Dr. Ross on Hearing Loss
RERC Assistive Listening Devices Project
by Mark Ross, Ph.D., Matthew Bakke, Ph.D.
It is an audiological “truism” that hearing-impaired people frequently exhibit a disproportionate amount of difficulty understanding speech in the presence of noise and reverberation (Ross, l992; Nabalek, l994). Acoustical conditions that may be only mildly troublesome to people with normal hearing can virtually eliminate speech comprehension for people with hearing loss. An example of this would be in large area listening situations where desired speech signals are weakened by distance, and contaminated by noise and reverberation, before arriving at the listeners. Many hearing-impaired people simply “give up” in anger or disgust rather than expose themselves again and again to another unsatisfying and frustrating listening experience.
Under optimal acoustical conditions, however, people with hearing loss often function quite well. What assistive listening systems (ALS) do is permit them to fully utilize their residual auditory capabilities. It does this by eliminating or reducing the deleterious effects of distance, noise, and reverberation upon their speech perception capabilities. An ALS can make it possible for a hearing-impaired person to more fully participate in, and benefit from, the many social and cultural activities offered by our society. Without such a system, many people with hearing loss withdraw and feel isolated from their normal milieu. We should not, therefore, underestimate and undervalue the potential significance of these devices for people with hearing loss.
Unfortunately, while ALS are required by the Americans with Disability Act to be installed in many venues, there have been frequent problems associated with their use. It was in response to consumer complaints that the U.S. Access Board, an independent federal agency whose primary mission is ensuring accessibility for people with physical and sensory disabilities, funded a project to examine ALS. The project was conducted by the Rehabilitation Engineering Research Center (RERC) at the Lexington Center. Specifically, the RERC was charged with writing a “state of the art” paper and completing a research project for the purpose of recommending electroacoustic performance standards for large-area ALS to the Access Board. This project is now completed and our final report has been submitted to the Access Board.
This paper can be considered an extensive summary of our final report. In it, we will (1) describe the three ALS now being used and major issues concerning them, (2) review the recommended measures to address consumer concerns, and (3) present the electroacoustic performance standards that emerged from the study. The complete report can be requested from the Lexington Center RERC or accessed on our website (hearingresearch.org).
In the course of the project, three focus groups were convened. In the first one, fifteen knowledgeable consumers were invited and asked to share their experiences with ALS. Participating in the second group were a number of representatives from manufacturers, installers, and providers (theatre and movie houses) of these systems. In the last focus group, all the previous participants were invited, plus several additional consumer advocates. For visual access, a court stenographer provided real-time speech-to-text captioning at all three meetings, while infrared and induction loop systems supplied auditory access. The captions were stored on hard copy and the complete drafts were reviewed prior to writing the final report to ensure that no comments and suggestions were overlooked.
Types of Assistive Listening Systems
There are three general types of ALS and we will briefly review some of the highlights of each of them below. The differences between the systems pertain to transmission mode, installation variables, types of interference, and the appropriateness in specific venues rather than to their relative listening advantages. Essentially, what these systems do – all of them – is increase the speech to noise ratio (S/N), undoubtedly the most effective measure that can be taken to improve speech perception. They do this by bridging the acoustical space between the source and the listener, either electromagnetically or with light or radio waves. When installed properly, they all do what they purport to do, i.e. improve speech perception performance compared to that obtained with or without hearing aids. Studies have shown that performance scores between the different types of large-area listening systems are insignificant compared to the differences between any of them and sound signals generated by a PA system (Bankoski & Ross l984; Nabelek, Donahue & Letowski l986; Nabalek & Donahue l986; Noe, Davidson & Mishler l997).
Induction Loop (IL) Systems
IL systems are the oldest and least employed large area listening system. In this country, they were initially used in schools and classrooms for deaf children in the middle l960’s to the early l970’s before being supplanted by personal FM systems. In an IL system, the output from an amplifier is led to a coil of wire placed around the circumference of a designated listening area. The electric current in the wire produces an electromagnetic field around the wire and targeted to remain within the looped area. The telephone coil in a hearing aid detects this electromagnetic field in exactly the same way it responds to the magnetic field generated by hearing aid compatible (HAC) telephones. The output from the telephone coils is processed by the hearing aid in exactly the same way as are microphone signals. Either live voice or recorded audio signals can be accommodated as the input to an IL system.
