WO2010057267A1

WO2010057267A1 - Adaptive hearing protection device

Info

Publication number: WO2010057267A1
Application number: PCT/AU2009/001518
Authority: WO
Inventors: Andrew Peter Bradley; Wayne James Wilson; Ian O'brien
Original assignee: The University Of Queensland
Priority date: 2008-11-21
Filing date: 2009-11-20
Publication date: 2010-05-27

Abstract

A hearing protection device (30) and a method of providing hearing protection by adaptive noise reduction. Noise is sampled at or near the ear of a user using a microphone (31) and a cancelling noise is produced and directed into the ear of the user by a non-occluding speaker (32). The noise reduction occurs by destructive interference between the incoming noise and the sound pressure wave produced by the speaker (32).

Description

TITLE

ADAPTIVE HEARING PROTECTION DEVICE FIELD OF THE INVENTION

The present invention relates to the field of hearing protection and in particular to a method of reducing, controlling and monitoring the noise exposure in situations where ambient sound has a large dynamic range and a user needs to maintain fidelity of hearing. A typical application is hearing protection for orchestral musicians without affecting their, or the orchestra's, ability to play. BACKGROUND TO THE INVENTION

Noise induced hearing loss (NIHL) in the workplace is a well- documented phenomenon that causes physical and psychological problems to those afflicted as well as economic damage to workplaces. Noise induced hearing loss is a gradual process and is often not noted until the damage is done. During the last few decades, both musicians and hearing specialists have become increasingly aware that both popular and classical music have the potential to produce NIHL. Auditory acuity and sensitivity are, of course, especially important to musicians and even a subtle deficit may detract from the perfection of a performance. In extreme cases, severe hearing loss could mean an end to a musician's career.

Management of this problem within the workplace is often confounded by a worker's need to clearly perceive their acoustic environment in order to carry out their job to a satisfactory level. This is especially true within the performing arts where musicians - often working in hazardous noise levels - are unable to adequately protect their hearing due to the need to hear themselves and their colleagues with great clarity. Opera, ballet and symphony orchestra musicians are particularly challenged by this problem. It has been reported that over 42% of classical musicians experience substantial noise-induced hearing loss. With approximately 950 professional orchestras worldwide employing around 100,000 musicians, there are a significant number of classical musicians at risk of noise-induced hearing loss.

Exposure to excessive sound levels can cause damage in two ways, mechanical trauma and sensorineural hearing loss:

1. A mechanical hearing loss indicates there is a problem with the mechanism that conducts sound from the environment to the inner ear. Problems in the external auditory canal (outer ear), ear drum or the bones of hearing (the middle ear) may cause a conductive loss. This type of loss can often be corrected by medication or surgery. If it cannot be corrected, the individual can usually do well with a hearing aid.

2. A sensorineural hearing loss indicates a problem in the organs or nerves of hearing. There may be damage to the cochlea, auditory nerve, or the auditory centers of the brain. An individual with sensorineural hearing loss may benefit from a hearing aid, cochlear implant, communication therapies or other medical management depending on the degree or cause of the loss.

Orchestral noise can be defined as the high-level sound produced by an orchestra whilst performing or practicing. This noise can damage the hearing of classical musicians and put orchestras in breach of occupational health and safety legislation. Although NIHL in this industry has been a problem for some time, the combination of ever louder orchestras and tightening noise exposure legislation is impacting on the repertoire played by professional orchestras and threatens the very nature of the industry. The musicians themselves are faced with either sustaining NIHL or wearing personal hearing protection devices such as earplugs and earmuffs. However, current devices suffer from a number of drawbacks, which discourages their use amongst orchestral musicians. Principally, the problem of occlusion, or the dominance of internal noises over external noise when the ear canal is occluded, makes it very difficult for wind, brass and some string players to play to an appropriate standard while they are wearing earplugs. Clearly, for some musicians the choice is currently either to risk sustaining a NIHL or wearing inappropriate hearing protection, both of these are unacceptable as they may result in the early termination of a musician's playing career.

In general, the potential for noise to damage hearing is determined both by the level of the noise and the exposure time to that noise. Noise levels are measured in decibels (dB), which is a logarithmic unit of measurement that expresses the magnitude of power or intensity of a sound, relative to a reference sound pressure level (normally, 20 micro- Pascals). In general, the higher the dB level, the louder the noise. However, weighting filters that approximate the human ear's sensitivity to sounds of different frequencies are often used to form better estimates of the sound pressure levels at the ear, for example, dBA and dBC are two commonly used weighting functions.

Standards set in the United States by Occupational Safety and Health Administration (OSHA) indicate that the exposure of unprotected ears to noise equal to or in excess of 85 dBA over an 8 hour period will cause a gradual hearing loss in a significant number of individuals. Louder noises will accelerate this damage with transient peak levels exceeding 14OdBC deemed to be hazardous. The allowed exposure time decreases by one-half for each 3 to 5dB increase in the average noise exposure level (often referred to as the "exchange rate"). Under Australian legislation, an employer is bound to take action if noise levels go beyond these benchmarks. Musicians routinely face sound pressure levels in the potentially hazardous range. Violins and violas can reach 10OdBA - comparable to the sound of a major sporting event. Brass instruments can reach 114dBA - the level of a chain saw. Symphonic music, at its peak, can reach 13OdBA - the level of a jackhammer. In comparison, a rock/pop band generally reaches 110-12OdBA. Therefore, the noise exposure of orchestras can clearly exceed the nominal benchmark of 85dBA. Considering that classical musicians often practice and/or perform for 4 to 8 hours a day, this matter requires urgent attention as it ultimately endangers the very existence of this industry.

