Regional and Jet Transport Pilot
Active Noise Reducing Headset Study:
A Report from the Line Pilot
Michael S. Hayes
Human Factors and Ergonomics Committee
August 1998
(Revised March 1999)
Michael D. New, Ph.D. and Michael S. Hayes
ALPA Human Factors and Ergonomics Committee
The first patent for an active noise cancellation headset was granted to Dr. Paul Leug in 1936. It was not until the early 1980s, however, that the technology was available to make a truly portable unit capable of actively reducing noise. Now all major manufacturers of aviation related communication devices feature at least one active noise reduction (ANR) headset. With so many choices available, your Human Factors and Ergonomics Committee decided to investigate the issues related to ANR technology and to conduct an informal survey where exemplar units would be tested by you—the line pilot.
The goal of this article is not to endorse any particular headset. Our intent is to familiarize the reader with the underlying concepts in the area of sound production, measurement, perception, and attenuation. We thought that you might also enjoy a review of some of the leading headsets and to hear what your colleagues had to say about those units. We decided to use a questionnaire because we wanted to have as many reviews as possible by actual pilots in real world operations. In the following sections, we will describe the nature of sound, its measurement, basic ANR theory, and issues relevant to its application. The methodology used in the study is then described followed by the questionnaire results.
Sound Production and Measurement
Normally, sound begins as a pressure wave that radiates outward from the source much like the ripple caused by a small stone being tossed into an otherwise calm pond. This "wave" consists of a period of increasing pressure (more air molecules) followed immediately by a decreasing pressure segment. If plotted, this pressure fluctuation would approximate a sinusoidal function. The number of air molecules (and consequently the pressure) involved is a direct result of the force of the mechanism causing the disturbance. For instance, if a person yells, they expel more air than if they whisper. The greater number of air molecules involved translates into greater pressure acting upon the receiver’s eardrum and consequently the yell is perceived as being "louder" than the whisper.
The intensity of the maximum pressure of each wave is measured by comparing the present value with a standard reference value (representing the minimum pressure the human ear is capable of perceiving) and conducting a log transformation on the ratio. These units are called decibels and are indicated by the abbreviation "dB." An increase of 10 dB translates into a perceived doubling of the loudness of the sound. Common examples of noise levels include: a whisper roughly equates to 30dB, normal conversation occurs at about 60 dB, a lawnmower operates at 90 dB, and a jackhammer emits 120 dB of noise. It is generally agreed that long exposure to noise greater than 85 dB can be damaging to your hearing. Fortunately for most workers in America, there are federal standards set forth by the Occupational Safety and Health Administration (OSHA) that regulate the duration of exposure to various noise intensities. At 90 dB the limit is set at 8 hrs per day. At 100 dB the limit is 2 hrs and at 110 dB the worker is only allowed to spend 30 minutes per day in this environment without hearing protection. Noise levels for airline pilots, however, do not fall under OSHA’s jurisdiction. Instead the FAA maintains control over all aspects of the design and manufacture of aircraft (40 CFR, 1975). What is the official FAA standard related to cockpit noise? Simply, "Vibration and noise characteristics of cockpit equipment may not interfere with the safe operation of the airplane." (14 CFR 25.771, 1965).
In addition to the pressure of a sound, the distance between the maximum values of the sinusoidal function can vary. This is commonly referred to as the wavelength of the sound. The number of cycles of the wavelength per second is called the frequency and is expressed in hertz (Hz) or kilohertz (kHz). Humans can perceive sounds from 20 Hz to 20,000 Hz (or 20 kHz) and subjectively refer to the variations as differences in pitch. Voice communications occur in the 100 Hz to 8000 Hz range. When searching for data describing typical frequencies and intensities of noise in the cockpit of a commercial airliner, we were surprised that those data are not readily available. The only study of the sound pressure as a function of frequency we could find was for the KC-135 aircraft (Farinacci, 1975). Those data are presented in Figure 1.
