can choose to attend to whatever signal they may be interested in hearing.
Figure 2. Figure illustrating a social gathering in a home. The man in the center talking to the woman is wearing hearing aids. His listening experience is different depending on whether he is wearing both hearing aids set in directional settings (depicted by the triangle) or an asymmetric fitting (depicted by the circle). The asymmetric fitting allows the listener the benefits of the directionality without cutting him off from the rest of the party.
The key to asymmetrical directionality is to understand that one hearing aid in the directional setting and one in the omnidirectional setting provides the same SNR benefit as using two hearing aids set in the directional settings. Several studies have verified this including Bentler et al,21 Cord et al,22 and Mackenzie and Lutman.23
Using hearing aids set asymmetrically comes with the added benefit of maintaining maximum auditory awareness for sounds arising from any direction. It was noted in this article that Walden et al16 determined that directional microphones work best when the signal of interest is:
- Close to the listener;
- In front of the listener; and
- The undesired signal (noise) is spatially separated from the signal of interest (usually speech).
In the real world, there are many environments where these conditions would not be true in a noisy environment. In fact, the signal of interest in real life is not always in front of the listener. The desired signal can be at any location. There can also be multiple signals of interest in an environment. For example, when a hearing aid user is sitting around a table with many speakers, the directional microphone settings might cut a listener off from what they want to hear. If a listener is using two hearing aids set to directional settings, they can be cut off from their environment, making it difficult to even be aware of sounds from other directions.
A picture of an acoustic scene is perhaps the best way to illustrate the benefit of using an asymmetric hearing aid fitting. Figure 2 illustrates a social gathering in a home. Let’s assume that the man talking to the woman wears hearing aids, and they are both standing. The experience he has will be different depending on if his hearing aids are both set to a directional setting or if the aids are set to an asymmetric fitting. A traditional binaural directional-microphone setting is depicted grossly by the triangle. Thus, the listener will hear what is in front of him, and he will find it difficult to hear what the men to his right are saying. In an asymmetric directional setting, he hears what is in front (ie, as in a traditional binaural directional setting), but has the added benefit of maintaining the ability to hear and monitor other sounds coming from other directions (depicted by the circle). He can then turn and attend to other conversations as desired. In the traditional setting, users may report a feeling analogous to tunnel vision. That is, they are cut off from their surroundings. Asymmetric fittings give the hearing-impaired listener an experience more like that of normal-hearing listeners.
Finally, Cord et al22 found improved ease of listening for asymmetric directional fittings as compared to bilateral directional fittings. Users do not feel as isolated from sounds originating from the sides and rear due to the environmental sound cues from the omnidirectional processing that is always available to them.
In summary, the use of an asymmetrical fitting gives the benefit to using two hearing aids set to directional settings without cutting the listener off from the environment. This allows the listener to determine what is important for them to listen to and not rely on an automatic-switching hearing aid that does not understand the listener’s intent and may not classify the acoustic environment accurately.
Steering of Directionality and Adjusting Directional Beam Width
Two other developments in directionality have centered on steering the pattern of the directional response and adjusting the beam width. Steering of directionality is the ability of the directional system to move the area of most sensitivity to a location other than the front of the listener. As noted in the previous section on asymmetric directionality, the signal of interest is not always facing the listener. Consider the often-cited situation where the driver of a car might want to hear the passengers in the backseat or the passenger sitting beside him. It is unsafe for drivers to turn their heads toward these speakers. Some switching systems give the user control over where to steer the microphone, making it most sensitive to the front, back, or the sides. Some devices automatically steer the directionality depending on the environmental input. There is little data to support again that automatic switching of pattern sensitivity is effective, but manual switching or asymmetric directionality may be useful to some patients in specific listening environments.
Adjusting the beam width of the pattern is the ability to make the pattern narrower so as to focus more sharply to the front. Narrowing the beam width has the benefit of filtering out more background noise to the sides and behind the listener. Hearing aids are available on the market for which the dispensing professional can set up a program with a narrower directional beam, or the beam can be narrowed automatically dependent upon the level of the signal to the front. Figure 3 shows the different beam width settings of a directional microphone system.
Figure 3. Examples of three settings of a hearing aid with adjustable directional beam widths. Left to right: Narrow, Medium, and Wide corresponding to listening scopes of approximately ±50°, 70°, and 90°, respectively.
Microphone Arrays
Wireless communication between hearing aids has allowed for the implementation of another approach to directionality that uses an array of microphones. In this approach, the two microphones on one hearing aid are linked to the two microphones on the other aid. This allows the null points of the beam to be moved further to the front and a narrower beam (approximately ±45°) to be created by the array; traditional dual microphone systems are limited to a front beam with an angle of ±60°.
A narrower beam provides the potential for a more favorable SNR. However, the narrow beam described here is limited to situations where a listener would want to focus on one speaker located in front of them in a diffuse noise environment. The implementation of this feature in current technology requires the user to switch to a separate program to use the narrow beam. As described above in the section on asymmetric fittings, the disadvantage of such a system is being cut off from all other signals coming from other directions.
A research study comparing a narrow beam (±45°) created by a microphone array and a more traditional beam (±60°) resulted in approximately a 1 dB improvement in a laboratory condition. This specific condition utilized a diffuse noise environment with speech shaped noise at ±45°, ±90°, ±135°, and 180°. In another condition tested, a diffuse noise environment with continuous babble and ICRA4 noise from ±60° and 90°, as well as babble noise from ±135° and 180°, showed no difference in performance between the two conditions.24 These results emphasize the specific environment in which this feature could be useful.