There are many factors that add to the complexity of orchestral noise. For example, the musician's position within the orchestra also influences the noise exposure. Regardless of the wider orchestra set-up, where an individual musician sits relative to their colleagues is largely dictated by the job they hold. This has a significant impact on the nature of the sound to which they are exposed. Moreover, a musician's hearing loss is often asymmetric, relating to the position of the instruments. The violinist hearing loss tends to be worse in the left ear (closer to the instrument), while the flute and piccolo players experience greater loss in the right ear. For further details of the nature of orchestral noise refer to the three year study by Ian O'Brien, Wayne Wilson and Andrew P. Bradley, published in the paper "Nature of Orchestral Noise," in the Journal of the Acoustical Society of America (JASA), 124 (2) August 2008, pp 926-939.

These and other factors, such as variability in the type and level of music played on a daily basis, make currently available noise canceling headphones, earplugs and ear muffs insufficient or inadequate as detailed below.

Currently, there are three broad approaches to hearing protection devices:

1. Passive hearing protection relies upon the damping (attenuation) of the sound by physical isolation of the ear drum using either a range of materials and/or non-powered acoustic filters. In passive protection the sound transmits from the input to the output as an acoustic or mechanical vibration or oscillation. Passive protection results in an increased hearing threshold for the user, which can adversely affect their ability to play. Placing your hands over your ears could be considered passive noise reduction. Etymotic Research lnc of Elk Grove, Illinois produce a series of ear plugs that attenuate sound by 9, 15 or 25dB that are commonly used by rock musicians. The plugs are custom made and designed to maintain high fidelity (that is, spectrally flat) sound reproduction, but at an attenuated level. Currently some orchestral musicians wear these unobtrusive passive ear plugs whilst performing. However, the vast majority do not, as they adversely affect their ability to play their instrument;

2. Electronic hearing protection devices consist of at least one microphone that senses the incoming sound, electronics that then amplify, attenuate, filter or combine these signals and then at least one speaker to output the processed sounds to the users ear. Electronic protection devices can be similar in function to passive devices, providing attenuation and frequency specific filtering. Alternatively, they can have additional functions such as level dependent amplification (as is common in hearing aids for example) and arbitrary spectral and/or spatial mixing of the incoming signals. Electronic hearing protection devices are utilized in both headphones and (in the ear) ear plugs and the later are often combined with behind-the-ear signal processing units. Currently, all electronic hearing protection devices are combined with headphones or earplugs that provide some degree of (additional) passive protection and so this negatively impacts the user's hearing and is detrimental to the musician's ability to play. In addition, devices that utilize ear plugs are prone to the occlusion effect;

3. Active noise cancellation is a form of electronic protection that involves the use of electronics to produce a phase-inverted reproduction of the incoming signal in order to dampen the level of that signal. Noise cancellation headphones measure and analyze the background noise and then emit "anti-noise" of the opposite polarity through a small microphone near the ear to actively cancel out (that is, reduce the sound pressure level of) the noise. Alternatively, systems with two or more microphones can actively cancel "noise" whilst allowing passage of a desirable "signal." For success of these systems however, the measurement and modeling of both the noise and signal is critical and strongly dependent on the application. Therefore, current active noise cancellation systems, designed for removing relatively constant levels of (often low frequency) background noise on airplanes etc., are unsuitable for orchestral musicians. In addition, all current active noise cancelling hearing protection devices combine both passive and active protection and so negatively impact the user's hearing thresholds and consequently their ability to play their instrument.

The majority of current hearing protection devices occlude the ear in some way, most often using either a passive or electronic attenuator inserted directly into the ear canal. This causes two specific problems: attenuation occurs when the protective device is not needed (during quieter passages); and musicians whose instruments are in direct contact with their head experience a predominance of their own sound due to the occlusion effect. Current hearing protection devices are either completely passive or use a combination of passive, electronic or active noise cancelation. Several of these protective devices are specifically designed to protect the hearing of musicians and as such will be described in detail in the following paragraphs.

US patent 4,852,683 "Earplug with Improved Audibility" marketed as a range of earplugs produced by Etymotic Research are an in-the-canal type passive earplug designed specifically for musicians. They attenuate at levels of 9dB, 15dB or 25dB dependent on an inserted filter. They offer limited spectral distortion of the incoming signal, delivered via a specially designed passive filter system. However, as already described, these earplugs are not commonly used by orchestral musicians particularly amongst brass and woodwind players, as they provide constant attenuation (resulting in increased hearing thresholds) and so must be removed during quiet passages. In addition, they do not overcome the occlusion effect. US Patent Application US2008/0044040 "Method and Apparatus for Intelligent Acoustic Signal processing in Accordance with a User Preference," marketed commercially as "Smart Hearing Protection" is an electronic device which passively attenuates incoming signal, with in-the- canal "ear buds," and then amplifies/attenuates and combines signals received by any one or combination of four mounted directional microphones according to the user's discretion. The device is intended to be able to reduce the level of sounds to the rear, the front or on either side of the musician, while rebalancing signals from other directions. However, as the device uses ear buds, it does not reduce the occlusion problem and the level of attenuation is based upon subjective decisions of the user. In addition, the device requires a user to use their hands to adapt the balance and/or attenuation and so is difficult to adjust during a performance.