Once properly installed – and this can be a major qualification for all ALS – an IL system is undoubtedly the most convenient and cost-effective ALS. To hear the audio signals, all a person has to do is enter the looped area and switch his/her personal hearing aids to the telecoil position. As long as a person’s hearing aids include telecoils, he or she always has an assistive device “receiver” available. There are no US standards that pertain to the installation of IL systems, although there is a European standard (IEC, 118-4, l981). In this country, Oval Window Audio, a manufacturer of audio loops, has extended and refined these standards in their recommended installation guidelines (see the RERC report for details). Two key considerations govern the installation of an IL system. The first one requires that ambient electromagnetic field be low enough to prevent inference with the reception of the desired electromagnetic signals, while the second one specifies that the desired electromagnetic field reach a target intensity levels within the entire looped area.
Obviously, an IL system requires that the user’s hearing aid include a telecoil. At the present time, however, only about 30 percent of modern hearing aids contain telecoils. With the trend toward smaller and smaller hearing aids, this percentage is (unfortunately, from our perspective) unlikely to increase. Therefore, the use of large area IL systems will probably continue to be limited. While there are special receivers that can be used with an IL system, this defeats the primary advantage of such systems – their convenience.
The telecoil response of a hearing aid may not duplicate the microphone response (Thibodeau & Abrahamson l988). This is less of a issue with programmable hearing aids or those that employ amplified telecoils (Noe, Davidson & Mishler l997). The physical orientation of the telecoil within the hearing aid will also vary the aid’s response. For optimal sensitivity to an induction loop, either floor or neckloop, the telecoil should be mounted perpendicular to the loop. This differs from the optimal (horizontal) orientation required for telephone listening (Preves l994).
Even though a hearing aid may contain a telecoil, there are times when it is desirable for the person to also be able to perceive a signal from the hearing aid microphone. This requires the capability to receive both microphone and telephone coil input simultaneously, either with an M/T switch position or a programmable option. Such capability would be particularly appropriate when a neckloop is being used with an infrared or FM receiver (see below).
In an IL system, it is difficult to confine the electromagnetic field within the looped area. Some of the energy “spills over” into adjacent areas, both horizontally and vertically. Also the intensity of the signal within large looped areas often varies. The further a person is from the loop, the weaker the signal picked up by the telecoils. Over the years, a great many creative loop configurations have been used in an effort to circumvent this problem. The most successful effort that we know of is the “3-D” loop developed by the Oval Window Company (Lederman & Hendricks l994). In the 3-D loop, four wires configured in a precise geometric pattern are embedded in a mat placed on the floor. Reportedly, the resulting elecromagnetic signal is not only contained within the looped area, but the 3-D pattern of the electromagnetic field reduces the impact of the telecoils orientation within the hearing aid.
Frequency Modulated (FM) Radio Systems
Large area FM systems work on the same principle as personal FM devices, but differ in that they are designed to be used in such venues as auditoriums, theatres, houses of worship, etc. Both types are basically FM radios in which the audio signal is broadcast to listeners wearing receivers tuned to the transmitting frequency. They can be utilized as “stand-alone” devices or connected to the amplifier in a Public Address system. They employ the same radio frequencies within the 72 to 76 MHz FM band as do personal systems. The FCC requires that they be “low-power” devices, with maximum power no more than 80 millivolts per meter measured at three meters from the antenna. This is ordinarily sufficient to provide an acceptable signal throughout all but the largest venues. Other users, such as pagers and emergency vehicles utilize the same frequencies but are not limited by the same power restrictions. When interference occurs with an FM system, the onus is on the ALS user to find a solution (like switching to a different frequency).
ALS manufacturers differ in how they allocate this band, using a varying number of narrow-band channels (50 kHz) and wide band (either 150 kHz or 200 kHz) channels. Some of the frequencies used by the different manufacturers may be identical, while others differ. In recent years, the FCC has also permitted ALS manufacturers to use the 216-217 MHz band as a low power source for auditory assistive devices (both personal and large-area). The specific allocation of bandwidth and carrier frequency within this band also differs between manufacturers. Except for power levels and the necessity to broadcast within the permitted channels, there are no requirements for companies to abide by any universal transmitting characteristics in an FM ALS. Transmitters of the different companies range in complexity and in the pre-processing strategies they employ (such as high frequency pre-emphasis and various compression options). While the rationale for employing a particular pre-processing strategy may appear to be convincing, there is no independent evidence that we are aware of that supports one over the other in listening tests conducted with hearing-impaired subjects.
