Adaptive Hearing Health Control

Abstract

A system (200) for controlling sound dose to a listener is presented. The system (200) comprises a cloud server (210) operatively connected to a remote storage (220) configured to store a plurality of audio profiles, each of which includes sound pressure data and a hearing profile associated with a particular user of the system (200). The system (200) comprises a transducer (15), a controller (100), and a current sound pressure (p) experienced by a listener (40). c and a microphone (17) positioned to measure a frequency response (Fp) of the listener (10) relative to the sound pressure data. The cloud server (210) and / or the controller (100) of the one or more audio playback arrangements (10) include control circuitry for estimating, based on the sound pressure data or the like, a current estimated time period representing an estimated time until the listener's sound dose exceeds an acoustic dose threshold.

Description

[Technical field]

[0001] The present invention relates to hearing health, and more particularly to controlling a listener's sound dose. [Background technology]

[0002] Access to audio in all forms has increased greatly with the introduction of portable electronic devices such as the Walkman and later mobile phones. An audiobook, a favorite song or an interesting podcast is always within reach. This, combined with the increase in general environmental noise from industry, traffic, etc., has led to an increased awareness of the deteriorating hearing health of a large proportion of the human population.

[0003] The World Health Organization (WHO) provides several studies reporting the dangers of listening to loud volumes for extended periods of time. For example, according to the WHO, more than one billion young people aged 12-35 are at risk of hearing loss due to exposure to loud volumes for recreational purposes. The WHO continues to estimate the annual cost of untreated hearing loss at $980 billion worldwide. The WHO estimates that 50% of hearing loss could be prevented by public health measures. Some prevention strategies target exposure to loud sounds and music, or individual lifestyle choices such as wearing protective devices such as earplugs. This can be aided by implementing audio standards for personal audio systems and devices.

[0004] As a result, the International Telecommunications Union (ITU) has published recommendation ITU-T H.870 entitled "Guidelines for safe listening devices / systems". This document explains how sound dose, or acoustic dose, is calculated and compared to limits that indicate safe listening. Further guidelines have been published, for example, by various national and regional occupational safety and health administrations, and other national institutes for occupational safety and health.

[0005] To prevent ear damage, headphones for consumer electronics are provided with a predefined maximum output level. This blunt approach does not take into account the duration of listening, and users risk developing long-term hearing health problems by listening at too high a level for too long a period of time.

[0006] From the above, it is clear that there is room for improvement. Summary of the Invention [Problem to be solved by the invention]

[0007] The object of the present invention is to provide a new type of audio control that improves upon the prior art and eliminates or at least mitigates the above-mentioned drawbacks. More specifically, the object of the present invention is to provide a system for monitoring sound dose for a listener that is convenient for the listener and reduces the risk of damaging the listener's hearing. These objects are achieved by the techniques described in the attached independent claims, preferred embodiments of which are defined in the associated dependent claims. [Means for solving the problem]

[0008] In a first aspect, a system for controlling a sound dose of a listener is presented. The system comprises a cloud server operatively connected to a remote storage. The remote storage is configured to store a plurality of audio profiles. Each audio profile includes sound pressure data and a hearing profile associated with a particular user of the system. The system further comprises one or more audio playback arrangements comprising a transducer, a controller, and a microphone. The microphone is arranged to measure a current sound pressure experienced by the listener. The one or more audio playback arrangements are operatively connected to the cloud server and configured to determine an identity of the listener. The system is configured to associate the determined identity with the particular user of the system. The cloud server and / or the controller of the one or more audio playback arrangements comprise a control circuit, the control circuit configured to cause the steps of obtaining an audio profile including sound pressure data and a hearing profile for the listener and obtaining the current sound pressure experienced by the listener. Based on the sound pressure data, the listener's hearing profile, and the most recent sound pressure, the control circuitry is further configured to cause the step of estimating a most recent estimated time period representing an estimated time until the listener's sound dose exceeds an acoustic dose threshold by frequency weighting the most recent sound pressure based on the listener's hearing profile. In response to the most recent estimated time period falling below a dose period associated with the acoustic dose threshold based on the sound pressure data and the most recent sound pressure, the control circuitry is further configured to cause the step of calculating an updated sound pressure where the most recent updated time period representing an estimated time until the listener's sound dose exceeds an acoustic dose threshold meets or exceeds the dose period. The control circuitry is further configured to cause the step of providing instructions to control the most recent sound pressure to meet the updated sound pressure.