Published International Patent Application numbers WO 2007/054589 "Hearing Device" and WO 2008/049646 "Method for Hearing Protecting and Hearing Protection System" together describe a low-cost adaptive passive attenuation earplug and a system to control the level of the attenuation for a single musician or group of musicians. There are various embodiments of the 049646 invention, where attenuation levels are decided by the user via a foot controller (as musicians have "busy hands"), via a single group controller operating a volume fader, or automatically using an electronic circuit sensing various sound pressures or light intensities/colours. Alternatively, the system can detect when a musician starts to play their own instrument and increase the level of attenuation accordingly. However, there are a number of limitations with this system such as: when attenuation levels are manually controlled the user may forget to reset to a low attenuation level during quiet passages; when attenuation is controlled centrally for a group of performers the controller does not know individual noise exposure levels and so may under or over attenuate certain musicians within that group; the system uses earplugs with adaptive passive attenuation and so will be prone to the occlusion problem; in "unprotected mode" the earplugs do not provide zero attenuation, rather they provide a "relatively low" level of attenuation which produces a detrimental increase in the musician's hearing threshold; finally it requires a "remote control" for varying the attenuation (either manually or automatically) rather than providing all the required functionality in a single unit.

The background art describes a number of hearing protection devices that are either built into or are adaptable to chairs, such as US 6,119,805 "Hearing Protector Adaptable to Chair," US 5,133,017 "Noise Suppression System" and US 4,977,600 "Sound Attenuation System for Personal Seat." However, these devices are bulky in nature and so are not readily portable. Therefore, musicians will find it difficult to protect their hearing at every location that they may play or practice at, e.g., when practicing at home or playing in venues without such protection devices installed. The 6,119,805 patent describes a passive personal acoustic screen that blocks sounds to the musician from the sides and the rear. It sits in a 'U' shape around the head and may be disengaged or engaged using a manual switching mechanism. Being completely passive it indiscriminately attenuates incoming sound. In addition, the 5,133,017 patent has constant attenuation and so will suffer from the reduced hearing threshold problem previously described.

In summary, none of these prior art devices rely solely upon active noise cancellation, they all attenuate using some form of passive or electronic occlusion of the ear canal. Therefore, a device that provides purely active noise cancellation in a single open (i.e., un-occluded) and portable hearing protection device, that does not increase the musician's hearing threshold and which automatically adapts to, and logs, their noise exposure levels would be a highly beneficial development.

Although the background has focused primarily on sound problems for orchestral musicians it will be appreciated that similar problems arise for any person in a high sound level environment who needs to maintain the fidelity of the sound that they hear, thus precluding prior art devices.

OBJECTS OF THE INVENTION It is an object of the present invention to overcome or at least ameliorate one or more of the above limitations including providing a method for protecting hearing whilst not affecting the fidelity of the sound that is heard.

It is a further object of the present invention to quantitatively monitor, log and/or control each individual's noise exposure levels in order to aid in the implementation and management of an effective hearing conservation program.

SUMMARY OF THE INVENTION

In one form, although it need not be the only nor indeed the broadest form, the invention resides in a hearing protection device comprising: at least one microphone that samples incoming sound and generates an inbound signal; a processor that analyses the inbound signal and generates an adaptive noise reduction signal if the inbound signal exceeds a threshold level; and at least one non-occluding speaker that delivers a noise reducing sound pressure wave from the adaptive noise reduction signal.

Suitably there are two speakers, one associated with each ear of a user, and two microphones, each positioned adjacent a speaker. The speakers, microphones and processor are suitably contained in a head band together with a power source.

The threshold level is suitably a sound level considered to be hazardous. A hazardous level may be determined by the user or by a third party such as an Occupational Health and Safety Authority. The hazardous level may relate to a single event or may be cumulative.

The speakers are preferably circumaural headphones but may be non-occluding ear plugs. The processor suitably generates the adaptive noise reduction signal by inverting the inbound signal and applying gain of the form: gdB = dBactive^*tanh((dBin - CutinGain)/dBrange) - dBactive; where:

• tanh is the hyperbolic tangent function; • dBrange is the range of dB over which to stretch the tanh function;

• dBactive is half of the range in dB over which the gain is varied;

• CutinGain is the dB value for the centre of the tanh function;

• dBin is the incoming sound level.

The hearing protection device may further include a remote controller device that allows manual control of operating parameters of the hearing protection device.

In a further form the invention resides in a method of adaptive noise reduction including the steps of: sampling incoming sound and generating an inbound signal; if the inbound signal exceeds a threshold level generating an adaptive noise reduction signal from the inbound signal by applying nonlinear gain and inverting the inbound signal; and applying the noise reduction signal to at least one non-occluding speaker to deliver a noise reducing sound pressure wave. The adaptive noise reduction occurs by destructive interference of the noise reducing sound pressure wave with the incoming sound. In addition, this system can monitor and/or log individual noise exposure levels over time and, based on this, adapt the degree of hearing protection required, or requested by individual musicians.