The major issue with an FM ALS is that radio signals are not contained within the facility. Those generated within a venue are transmitted outside the physical confines of a facility, while at the same time radio signals from the outside penetrate the facility. In both instances, this can cause a problem. When signals leak from a facility, they may interfere with other authorized users in the same band as well as compromise privacy. Those that enter the venue may interfere with the quality of the signal reception. The possibility of such interference is a key consideration when the installation of a FM ALS is contemplated.
All FM receivers are basically FM radios. Manufacturers provide FM receivers that vary in complexity and secondary features, but all are designed to accord with the unique design characteristics of their own transmitters. This makes using receivers from one company with the transmitter of another a problematical situation, even when both are tuned to the same frequency. Generally, except for the possibility of using a “universal receiver” (see below), such an interchange is probably not a good idea.
The receivers for an ALS FM system are all body units, meant to be used either with earphones or coupled to personal hearing aids. Those people whose hearing aids do not include a “T” coil can use earphones or earbuds. Acoustical coupling with earphones is ordinarily feasible for CIC hearing aid users; i.e. the earphones can be placed right over the ears. Acoustical coupling is also possible for many ITC and ITE hearing aid users. With these hearing aids, however, there is no way to predict in advance whether an individual can couple satisfactorily without acoustical feedback.
For those whose hearing aids include a telecoil - appropriate in our judgement for anybody whose hearing loss is moderate or greater - inductive coupling is a convenient way to access an FM ALS. (This is advantage of telephone coils that should be considered during the hearing aid selection process.) Inductive coupling is usually accomplished with a neckloop, though for some people with the most severe hearing losses, silhouette inductors would be appropriate. Stereo reception is not possible with a neckloop, since the neckloop is basically a single output transducer; stereo reception is possible with silhouette inductors, providing an appropriate adapter in plugged into the earphone jack of the FM receiver.
Receivers can be powered by either disposable or rechargeable batteries. Battery life for the rechargeable batteries range from 6 to 10 hours (one report claims as much as 35 hours). Convenient pocket recharging-carrier cases permit the receivers to be recharged while being stored. The reported life span of the disposable batteries range from 18 to 70 hours, depending upon volume setting and type of coupling. Unlike many personal FM systems, those with an ALS do not incorporate a low battery warning light.
Infrared (IR) Systems
An IR system transmits audio signals via invisible light waves at frequencies between 700 nm to about 1000 nm. Audio signals are used to frequency modulate an RF sub-carrier that in turn amplitude modulates the IR carrier (Lieske l994). What results is a double modulation of the IR light wave, first FM and then AM. The bandwidth of the IR carrier is usually about 50 nm wide, thus permitting a number of RF sub-carriers to be simultaneously carried by the same IR light wave (useful for such applications as simultaneous translation into different languages or to receive a stereo signal).
All IR systems are composed of three basic components: the transmitter (also called the modulator), the emitter and the IR receiver. Before being emitted, the signal can be pre-processed similarly to the ways that FM transmitters process signals. The actual light waves are delivered by an emitter composed of a number of light emitting diodes (LED’s). Although IR light waves are invisible to the naked eye, they are light waves and thus display the same characteristics as other light sources. This fact explains many of the issues and advantages relating to IR systems, such as the effects of direct sunlight and the impact that the color and texture of room surfaces will have upon the IR reflections. It is because of the reflecting properties of some wall surfaces, coupled to the increasing power of IR transmitters, that strict “line of sight” transmission, long thought to be a limiting characteristic of IR systems, may no longer be applicable in many locations. In darkened theatres, however, line of sight limitations would still be a major consideration. Whatever the room surfaces and however a room is configured, the light waves are contained with the room enclosure - the major advantage of IR systems over IL and FM systems.