[0009] In one variant, the sound pressure data of the audio profile comprises historical sound pressure data comprising previous sound pressures experienced by the listener, which is beneficial as it makes the estimation of the latest estimation period more accurate.

[0010] In one variant, the sound pressure data of the audio profile comprises the cumulative sound pressure experienced by the listener, which is advantageous since it allows for storing and sharing the cumulative sound pressure. The obtained cumulative sound pressure may for example be accumulated while the listener is listening to different audio reproduction configurations.

[0011] In one variant, the sound pressure data of the audio profile comprises a trend in the sound pressure experienced by the listener, which is beneficial as it makes the estimation of the latest estimation period more accurate.

[0012] In one variant, the latest sound pressure is obtained from microphones placed at the listener's ears, which is advantageous as it allows the latest sound pressure to be obtained directly without excessive processing.

[0013] In one variant, the step of estimating a current estimated period representative of an estimated time until the acoustic dose of the listener exceeds the acoustic dose threshold comprises determining a prediction model, which is advantageous as it makes the estimation of the current estimated period more accurate.

[0014] In one variant, the forecast model comprises an average model, which is beneficial as it makes the estimation of the latest estimated period more accurate.

[0015] In one variant, the forecast model comprises a recurrent neural network (RNN) model, which is beneficial as it makes the estimation of the latest estimated period more accurate.

[0016] In one variant, the control circuitry is further configured to cause updating of sound pressure data of the audio profile for the listener based on the most recent sound pressure, which is beneficial as it simplifies, for example, the storage and sharing of the accumulated sound pressure.

[0017] In one variant, the control circuitry is configured to cause the step of providing instructions for controlling the current sound pressure for the left and / or right ear of the listener, the audio profile comprising separate sound pressure data for each of the left and right ears of the listener, which is advantageous as it allows for more precise control of the sound dose.

[0018] In a second aspect, a method for controlling a sound dose of a listener is presented. The method includes obtaining an audio profile including sound pressure data of the listener and obtaining a latest sound pressure experienced by the listener. The method further includes estimating a latest estimated period representing an estimated time until the sound dose of the listener exceeds a sound dose threshold based on the sound pressure data and the latest sound pressure. If the latest estimated period is below a dose period associated with the sound dose threshold, the method includes calculating an updated sound pressure based on the sound pressure data and the latest sound pressure such that a latest updated period representing an estimated time until the sound dose of the listener exceeds a sound dose threshold meets or exceeds the dose period, and providing instructions for controlling the latest sound pressure to meet the updated sound pressure.

[0019] In one variant, the sound pressure data of the audio profile comprises historical sound pressure data comprising previous sound pressures experienced by the listener, which is beneficial as it makes the estimation of the latest estimation period more accurate.

[0020] In one variant, the sound pressure data of the audio profile comprises the cumulative sound pressure experienced by the listener, which is advantageous since it allows for storing and sharing the cumulative sound pressure. The obtained cumulative sound pressure may for example be accumulated while the listener is listening to different audio reproduction configurations.

[0021] In one variant, the sound pressure data of the audio profile comprises a trend in the sound pressure experienced by the listener, which is beneficial as it makes the estimation of the latest estimation period more accurate.

[0022] In one variant, the latest sound pressure is obtained from microphones placed at the listener's ears, which is advantageous as it allows the latest sound pressure to be obtained directly without excessive processing.