The invention described here-in allows a musician's ears to remain completely un-occluded, with the ear hearing unfiltered acoustic sound. As the music approaches hazardous levels, or other predetermined level, the amount of active noise cancellation is automatically increased attenuating the incoming sound at the ear. In particular, the level of the incoming sound determines the amount of noise cancellation (attenuation), not the subjective judgment of the user. Therefore, after initially setting up the device it operates completely automatically adaptively controlling the user's noise exposure to remain within their requested noise exposure limits.

Further features and advantages of the present invention will become apparent from the following detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

To assist in understanding the current invention and to enable a person skilled in the art to put the invention into practical effect, preferred embodiments of the invention will be described by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is a graph of the input (dB_in) to output (dB_0Ut) relationship of: no protection (dB_0Ut = dB_in), passive protection only (dB_0Ut = dB_in - 9dB), constant active protection only (dB_0Ut = dB_in - 15dB), passive and constant active protection (dB_0Ut = dBj_n - 24dB), and the adaptive active hearing protection system disclosed in the current invention;

Figure 2 is a graph of the input (dB_in) to output (dB_0Ut) relationship of the adaptive active noise cancellation system for the cut-in intensities of: 10OdB, 8OdB and 6OdB respectively;

Figure 3 is a schematic of one embodiment of the current invention; Figure 4 shows a non-linear gain function of the inverting amplifier of Figure 3;

Figure 5 demonstrates a typical mono input waveform (top), gain of the inverting amplifier in response to this input (second-top), the output when the input is attenuated by a constant 9dB (second-bottom) and the output when the input is attenuated by the adaptive active noise cancellation system of the current invention (bottom).

Figure 6 is a flow chart of a process for both adapting the cut-in intensity of the hearing protection and logging actual exposure levels for the adaptive active noise cancellation system disclosed in one embodiment of the current invention.

Figure 7 shows the mean dBA and maximum dBC levels on the opening three minutes of Richard Strauss's Also Sprach Zarathustra for: the original music, original attenuated by a constant 9dB (either passive, active of electronic) and the original attenuated by the adaptive active noise cancellation system of the current invention at an average attenuation of 9dB. It also shows how the gain of the inverting amplifier and the cut-in level of the active attenuation is varied as the piece of music progresses in accordance with the flow chart of figure 6; Figure 8 is an illustration of an embodiment of the present invention utilising open circumaural headphones and detailing an optional remote control device with a graphical user interface and wireless communication to the headphones; and

Figure 9 shows an illustration of an embodiment of the device in Figure 8 being worn by a musician. Note that in this embodiment the headband is positioned behind the head for the sake of discretion.

Figure 10 shows a schematic of an embodiment of the current invention highlighting an analogue and digital pathway, plus the ability to present tones to the user to indicate events such as hazardous noise exposure. Figure 11 shows the inputs, outputs and processing undertaken by the digital signal controller in the current embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION Embodiments of the present invention reside primarily in method steps for controlling the gain of the active noise cancelling amplifier based on the sound pressure level of the orchestral noise. Furthermore, the present invention monitors and logs noise exposure levels and can then adapt the overall amount of hearing protection. Accordingly, the method steps have been illustrated in concise schematic form in the drawings, showing only those specific details that are necessary for understanding the embodiments of the present invention, but so as not to obscure the disclosure with excessive detail that will be readily apparent to those of ordinary skill in the art having the benefit of the present description. For ease of understanding the preferred embodiments will be described in terms of the application to aural protection for orchestral musicians. It will be appreciated that the invention is not limited to this specific application.

In this specification, adjectives such as first and second, left and right, and the like may be used solely to distinguish one element or action from another element or action without necessarily requiring or implying any actual such relationship or order. Words such as "comprises" or

"includes" are intended to define a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed, including elements that are inherent to such a process, method, article, or apparatus.

As described in the background to the invention, current hearing protection devices, be they passive, electronic, active or a combination thereof, commonly provide the same amount of attenuation at all noise exposure (input) levels. This is illustrated in Figure 1 for passive attenuation of 9dB, a constant active attenuation of 15dB and a combined attenuation of 24dB. As Figure 1 illustrates, an input (exposure) noise level is reduced by a constant amount, so that an input of 85dB is reduced to 76, 70 and 61 dB for passive, active and combined protection respectively. However, input noise levels at or below 9, 15 and 24dB will be attenuated to OdB for passive, active and combined protection respectively. This means that these noise levels are attenuated below the user's hearing threshold and so they cannot be heard at all. In applications such as industrial and commercial noise reduction this may not be a problem (and may in fact be desirable), but for orchestral musicians this means that they may not be able to hear either themselves or parts of the orchestra in quiet passages. Figure 1 illustrates the intensity-attenuation function of the current invention which provides an increasing amount of attenuation to increasing sound levels. That is, loud sounds are attenuated and quiet sounds are not. For example, it can be seen in Figure 1 that an input noise level of 85dB is attenuated by 15dB by both the active only and the current invention, whilst noise levels below 5OdB are still attenuated by 15dB by the active only, but not attenuated at all by the current invention. In this way, the musician's hearing is protected from the loud sound (85dB) whilst still being able to hear quiet sounds (say below 5OdB).