Perception of Sound
The eardrum flexes in response to the frequency and intensity of the sound pressure. These vibrations are transmitted to the inner ear via three small bones collectively referred to as the ossicles. The fluid filled inner ear contains the organ of Corti. Inner and outer hair cells (cilia) are attached to the organ of Corti. It is the cilia that translate the vibration sensed in the inner ear into electrical energy that makes its way to the brain via the auditory nerve. Although there is some debate as to the mechanics involved, it is generally accepted that prolonged exposure to loud noise (particularly in the higher frequencies) leads to damage of the outer cilia and, therefore, loss of auditory acuity. There are other problems that have been linked to exposure to loud noise (irritability, fixation, and fatigue), but the major threat to our health is related to hearing loss. Recently, our ability to counter this threat has been advanced through the use of active noise reduction (ANR) headsets.
Sound Attenuation
There are two techniques for reducing the amount of noise in the cockpit--passive and active attenuation. Passive attenuation refers to providing a physical barrier to the sound waves. Earplugs and full-cup headsets take advantage of this technique. Approximate passive attenuation values for earplugs are 5 – 30 dB below 1000 Hz and 30 – 40 dB above 1000 Hz. Headsets provide noise reduction of less than 20 dB below 1000 Hz, but are very efficient above 1000 Hz (25 – 40 dB). The most important aspect of passive noise attenuation to note is that headsets alone provide little protection in the lower ranges of the frequency spectrum. To protect against those lower frequencies, the operator would have to use earplugs, but this would limit his ability to perceive most voice communications. Thanks to modern technology, ANR headsets are capable of protecting against unwanted lower frequency noise while allowing communications to continue.
The typical Active Noise Reduction headset measures the noise near each ear via small microphones. This information is then fed into the circuitry and a signal is calculated describing the prevalent lower frequency noise. The ANR unit then generates an opposite phase signal of the same magnitude calculated from the ambient noise and sends this to a speaker located in the earcup. When this new sound is combined with the ambient noise, they cancel each other and silence results. There are three hardware variables that can influence the ability of the headset to perform this task. First, microphone location is a key factor. Optimally, the microphone should be located as near as possible to the entrance of the ear canal. Second, the electronics of the unit should be of sufficient quality to analyze the incoming sound and to generate an adequate out-of-phase tone. Third, the speaker used to fill the headset with the "anti-noise" sound must be able to generate a faithful reproduction of the tone received from the electronics. Our intent, however, was not to measure the relative merits of the hardware (although we have provided a synopsis of the technical specifications). In this study, we wanted to focus on the "usability" of the headsets and to offer insights that might make your purchase easier, therefore, the questionnaire reflected that perspective. The results will be presented, but first, a quick introduction to how the study was conducted.
The Study
WHO were the evaluators? They were pilots just like you. We asked 42 pilots from various airlines to participate. It appeared that this was a hot topic given the enthusiastic response from the line. As you can imagine, there is a vast difference in the type and magnitude of noise in different cockpits, therefore, we used data from 11 different aircraft types.
WHICH headsets did we evaluate? We asked every major manufacturer of ANR headsets to participate. Three manufacturers agreed to supply us with examples of their products. We used three different models from Sennheiser Electronic Corporation, three models from Telex Communications, and one model from LightSPEED Technologies.
HOW were the headsets evaluated? Each of the volunteers was asked to evaluate two to four different models. Typically, they would accumulate at least 20 hours on each headset and then complete a short questionnaire. At the end of the study, we collected 66 completed surveys from regional turbo-prop pilots and 57 surveys from pilots at major air carriers.
The survey consisted of general instructions and four blocks of questions. In the first block, simple identification and tracking information was gathered. The second block asked the participant to evaluate the sources of extraneous noises in their cockpit. The third block asked questions related to comfort and perceived performance of the unit. A blank forth section was included to allow the evaluator the opportunity to comment on areas not addressed by the previous questions.
These data were then analyzed to determine if statistically significant differences existed between the groups. The responses for each question are reported in the graphs below by airline type (i.e., Regional and Major). The following icons were used to indicate the respective rating and were based on the mean response for each group:
Following the tables, we discuss the major findings of the study and summarize the most common responses from the "comments" section of the questionnaire.