[0023] In one variant, the audio profile further comprises a hearing profile of the listener. In this variant, the latest estimated time may further be based on the hearing profile of the listener by frequency-weighting the latest sound pressure based on the hearing profile of the listener. This is beneficial because the weighted latest sound pressure is the actual sound pressure perceived by the listener. As an extreme example, if the listener cannot hear certain frequency bands, the sound pressure in these frequency bands is substantially harmless to the listener.

[0024] In one variant, the step of estimating a current estimated period representative of an estimated time until the acoustic dose of the listener exceeds the acoustic dose threshold comprises determining a prediction model, which is advantageous as it makes the estimation of the current estimated period more accurate.

[0025] In one variant, the forecast model comprises an average model, which is beneficial as it makes the estimation of the latest estimated period more accurate.

[0026] In one variant, the forecast model comprises a recurrent neural network (RNN) model, which is beneficial as it makes the estimation of the latest estimated period more accurate.

[0027] In one variant, the method further comprises updating the sound pressure data of the audio profile for the listener based on the latest sound pressure, which is beneficial as it simplifies, for example, the storage and sharing of the accumulated sound pressures.

[0028] In one variant, the audio profiles are retrieved from a remote storage, which is beneficial as it simplifies, for example, the storage and sharing of the audio profiles.

[0029] In one variant, the method is performed for the left and / or right ear of the listener and the audio profile comprises separate sound pressure data for each of the listener's left and right ears, which is advantageous as it allows for more precise control of the acoustic dose.

[0030] In a third aspect, a controller is presented for controlling an audio playback arrangement. The audio playback arrangement comprises an audio transducer and a microphone arranged to measure a most recent sound pressure experienced by a listener of the audio playback arrangement. The controller comprises a control circuit, the control circuit configured to cause the steps of obtaining an audio profile including sound pressure data for the listener and obtaining a most recent sound pressure experienced by the listener. Based on the sound pressure data and the most recent sound pressure, the controller is configured to cause the steps of estimating a most recent estimated time period representing an estimated time until an acoustic dose of the listener exceeds an acoustic dose threshold. In response to the most recent estimated time period falling below a dose period associated with the acoustic dose threshold, the controller is configured to cause the steps of calculating an updated sound pressure based on the sound pressure data and the most recent sound pressure, where the most recent updated time period representing an estimated time until an acoustic dose of the listener exceeds an acoustic dose threshold meets or exceeds the dose period. The controller is further configured to cause the steps of providing instructions for controlling the most recent sound pressure to meet the updated sound pressure.

[0031] In a fourth aspect, an audio playback arrangement is presented comprising the controller of the second aspect.

[0032] In one variant, the audio reproduction arrangement further comprises an audio transducer and a microphone arranged to measure the current sound pressure experienced by a listener of the audio reproduction arrangement.

[0033] In one variation, the controller is configured to obtain an audio profile including sound pressure data for the listener and obtain a latest sound pressure experienced by the listener. Based on the sound pressure data and the latest sound pressure, the controller is configured to estimate a latest estimated time period representing an estimated time until the listener's sound dose exceeds an acoustic dose threshold. If the latest estimated time period is below a dose period associated with the acoustic dose threshold, the controller is configured to calculate an updated sound pressure based on the sound pressure data and the latest sound pressure such that the latest updated time period representing an estimated time until the listener's sound dose exceeds an acoustic dose threshold meets or exceeds the dose period. The controller is further configured to provide instructions for controlling the latest sound pressure to meet the updated sound pressure.

[0034] In one variant, the controller is further configured to execute the method according to the first aspect.

[0035] In one variant, the controller is further configured to control an audio transducer to reproduce the source signal S received by the controller.

[0036] In one variation, the controller is further configured to process the source signal to adjust the intelligibility of the source signal.

[0037] In one variant, the audio reproduction arrangement further comprises an external microphone, and the intelligibility of the source signal is adjusted based on background noise acquired from the external microphone.