From the foregoing, it should be emphasized that in the preferred embodiment of the current invention there should be minimal passive attenuation of the incoming sound. In this way, hearing protection (attenuation) can be controlled actively and hearing thresholds (at low sound levels) not adversely affected. In one embodiment of the invention "open" circumaural headphones are used (such as the Sennheiser HD 650 available from Sennheiser Electronics Corporation of Old Lyme, Connecticut, which provides approximately OdB of isolation between 30Hz and 2kHz). Alternatively, insert headphones could be used provided that they do not occlude the ear canal. One method of at least partially achieving this is for musicians' to get their ear plugs molded to just past the bony cartilage to secure the plug to the bone. Alternatively, the be9/be7 hearing devices marketed by ReSound lnc of Bloomington, Minnesota, have "invisible open technology™," wherein the ear canal device allows air to travel freely in and out of the ear, ensuring the user can hear both their own, and other voices, as naturally as possible.

Figure 2 demonstrates how the "cut-in" intensity of the variable intensity-attenuation function can be varied to suit different average or expected sound levels of orchestral noise (say, due to variations in repertoire, venue or position in orchestra). In this way, the definition of "loud" (attenuation required) and "quiet" (no attenuation required) can be varied whilst maintaining a smooth transition over sound increasing levels. Figure 2 illustrates cut-in intensities of 100, 80 and 6OdB that may be suitable to low, average and high exposure levels respectively.

Figure 3 illustrates a device outline of one embodiment of the current invention. The device 30 consists of a microphone 31 , speaker 32, variable gain amplifier 33, inverter 34, frequency weighting filter 35 and sound level estimation circuits 36. It is speculated that for optimal results a circuit such as this is required to protect each ear. However, a system utilizing a single microphone or sound level estimator may also produce the desired results in some situations. It is also realized that relative positioning of the microphone and speaker are of importance to the effectiveness of active noise cancellation systems and a number of schemes are known in the prior art such as that disclosed in US patent 6,831 ,984 "Noise Reducing" and evaluated by S. M. Kuo, S. Mitra and W.S Gan in "Active Noise Control System for Headphone Applications," IEEE TRANSACTIONS ON

CONTROL SYSTEMS TECHNOLOGY, VOL. 14, NO. 2, MARCH 2006.

The specific choice of components will depend on the specific application. Applications such as orchestral situations are anticipated to require high-end microphones, speakers and electronics. On the other hand, hearing protection for air travelers may use lower standard components. Sound detected by the microphone 31 is weighted in the filter 35 using A-weighting or C-weighting as described below. The sound level is estimated in sound level estimator 36 and the combination is applied to the sound signal to generate a gain curve for the variable gain amplifier. The signal is inverted by inverter 34 and the inverted signal is applied by the speaker 32. It will be appreciated that individual functions may not correlate to discrete devices but could be performed by a specific circuit or combination of circuits.

The form of the input (exposure) noise level to inverting amplifier gain function for one embodiment of the current invention is illustrated in

Figure 4, showing the function itself, some ideal (un-weighted) dB levels and dB levels with an A-weighting (dBA) both estimated from a single piece of (mono) music (in this case, a snippet of Handel's Hallelujah Chorus). In the preferred embodiment the form of the function to determine the gain of the inverting amplifier (gdB) is: gdB = dBactive^*tanh((dBin - CutinGain)/dBrange) - dBactive; Where:

• tanh(x) is the hyperbolic tangent of the elements of x;

• dBrange (typically set to 20) is the range of dB over which to stretch the tanh function;

• dBactive (typically set to 15) is half of the range in dB that you want the gain of the inverting amplifier to vary over. In this way, a dBactive of 15dB would ideally result in a maximum active attenuation of 3OdB;

• CutinGain is the dB value for the centre of tanh function (typically set to 60);

• dBin is the sound level of the input (exposed) noise.

Figure 4 illustrates how the gain of the amplifier is dependent upon the intensity of the incident sound measured by the microphone. As is well known in the prior art, there are numerous methods available to estimate the sound level recorded by the microphone in Figure 3 (dBin). Typically a frequency weighting filter such as A or C is applied in filter 35 and then the root mean square (RMS) sound pressure is estimated, Prms, in sound level estimator 36. The sound pressure level (SPL) in dB can then be calculated as: dBin = 20^*log10(Prms/Pref);

The commonly used reference sound pressure in air is Pref = 20 μPa (RMS), which is usually considered the threshold of human hearing. Figure 5 demonstrates a typical mono input waveform (top), gain of the inverting amplifier (gdB) in response to this input (second-top), the output when the input is attenuated by a constant 9dB (second-bottom) and the output when the input is attenuated by the adaptive noise cancellation system of the current invention (bottom). Figure 5 clearly demonstrates that the current invention preserves the low amplitude portions of the input whilst attenuating the large amplitude portions, as compared to the constant attenuation that reduces the level of both the low and large amplitudes by the same amount (9dB). It can also be seen, for example at 50, that the gain of the inverting amplifier is modified in proportion to the amplitude or intensity of the input signal.

The current invention's non-linear compression of large amplitude signals evident in Figure 5 can be viewed as an "acoustic companding" or "acoustic dynamic range compression." Well known in the telecommunications and audio engineering prior art, companding is a method of mitigating the detrimental effects of a telecommunications channel with limited dynamic range, while dynamic range compression (DRC) is a process that reduces the dynamic range of audio signals and is used in sound recordings, live sound reinforcement and broadcasting. However, the current invention differs from this prior art in that the compression of large amplitude signals is happening acoustically, that is, via destructive interference (active cancellation) of the sound pressure waves, rather than electronically as in companding and DRC (where the control circuit monitors the amplitude of the input signal and controls the gain applied to that signal before it is broadcast or output to speakers). In the current invention the control circuit again monitors the amplitude of the input signal, but controls the gain of an inverting amplifier that outputs an out of phase signal to the speaker which then destructively interferes with the sound just outside the ear canal, thus decreasing the amplitude of the sound pressure wave entering the ear.