The transparent lens found on every IR receiver contains the photo detector diode that detects the IR light wave. The receiver then demodulates the RF sub-carrier, and the audio signal is retrieved, processed, and delivered to the transducers. One reason why IR systems of different companies are not always completely compatible, even though they may employ the same sub-carrier frequency, is that they can differ in many other respects, such as the electrical selectivity and filter characteristics of the receiver, the nature of the transmitter pre-processing strategies and the resulting compensatory receiver characteristics.
IR receivers come in all shapes and sizes, from units that dangle under the chin, large self-contained headphones (incorporating amplifier and volume controls), to body receivers similar to those used with FM systems. Neckloops and silhouette inductors can be plugged into these latter units. While some under-chin receivers include an output jack to permit inductive coupling, these are generally micro-mini plugs that will accept the usual neckloop. Some IR receivers include an environmental microphone to permit side conversations (conceptually similar to an M/T position). One recently introduced IR receiver includes a low-battery light feature. Battery life and characteristics are similar to those obtaining with FM systems.
Until recently, 95 kHz has been the sub-carrier frequency most often used by manufacturers. When stereo reception was required, in other than large area listening situations, a sub-carrier of 250 kHz was simultaneously employed. This unofficial standard no longer applies. Other sub-carrier frequencies are now being used (300 kHz, 2.3 MHz and 2.8 MHz). One classical advantage of IR ALS was that the same IR receiver could be used in a number of large venues (as well as for IR TV listening devices). The reason given for the necessity of moving to a higher sub-carrier frequency is the electromagnetic interference (EMI) at 95 kHz produced by the newly introduced T-12 fluorescent ballasts. One manufacturer asserts that filters on their newly introduced IR receivers can eliminate EMI at 95 kHz, but this has not yet received independent verification. Since it is unlikely that mandatory standards regarding a specific sub-carrier frequency will be promulgated for ALS, it is clear that we will be faced with the reality of a increasing number of sub-carrier frequencies in the future.
A recurrent complaint by consumers was the uncertainty of reception while using an IR system in a public place. Consumers frequently observed that at some seats in an auditorium, the IR signals were either absent, weak or distorted. Unlike FM systems, there are no formal power requirements for the IR transmission. While with proper installation, the units provided by manufacturers are capable of providing excellent reception throughout a facility, this may not be realized because of a faulty installation. Given the proper placement and power of emitters, however, it is perfectly feasible to ensure an adequate IR signal at all locations in a facility, no matter its geometry. It is apparent, however, that the proper installation of an IR system is not something that can be taken for granted. Clearly, it takes focus and skill. Furthermore, as pointed out by the installers and manufacturers in our focus groups, even when an IR system is properly installed, local facility managers may make modifications for reasons of their own (like covering emitters with drapes, for example).
Addressing Consumer Concerns
In the comments below, we will summarize the concerns and recommendations offered by the consumers during the focus groups. Most are clearly common sense observations (unfortunately, too often in short supply). Most of these points came up time and again, attesting to their personal significance for people with hearing loss who would like to continue to attend and profit from various social and cultural events.
People have to know that ALS are available before they attend some event. This can be easily accomplished by requiring that all media advertisements (newspaper, TV, radio) and telephone announcements note that ALS are available. These notices should be of the order of visibility or prominence as other attributes of a performance (such as “Dolby” sound, etc).
Upon their entry into a facility, there should be clear signage indicating the precise location where personal receivers can be obtained. The sign should include information on the transmission characteristics of the FM or IR systems (this would be important for people who possess a personal ALS receiver).
The same person, in the same physical location, should be responsible for checking the receivers in and out. This person’s responsibilities must include (1) verifying that all receivers are functioning appropriately before they are checked out, (2) helping people select an appropriate coupling arrangement (neckloop, headphones, earbuds), and (3) briefly instruct recipients in the operation of the receiver when necessary. After the performance, this person should take the appropriate hygienic measures (spraying or wiping foam cushions with an antiseptic solution and replacing disposable cushions).
Receivers and Coupling Arrangements
The Americans with Disability Accessibility Guidelines (ADAAG) issued by the Access Board specifies the minimum number of receivers according to a sliding scale. In facilities with seats of 500 or less, the requirement is that the number of receivers should total 4% of the total number of seats. Using a sliding scale, the requirement is 3.5% for facilities with 501 to 1000 seats and 2.75% for places with 1001 to 2000 seats. The revised ADAAG guidelines recommends that a portion of the receivers, (25% but no less than 2) be compatible with hearing aids equipped with telecoils.