[0038] In one variant, the audio profile further comprises a hearing profile describing the hearing impairment of the listener, and the controller is further configured to process the source signal based on the hearing profile. This is beneficial because the weighted latest sound pressure is the actual sound pressure perceived by the listener. As an extreme example, if the listener cannot hear in certain frequency bands, the sound pressure in these frequency bands is substantially harmless to the listener.

[0039] In one variation, the controller is further configured to control the transducer based on the instruction to control the latest sound pressure to meet the updated sound pressure.

[0040] In a fifth aspect, a system for controlling sound dose for a listener is presented. The system comprises a cloud server operatively connected to a remote storage configured to store an audio profile including sound pressure data for the listener, and one or more audio playback arrangements comprising a transducer, a controller, and a microphone positioned to measure the current sound pressure experienced by the listener. The one or more audio playback arrangements are operatively connected to the cloud server, and the cloud server and / or the one or more audio playback arrangements comprise the controller of the second aspect.

[0041] In a variant, the remote storage is configured to store a plurality of audio profiles, each of which includes sound pressure data associated with an individual user of the system. The one or more audio reproduction configurations are configured to determine an identity of a listener. The system is further configured to associate the identity with an individual user of the system. This is advantageous since audio reproduction configurations may be shared between multiple listeners and / or listeners may use different audio reproduction configurations and still have precise control of the acoustic dose.

[0042] In a sixth aspect, a computer program product is presented comprising a non-transitory computer readable medium having thereon a computer program including program instructions, the computer program being loadable into a controller and configured to cause a method according to the first aspect to be performed when the computer program is executed by the controller. [Brief description of the drawings]

[0043] Embodiments of the present invention will now be described with reference to the accompanying schematic drawings, which show non-limiting examples of how the inventive concepts may be put into practice.

[0044] [Figure 1] FIG. 1 is a block diagram of an embodiment of the present disclosure. [Figure 2a] 1 is a time series plot of acoustic dose according to an embodiment of the present disclosure. [Figure 2b] 1 is a time series plot of acoustic dose according to an embodiment of the present disclosure. [Figure 3a] FIG. 2 is a schematic diagram of an audio playback configuration according to an embodiment of the present disclosure. [Figure 3b] FIG. 2 is a schematic diagram of an audio playback configuration according to an embodiment of the present disclosure. [Figure 3c] FIG. 2 is a schematic diagram of an audio playback configuration according to an embodiment of the present disclosure. [Figure 3d] FIG. 2 is a schematic diagram of an audio playback configuration according to an embodiment of the present disclosure. [Figure 4] FIG. 1 is a block diagram of a system according to an embodiment of the present disclosure. [Diagram 5] FIG. 2 is a signaling diagram of a system according to an embodiment of the present disclosure. [Figure 6] 1 is a time series plot of acoustic dose according to an embodiment of the present disclosure. [Figure 7] FIG. 2 is a block diagram of a method according to an embodiment of the present disclosure. [Figure 8a] FIG. 2 is a schematic diagram of a controller according to an embodiment of the present disclosure. [Figure 8b] FIG. 2 is a schematic diagram of a controller according to an embodiment of the present disclosure. [Figure 8c] FIG. 2 is a schematic diagram of a controller according to an embodiment of the present disclosure. [Figure 8d] FIG. 2 is a schematic diagram of a controller according to an embodiment of the present disclosure. [Figure 8e] FIG. 2 is a schematic diagram of a controller according to an embodiment of the present disclosure. [Figure 9] FIG. 2 is a schematic diagram of a computer program according to an embodiment of the present disclosure. [Figure 10a] FIG. 2 is a schematic diagram of a computer program according to an embodiment of the present disclosure. [Figure 10b] FIG. 2 is a schematic diagram of a computer program according to an embodiment of the present disclosure. [Figure 10c] FIG. 2 is a schematic diagram of a computer program according to an embodiment of the present disclosure. [Figure 10d] FIG. 2 is a schematic diagram of a computer program according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0045] In the following, specific embodiments will be described in more detail with reference to the accompanying drawings. However, the present invention may be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided as examples so that this disclosure will be thorough and complete, and will fully convey the scope of the invention as defined in the appended claims to those skilled in the art.