Figure 6 illustrates a flow chart of an Adaptive Control Loop process for both adapting the cut-in intensity of the hearing protection and logging actual exposure levels for the adaptive attenuation system described in one embodiment of the current invention. In the flow chart of Figure 6 the system first initialises some parameters that define the operation of the device, in the preferred embodiment this is done via a graphical user interface on a remote control, where:

• Delta defines the step size (increments) that the CutinGain is modified by in each iteration of the algorithm (typically 3dB, as this is the smallest change in sound level that can be perceived);

• MidGain defines the current cut-in gain of the inverting amplifier (typically, it would be initialised to MaxGain below);

• MinGain and MaxGain define the minimum and maximum allowed cut-in gains respectively (typically 40 and 8OdB respectively).

Next, the requested equivalent sound level, LEQR, is set either directly by the user or based on legislation of orchestral guidelines (and may be specified as a dBA, dBC or unweighted (dBZ) or similar equivalent level). The value specified by LEQ_R is the sound level that the hearing protection device of the current invention will attempt to achieve (that is, keep the users noise exposure level below) through the modification of the CutinGain (in the current embodiment of the invention). The following steps are performed continually whilst the device is powered on (that is, being worn by the user). First, the short-term equivalent sound level, LEQST, is estimated over a suitably short time period, say between 1 and 10 seconds. Next, the current estimate of the long-term (average) equivalent sound level, LEQLT, is estimated. In the preferred embodiment an iterative method is used to estimate the long-term RMS sound pressure, RMS_Lτ, from the current short-term RMS sound pressure, RMS_ST, and the previous estimate of RMSLT as follows:

RMSLT = ((iter - 1)^*RMS_LT + RMS_ST)/iter; Where iter is the number of iterations performed at the current iteration. Next, LEQ_LT (in dB) is estimated directly from the logarithm of RMSLT. Both LEQ_Sτ and/or LEQLT, or any measure derived or related to them, can be stored in non-volatile memory on the device and later transferred either by a wired or wireless link to the remote control device (in Figure 8). Alternatively, these measures can be transferred to a personal computer so that they can be monitored on an orchestral (group) level to ensure that the orchestra is complying with occupational health and safety requirements in relation to noise exposure or as part of the ongoing management of an effective orchestral hearing conservation program. In the current embodiment of the invention, the short-term equivalent sound level, LEQ_ST, is calculated for both the original (exposed) sound level that the ear would have been exposed to assuming no adaptive noise cancellation (that is, via the microphone in Figure 3) and an estimate of the sound level at the ear after active noise cancellation (that is, the sound level that the ear is actually exposed to) via the following equation which estimates the effectiveness of the adaptive noise cancellation:

Xear(t) = Xinput(t) - g*(Xinput(t + φ))i

Where g is the gain of the inverting amplifier and φ is the estimated phase delay (or advance) of the adaptive noise cancellation circuit of Figure 3. That is, the phase difference between: a sound being recorded at the microphone and being output by the speaker; and a sound at the ear when there is no noise cancellation, x_input. A number of methods for estimating this phase delay will be known to those possessing an ordinary skill in the background art and it will also be realized that the phase delay may additionally be a function of the frequency, φ(/). In an alternative embodiment of the current invention the effectiveness of the adaptive noise cancellation (and hence the short-term equivalent sound level, LEQ_Sτ) can be estimated by measuring the sound pressure level in, or at the entrance to, the ear canal (often referred to as a "real ear" measurement). The prior art describes a number of arrangements involving an air tube connected to an additional microphone that can achieve this purpose.

Next, if LEQLT is greater than LEQR then MidGain is reduced by Delta. If, after being reduced, MidGain is less than MinGain then it is set to MinGain. If LEQLT is less than LEQ_R then MidGain is increased by Delta. If, after being increased, MidGain is geater than MaxGain then it is set to MaxGain. If LEQLT is equal to LEQ_R then MidGain is not changed. Once the current value of MidGain has been determined it is used to update the nonlinear gain function of the inverting amplifier that performs the adaptive noise cancellation (as per Figure 2). In a further embodiment of the invention CutinGain, dBrange and dBActive are varied in order to maintain LEQLT below LEQR. Alternatively, methods that adapt the value of Delta so that MidGain changes by a larger amount when LEQLT has been either above or below LEQR for a number of iterations and decreases Delta when LEQLT is alternating above and below LEQR in subsequent iterations are known in the prior art (for example, adaptive delta modulation applies a similar extension to the delta modulation scheme known in the telecommunications field).

Figure 7 show for the preferred embodiment of the current invention the mean dBA and maximum dBC levels on the opening three minutes of Strauss's Also Sprach Zarathustra for: the original music, original attenuated by a constant 9dB and attenuated by the adaptive active noise cancellation system of the current invention at an average attenuation of 9dB. It also shows how the gain of the inverting amplifier and the cut-in (CutinGain) level of the active attenuation are varied as the piece of music progresses. Figure 7 clearly demonstrates how the inverting amplifier's gain is increased during noisy passages and decreased during quiet passages. It also shows how the cut-in gain (Centre dB in Figure 7) is adapted according to the flow chart of Figure 6 over the same period. Finally, it shows how the equivalent (average) sound level, LEQ_Lτ (Mean dBA in Figure 7), maintained below LEQR (in this case 64dBA).