Companion Microphones
While typically not a directional microphone, companion microphones can dramatically increase the ability to hear from a specific direction. Early systems used a microphone on a wire plugged directly into a hearing aid. More commonly, FM systems have provided this benefit with fewer wires, and the hardware has been reduced significantly over the years. SNR improvements with FM systems can be on the order of 15 to 18 dB depending on the listening environment. These devices have been most commonly used for children in the classroom and are not widely used by adults—even though personal FM systems that can be used with hearing aids have been on the market for many years. The cost of these systems may be one reason for this.
Figure 4. This figure depicts on the right a speaker wearing a companion microphone. This speaker’s voice is then wirelessly directed to the hearing aid user’s hearing aids.
Wireless connectivity in hearing aids has great potential to provide listeners with solutions that provide SNR improvement comparable to FM while being cosmetically acceptable and low in cost. Figure 4 shows a listener receiving the speaker’s voice directly to his/her hearing aids through a wireless connection. This type of accessory is beneficial in many environments, such as restaurants, meetings, lectures, and an automobile. One such device has been introduced to the market in which a microphone is worn by a speaker and the voice is sent through an intermediary device to the hearing aids.
In the future, expect more of these systems to be introduced. Companion microphones can be given to a speaker in a difficult listening situation, and the speaker’s voice will be picked up and sent wirelessly to the hearing aids (Figure 4). These types of microphones, with ranges of up to 15 meters (about 50 feet), will bring the ability to hear in noisy situations to more users than ever before. Future developments will likely make these types of microphones available for more than one speaker so that a hearing aid user can listen to multiple speakers in a noisy environment.
Summary
During the last decade, several discoveries and new technologies have improved directional hearing aid systems. These developments offer better sound quality, a more natural listening experience, and even better directionality than what was available when these digital directional microphones were introduced in the early 2000s.
Correspondence can be addressed to HR or to Laurel A. Christensen, PhD, at .
References
- Christensen LA. Signal-to-noise ratio loss and directional-microphone hearing aids. Seminars in Hearing. 2000;21(2):179-200.
- Killion MC. SNR loss: I can hear what people say, but I can’t understand them. Hearing Review. 1997;4(12):8-14.
- Fikret-Pasa S. The Effect of Compression Ratio on Speech Intelligibility and Quality [PhD thesis]. University Microfilms, Ann Arbor, Mich: Northwestern University; 1993.
- Killion MC, Schulein R, Christensen L, et al. Real-world performance of an ITE directional microphone. Hear Jour. 1998;51:24-38.
- Nilsson M, Soli SD, Sullivan J. Development of the Hearing In Noise Test for the measurement of speech reception thresholds in quiet and in noise. J. Acoust Soc Am. 1994;95:1085-1099.
- Valente M, Fabry D, Potts L. Recognition of speech in noise with hearing aids using dual microphones. J Am Acad Audiol. 1995;6:440-450.
- Pumford J, Seewald R, Scollie S, Jenstad L. Speech recognition with in-the-ear and behind-the-ear dual-microphone hearing instruments. J Am Acad Audiol. 2000;11:23-35.
- Walden BE, Surr RK, Cord MT, Edward B, Olson L. Comparison of benefits provided by different hearing aid technology. J Am Acad Audiol. 2000;11:540–560.
- Walden BE, Surr RT, Cord MT, Grant KW, Summers V, Dittberner AB. The robustness of microphone preferences in everyday listening environments. J Am Acad Audiol. 2007;18:358-379.
- Kochkin S. 10-year customer satisfaction trends in the US hearing instrument market. Hearing Review. October 2002;9:14-25,46.
- Kochkin S. Customer satisfaction with single and multiple microphone digital hearing aids. Hearing Review. November 2000;7:24-29.
- Kochkin S. MarkeTrak VIII: Consumer satisfaction with hearing aids is slowly increasing. Hear Jour. 2010;63(1):19-32.
- Ricketts T, Henry P. Low-frequency gain compensation in directional hearing aids. Am J Audiol. 2002;11(1):29-41.
- Wu Y, Bentler RA. Impact of visual cues on directional benefit and preference: Part I—laboratory tests. Ear Hear. 2010;31:22-34.
- Wu Y, Bentler RA. Impact of visual cues on directional benefit and preference: Part II—field tests. Ear Hear. 2010;31:35-46.
- Walden BE, Surr RK, Cord MT, Dyrlund O. Predicting hearing aid microphone preference in everyday listening. J Am Acad Audiol. 2004;15(5):365-96.
- Groth J, Laurenyns M, Piskosz M. Double-blind study indicates sound quality preference for surround sound processor. Hearing Review. 2010;17(3):36-41.
- Keidser G, Convery E, Volkmar H. The effect of gain mismatch on horizontal localization performance. Hear Jour. 2011;64(2):26-33.
- Groth J, Laureyns M. Preserving localization in hearing instrument fittings. Hear Jour. 2011;64(2):34-38.
- Cord MT, Surr RK, Walden BE, Dyrlund O. Relationship between laboratory measures of directional advantage and everyday success with directional microphone hearing aids. J Am Acad Audiol. 2004;15:353-64.
- Bentler RA, Egge JLM, Tubbs JL, Dittberner AB, Flamme GA. Quantification of directional benefit across different polar response patterns. J Am Acad Audiol. 2004;15:649-659.
- Cord MT, Walden BE, Surr RK, Dittberner AB. Field evaluation of an asymmetric directional microphone fitting. J Am Acad Audiol. 2007;18:245-256.
- Mackenzie E, Lutman ME. Speech recognition and comfort using hearing instruments with adaptive directional characteristics in asymmetric listening conditions. Ear Hear. 2005;26:669-79.