As pointed out above, many people with hearing loss must (because their hearing aids do not include telecoils) prefer to place headphones right over their hearing aids. One of the recommendations made by the consumers was that headphones be provided that can comfortably fit over all types of ear-level hearing aids (ITE, ITC, CIC). These headphones should permit users to either couple acoustically or inductively (permitting stereo reception in the latter instance). Furthermore, the fit must be such that “bleed” from the earphones not exceed the ambient noise levels at adjacent seats.
There are no universal standards applicable to the transmission characteristics of either FM or IR signals. The only formal limitation to FM ALS are that they be low power and that the carrier frequencies fall within the permitted channels. IR systems do not have even these limitations. In practice, venues can choose from a large number of FM (narrow and wide band) and IR sub-carrier frequencies. This effectively prevents consumers from purchasing a personal ALS receiver, something that many prefer to do for a number of reasons (hygienic, personal adjustments, assurance of proper operation, etc.). Until quite recently, a “universal receiver” had been available, one that consumers could purchase and employ in just about any facility using any type of ALS (IL, FM, or IR). This universal receiver is no longer being manufactured.
This is an option that should now be resurrected and available to consumers. In our judgement, such a receiver should be (1) tunable to any FM frequency in the 72-75 MHz or 216-217 MHz range, either wide or narrow channel, (2) permit the reception of any of the IR sub-carrier frequencies now being employed with such system, and (3) include a telecoil for IL reception. Other desirable features are environmental microphones, a “low battery” indicator, and tone controls. These are just basic recommendations; we have no doubt but that creative manufactures could add and refine these suggestions for the further benefit of consumers.
Developing Performance Standards for ALS
In developing performance standards, RERC researchers decided at the outset to focus on the last stage of the transmission process, i.e. the signals actually being delivered to listeners through ALS. The input could be an audio track or a live speech signal. By comparing the input to the output, all the variables and factors that ALS can impose upon the quality of signals would be subsumed. These would include type of ALS, whether the signals to the ALS are derived from microphone or recorded inputs, types of microphones and listening environment, the distance between the talker and the microphone, the quality of the various components, and the nature of the many interconnections between the components in the transmission process. This approach provides manufacturers and installers with a performance goal without dictating how they arrive at this end point. The primary metric used to define the quality of the output signal was the Speech Transmission Index (STI). Basically, what the STI does is compare the integrity of a signal at two points (in this instance, at the input and the output). It does this by measuring the “fill” between adjacent peaks in a simulated or actual speech envelope. Since this fill represents the addition of noise and reverberation to the basic signal, the more the fill, the lower the STI. An STI of 1.0 would mean a perfect reproduction of the input signal.
Subjects and Test Conditions
Fifty-nine adult listeners participated in the study, ten of whom had normal hearing. The hearing-impaired group was divided into six groups according to the degree and configuration of their hearing loss. These ranged from moderate flat hearing losses to precipitous high frequency hearing losses. Sentence stimuli were recorded under different types of either live or computer simulated listening conditions. Three types of distortions were created: (1) reverberation plus noise recordings at different S/N ratios in a classroom, an auditorium, and a conference room, (2) digital recording of noise created by a poorly installed induction loop at different S/N ratios and, (3) noise created by different degrees of symmetrically peak clipping.
The subjects listened binaurally through TDH 49 earphones to sentences recorded under the condition noted above. They were asked to judge the quality of the recorded signals using a four point scale: excellent # 4, good # 3, marginal # 2, and unacceptable # 1. A “minimally acceptable” listening rating of 2.25 was chosen. In the study, this figure would represent an average of the four presentations each subject received in each condition, in which three judgements were rated as “marginal” (a # 2 rating) and one as “good” (a # 3 rating). It should be emphasized that this a “minimally acceptable” level, one that we would hope and expect to be exceeded in the real world. The subjects were asked to rate four listening conditions: (1) minimally acceptable output levels, (2) minimally acceptable Speech Transmission Index (STI), (3) minimally acceptable S/N for internally generated noise from induction loops and (4) minimally acceptable peak clipping levels.