In a further embodiment of the invention the flow chart of figure 6 is modified based on previous noise exposure pattern experienced by the user (eg. venue, position, repertoire etc).

Figure 8 shows one embodiment of the current invention where the device 80 is a discreet open-ear pair of circumaural headphones 81 with a power source such as a battery pack enclosed in the head band 82, controlled by a remote device 83. Discreet externally mounted microphones 84 sample the incoming sound and feed level information to a processor in the headband 82. If the level is hazardous this triggers a set amount of adaptive noise reduction delivered via speaker diaphragms 85 suspended at the ear driven by the inverting amplifier 33/34, which amplifies the signal from the processor. However, in an alternative embodiment of the invention the device could be worn at least partially in the ear provided that it does not fully occlude the ear canal.

Figure 9 illustrates the open circumaural embodiment of the current invention being worn by a musician.

Figure 10 shows an expanded schematic of an embodiment of the current invention highlighting a wholly analogue pathway from microphone

100 to headphone speaker 110 that performs the active noise cancellation and a digital pathway that estimates the noise level and actively adapts the gain and phase of the analogue pathway. The advantage of a wholly analogue pathway between microphone and speaker is that it minimises the absolute (input to output) delay and therefore improves the performance of the noise cancellation, especially at higher frequencies. However, a potential disadvantage of a completely analogue pathway (including the frequency responses of the microphone and headphone) is that it will typically have a delay that is a non-linear function of frequency. That is, it will demonstrate a measurably non-linear phase response, where certain frequencies take longer to progress from input to output than other frequencies and so adversely affecting noise cancellation performance. Therefore, in one preferred embodiment of the invention there is an adaptive phase control unit to allow for the compensation (this is, linearisation) of the phase response from microphone to headphone. It would be known to those skilled in the art that once all of the components in the analogue pathway have been selected and the relative positioning on microphone and headphone determined then the phase response of the system is essentially fixed, can be measured and then linearised (compensated). Therefore, the phase control does not typically need to be adapted during normal operation of the devices. In addition, Figure 10 illustrates a signal selector 104 which enables the digital signal controller 105 to present tones from a tone generator 106 to the user to indicate events such as exposure to hazardous noise levels or provide user feedback such as indication of particular operational modes.

The embodiment outlined in Figure 10 measures noise levels and creates deconstructive interference to limit noise exposure at the ear. Deconstructive interference implies that the open loop response of the system (electronic and acoustic) should be 180° out of phase (or as close to it as possible) across the audible frequency range. The open loop gain of the system is used to control the amount of cancellation, and is determined on-the-fly by the Adaptive Control Loop as shown in Figure 6.

The system begins with a miniature microphone 100 mounted within the headphone, mounted close to the headphone driver. This close mounting ensures minimal acoustic phase delay between the microphone and headphone driver, thus eliminating the need for delays in the electronic signal path. The current prototype utilises an AKG C417 Lavalier Microphone (details available from http://www.akg.com). This microphone was chosen for its omnidirectional polarity pattern and small size.

The microphone signal is then amplified by a discrete bipolar junction transistor (BJT) based amplifier 101. This microphone amplifier design is based on Phil Allison's Low Noise Balanced Microphone Preamp (details available from http://sound.westhost.com/project66.htm). This design was originally chosen for its linearity in gain across the audible frequency spectrum. However it would be desirable to integrate the microphone amplifier with other circuitry so as to reduce the physical size and power requirements of the device.

Once the microphone signal has been amplified it is passed through the Adaptable Phase Controller 102 and Adaptable Gain Controller 103. These two systems are used to ensure that the overall (electronic and acoustic) open loop gain is 180° out of phase, and to control the amount of active noise cancellation as determined by the Adaptive Control Loop as shown in Figure 6. In its simplest embodiment the phase control is an operational amplifier based inverter. However, as previously explained, alternative embodiments will have a filter 107 with specific phase response across the required range of frequencies, which matches the acoustic response of the microphone and headphone driver and compensates for their specific phase delays. The adaptability of the phase control system will allow different combinations of headphones and microphones to be utilised and could allow on-the-fly tuning to ensure optimal active noise cancellation in a final product. In the current embodiment of the invention the adaptable gain control system is based around a Dallas DS1267 Dual Digital Potentiometer (details available from http://www.maxim- ic.com/quick_yiew2.cfm/qv_pk/2676). However, again a fully integrated design may be preferred for both physical size and power requirements. The signal selector 104 is used to switch the device between cancellation mode and tone generator mode. In tone mode, tones can be played to the user to notify them of changes in mode or to warn them of dangerous exposure levels.

The headphone amplifier 109 is used to drive the headphones with the cancellation signal. Currently, the design is based off the PIMETA v1 Headphone Amplifier (details available from http://tangentsoft.net/audio/pimeta/). This design was chosen for its high fidelity. However, like the microphone amplifier it would be obvious to those skilled in the art that this could be replaces with an integrated package.