Specifications: (As supplied by manufacturer)
(S = Sennheiser, T = Telex, and LS = Lightspeed)
(S = Sennheiser, T = Telex, and LS = Lightspeed)
(S = Sennheiser, T = Telex, and LS = Lightspeed)
Discussion
We conducted the study in two parts. The first part surveyed the pilots from regional airlines; the second concentrated on pilots from major air carriers. The number of headsets available for the majors was limited because of time constraints and the questionnaire form used by the majors was modified to take advantage of the lessons learned during the initial phase. Since there were differences in the two parts, we did not collapse the data across airline types. This fact aside, we feel the following comments accurately reflect input received from the participants and may assist you in your purchasing decision.
There were two dominant themes that evolved from these data. First, the pilots from regional aircraft placed a high priority on noise reduction. They preferred the headsets that reduced the most noise (both actively and passively). Size and weight were not as important to these pilots as was durability and quality of construction. Of the headsets tested, the regional pilots preferred the "ANR" by Telex (T-ANR&BATT, part number 70800-100). This particular headset was shown to be statistically better than other headsets in several categories (clarity of earphone, quality of construction, active and passive noise reduction). With an overall rating of 4.3 out of 5, it was not surprising that this headset also received the highest mean score in all but one category (the pilots rated the Sennheiser HMEC-200 better at passive noise reduction).
The second theme revealed by these data was that pilots of major airlines were more concerned with portability issues than passive attenuation of noise. In a quieter cockpit without a "hot" intercom system, these pilots seemed to view the bulkier headsets as providing too much attenuation. Several pilots commented that they had to pull the earphone off of one ear just to hear the requests of the other pilot, thus defeating the purpose of the headset. Of the four headsets tested by these pilots, there were no differences reported between the Sennheiser HMEC25K, HMEC45KA, or the Telex ANR-200 (at 567 grams the Telex ANR-4100 was rated as less portable, heavier, and larger than the others). Although these pilots seemed to prefer a smaller headset, there were several "WOW" comments from first time users of this technology. When asked if they would use the headset if provided by their company, they generally answered yes (particularly for the lighter units such as the Sennheiser HMEC45KA). When asked if they would purchase one of the units reviewed, these pilots felt the headsets were still too expensive and said, "no."
There were no differences between the pilots from different airline types with respect to comfort of the headset. Generally, the longer the headset was worn the less comfortable it became and the lighter headsets were more comfortable with one notable exception. The Lightspeed Technologies 20K headset was a favorite among the pilots that tested the unit. They gave the headset the highest rating for comfort through two hours of use. Only the Sennheiser HMEC45KA was close to the 20K in the comfort category. (As an aside, Lightspeed recently released an improved version of their "20K" headset. It appears they have made serious progress toward addressing those categories that were rated as "poor" during this evaluation. Their willingness to modify the existing design demonstrates the responsiveness of the industry to pilot input. If you have a suggestion, therefore, let them know.)
In conclusion, our results show that pilots of regional airlines prefer headsets that attenuate as much noise as possible while pilots of aircraft typically found at major airlines prefer small headsets that allow intra-cockpit communication while attenuating low frequency noise. An important aspect of this investigation has been to identify the lack of accurate noise data for airline cockpits. Your Human Factors and Ergonomics Committee has already taken the first steps in attempting to fill this void. We will keep you posted on our progress. Until then, wear your hearing protection and continue to guard yourself from long exposure to high frequency sounds.
References
Farinacci, N. A. (1975). USAF Bioenvironmental noise data handbook, vol. 32. (AMRL TR-75-50-32). Wright-Patterson AFB, Ohio: Aerospace Medical Research Laboratory.
Occupational Safety or Health Standards for Aircraft Crewmembers, 40 Fed. Reg. 17859 ( July 10, 1975).
Pilot Compartment, 14 CFR 25.771 (1965).