Results and Recommendations
1. Since except for the most severe group, all the hearing-impaired subjects made similar ratings only the average results for the six groups will be reported here. (Details can be found in the original report.) In making our recommendations, we have used the percentile point where 75% of the subjects exceeded 2.25 or greater in their ratings for any specific listening condition. Ratings and STI for the auditorium condition are shown in Table 1.
Table 1. Listening ratings and the STI at the seventy-fifth percentile rating point in the auditorium condition.
What these results show is that the criterion listening rating of 2.25 can be reached by 75 percent of the subjects only at three feet from the source in this particular auditorium. Beginning at 6 feet, the ratings of the sentences fall between unacceptable and marginal. Of all the results in this study, this is the one that clearly demonstrates the need for people with hearing loss to utilize an ALS in a large-area listening situation. The auditorium in which the study was carried out is considered to be a “good” auditorium (located within the Lexington School for the Deaf). If this kind of result occurred here, similar or worse findings would undoubtedly be apparent in other such venues.
2. The ratings and STI for the classroom condition can be found in table 2. The criterion listening level of 2.25 or higher can be met only somewhere between two and four feet from the talker. The degradation of speech signals as distance from the source is increased is also apparent in these results, clearly supporting the necessity of an assistive listening device in this type of situation. Over the years, a number of objective studies has demonstrated the inadequacy of the listening environment in classrooms for children; the results of this study now add the subjective judgement by adults to this body of information (Crandall, Smaldino & Flexer l995).
Table 2. Listener ratings and the STI at 75 percentile point for the classroom condition.
3. The ratings and the STI for the conference room condition is shown in table 3. As is evident, even at three feet from the source, the signal received by listeners in this venue is clearly unacceptable. Probably, these extremely poor results can be attributed to the existence of a permanent hum from the ventilation system in this particular conference room - a not uncommon situation in such rooms. What we would advise here is the use of some sort of conference microphone, i.e. a small-scale ALS.
Table 3. Ratings and the STI at the 75 percent level for the conference room condition.
4. In order to arrive at the recommended STI criterion, the 75 percentile ratings for the above three listening conditions were averaged and plotted relative to the measured STI at that point. This analysis showed that at an STI of .84 only a few of the people from the more severely hearing-impaired groups failed to reach the 2.25 listening rating level. What this indicates is that the severity of these subjects’ hearing loss precluded a satisfactory listening experience, not it is important to note that they would be unable to benefit from an ALS. In other words, without an ALS their performance is likely to be even poorer. At an STI of .84, all the other subjects achieved the minimally acceptable listening rating.
5. Three other general types of listening conditions were evaluated in this study: minimally acceptable SPL output levels, minimally S/N in an IL listening condition, and minimally acceptable peak clipping levels. These results hold no surprises:
Measuring the Speech Transmission Index
RERC engineers developed a simplified software version of the original STI first developed by Houtgast & Steeneken in l973 and since used extensively to define the acoustical conditions of enclosures. The RERC version can be used to measure the STI in large-area listening situations where the input is either from a live microphone or from an audio track. In the case of an assistive listening system using a live microphone, a test loudspeaker is placed at the location of the talker, performer or sound source. The output of the STI measurement system is broadcast from the test loudspeaker, picked up by the microphone, and passed on to the assistive listening system to be measured. In the case of an audio track input to the ALS, the audio track is simply replaced by the line output of the STI measurement system. In either case, the output of the ALS is monitored either by means of a line output from the ALS, or through a coupler of some kind (e.g., Zwizlocki coupler). The output from the assistive listening system is connected to the STI measurement system through the line input port of the computer’s sound card. The person testing the system can then run the software and in about three minutes, the STI measurement is complete.
It is important to note that the STI is applicable in all types of listening situations, not just with large area ALS. It can be employed in any size room, using any type of listening system, in such places as small meeting rooms or courtrooms. It can even be used in a completely live-voice situation, simply by comparing the speech signal at a talker’s mouth (the input source) to that recorded at any point in a room (the output). Given an STI of at least .84, one can be assured that the signals received by a listener are minimally acceptable. In our judgement, all facilities should attempt to exceed this minimal figure as well as the other electroacoustic recommendations derived from the results of this investigation. The STI software developed by the RERC can be obtained through our website (http://www.hearingresearch.org), or by calling the Research Department at the Lexington School for the Deaf (718 350-3200).