The Digital Signal Controller 105 is used to control the amount of cancellation applied to the headphones, having been programmed with the adaptive control loop as per Figure 6. In the current embodiment the controller performs digital analysis of the incoming signal and is not directly part of the analogue cancellation signal path. Rather, the digital signal controller is used to control the amount of cancellation by adjusting the adaptable gain control (and optionally the adaptable phase control system). In the current embodiment the controller is a Microchip dsPIC33FJ128GP802 DSC (details available at http://www.microchip.com/wwwproducts/Devices. aspx?dDocName=en5322 98). The controller calculates the root mean square (RMS) value of the microphone signal weighted by either A, B or Z frequency weightings and utilises this value in the adaptive control loop, in one embodiment the controller retrieves gain and phase values from a non-linear lookup table based on the weighted RMS microphone signal but in another embodiment could calculate the values directly. The RMS readings are also stored in the microcontroller logging purposes or for possible off-line analysis.

Figure 11 shows the inputs, outputs and processing undertaken by the digital signal controller in the current embodiment of the invention. The input from the microphone is first pre-filtered 107 to prevent aliasing and then converted to digital by the integrated analogue to digital converter (ADC) 1051. Next, the frequency weighting filter 1052 is applied and the RMS value 1053 calculated. Based on this RMS value the digital signal controller 105 applies the control loop 1054 as described in Figure 6 to adjust the adaptable gain controller 103 and adaptable phase controller 102. Local storage 1055 provides data logging for off-line analysis.

The invention has particular advantage because it allows the ear to remain completely un-occluded, with the ear hearing unfiltered acoustic sound. When the music reaches hazardous levels, active noise cancelling occurs to attenuate the incoming sound at the ear and thus providing hearing protection.

The above description of various embodiments of the present invention is provided for purposes of description to one of ordinary skill in the related art. It is not intended to be exhaustive or to limit the invention to a single disclosed embodiment. As mentioned above, numerous alternatives and variations to the present invention will be apparent to those skilled in the art of the above teaching. Accordingly, while some alternative embodiments have been discussed specifically, other embodiments will be apparent or relatively easily developed by those of ordinary skill in the art. Accordingly, this invention is intended to embrace all alternatives, modifications and variations of the present invention that have been discussed herein, and other embodiments that fall within the spirit and scope of the above described invention.

Claims

1. A hearing protection device comprising: at least one microphone that samples incoming sound and generates an inbound signal; a processor that analyses the inbound signal and generates gain function parameters for an outbound signal if the inbound signal exceeds a threshold level; an amplifier that produces the outbound signal according to the gain function parameters; and at least one non-occluding speaker that receives the outbound signal and delivers a noise reducing sound pressure wave.

2. The hearing protection device of claim 1 comprising two speakers, one associated with each ear of a user and two microphones, each positioned adjacent a speaker.

3. The hearing protection device of claim 1 comprising a headband, wherein the headband contains the least one microphone, the at least one non-occluding speaker, and the processor.

4. The hearing protection device of claim 3 further containing a power source in the headband.

5. The hearing protection device of claim 1 wherein the threshold level is a sound level considered to be hazardous.

6. The hearing protection device of claim 5 wherein the hazardous level is determined by the user.

7. The hearing protection device of claim 5 wherein the hazardous level is determined by a third party.

8. The hearing protection device of claim 5 wherein the hazardous level relates to a single event.

9. The hearing protection device of claim 5 wherein the hazardous level is cumulative.

10. The hearing protection device of claim 1 wherein the speakers are circumaural headphones.

11. The hearing protection device of claim 1 wherein the speakers are non-occluding ear plugs.

12. The hearing protection device of claim 1 wherein the processor generates the gain function parameters from a gain function of the form: gdB = dBactive*tanh((dBin - CutinGain)/dBrange) - dBactive; where: tanh is the hyperbolic tangent function; dBrange is the range of dB over which to stretch the tanh function; dBactive is half of the range in dB over which the gain is varied; CutinGain is the dB value for the centre of the tanh function; and dBin is the incoming sound level.

13. The hearing protection device of claim 1 further comprising a remote controller device that allows manual control of operating parameters of the hearing protection device.

14. A method of adaptive noise reduction including the steps of: sampling incoming sound and generating an inbound signal; if the inbound signal exceeds a threshold level generating gain function parameters for an outbound signal; generating the outbound signal by applying non-linear gain and varying the phase of the inbound signal; and applying the outbound signal to at least one non-occluding speaker to deliver a noise reducing sound pressure wave.

15. The method of claim 14 wherein the threshold level is a sound level considered to be hazardous.

16. The method of claim 15 wherein the hazardous level relates to one or more events and is cumulative.

17. The method of claim 14 wherein the gain function parameters are derived for a gain of the form: gdB = dBactive^*tanh((dBin - CutinGain)/dBrange) - dBactive; where: tanh is the hyperbolic tangent function; dBrange is the range of dB over which to stretch the tanh function; dBactive is half of the range in dB over which the gain is varied; CutinGain is the dB value for the centre of the tanh function; and dBin is the incoming sound level.

18. The method of claim 14 further including the steps of: monitoring individual noise exposure levels over time; and adapting the gain function parameters to suit user preferences.

19. The method of claim 14 further including the steps of: monitoring individual noise exposure levels over time; and adapting the gain function parameters to avoid hazardous levels.

20. The method of claim 14 wherein the gain function parameters are determined from a lookup table.