[0046] The term "coupled" is defined as connected, but not necessarily directly, and not necessarily mechanically. Two or more items that are "coupled" may be integral with one another. The terms "a" and "an" are defined as one or more, unless the disclosure expressly requires otherwise. The terms "substantially," "approximately," and "about" are defined to be generally, but not necessarily entirely, as would be understood by one of ordinary skill in the art. The terms "comprises" (and any form thereof, e.g., "comprises" and "comprising"), "has" (and any form thereof, e.g., "has" and "having"), "includes" (and any form thereof, e.g., "includes" and "comprising"), and "contains" (and any form thereof, e.g., "contains" and "contains") are open-ended linking verbs. As a result, a method that "comprises," "has," "includes," or "contains" one or more steps is not limited to possessing one or more of those steps, but only possessing one or more of those steps.

[0047] In this specification, the terms playback system, playback device, or playback volume shall mean playback in the sense of playing received audio data. This term does not mean playback in the limited form of playing stored audio or media, but encompasses the playback of all forms of media, e.g., real-time media, stored media, streamed media, progressively downloaded media, etc. In addition, the term compensation shall mean compensation in its broadest form and can be considered synonymous with the term processing in this disclosure. This is applicable not only when describing actions, but also when the term is used to describe objects, e.g., compensation filters.

[0048] Starting with FIG. 1, the present disclosure is directed to controlling the acoustic dose D experienced by a listener 40 such that the risk of damage or injury to the listener's hearing is reduced. The objective is to achieve this without unduly limiting the listening experience and listening comfort of the listener 40. The acoustic dose D is a measure of the cumulative sound pressure experienced by a listener over a particular period of time. The acoustic dose D is:

number

number

[0049] It should be mentioned that the listener 40 referred to in this disclosure means the particular user currently using, i.e. currently listening to, the audio reproduction arrangement 10 .

[0050] A listener 40 is listening to the audio reproduction arrangement 10, which is described in more detail in other sections of this disclosure, and which is connected to a current sound pressure p c The latest sound pressure experienced by listener 40 is p c may be caused at least in part by the transducer device 15, or transducer 15 for short, of the audio speaker arrangement 10. c is preferably obtained from one or more microphones 17 of the audio reproduction arrangement 10.

[0051] Preferably, the latest sound pressure p c is an A-weighted sound pressure, but this is only one preferred embodiment. The A-weighted scale of sound pressure is the industry standard, and limits and recommendations are typically communicated as A-weighted sound pressure, or, when converted to SPL, as SPL presented in dBA. However, those skilled in the art will be familiar with the current sound pressure p c It will be appreciated that A does not have to be an A-weighted sound pressure, but may be an unweighted sound pressure or a sound pressure weighted in any other suitable manner. In this disclosure, sound pressure notation is preferred, but those skilled in the art will appreciate that this is interchangeable with SPL, so long as the correct notation is used in the respective calculations.

[0052] Latest sound pressure p c The acoustic dose D of the listener is calculated by accumulating the acoustic dose D over a dose period T as shown in Figure 2a. D The latest sound pressure p accumulated between c and the latest sound pressure p c 1 illustrates how samples of are accumulated over time for an acoustic dose D, which is measured over a dose period T D The latest sound pressure p squared over c, which may be described as a sliding average of the acoustic dose D. To avoid discomfort, injury or damage to the listener 40's burden, the acoustic dose D must not exceed an acoustic dose threshold L, shown by the dashed line in FIG. 2a. The blunt force approach simply controls the audio reproduction arrangement 10 to stop providing audio to the transducer 15, thereby muting the audio from the audio reproduction arrangement 10, and reduces the latest sound pressure p c to background sound. Such muting would have to be in effect until the acoustic dose D falls below the acoustic dose threshold L, which would severely degrade the audio experience of the listener 40.

[0053] FIG. 2b is a partial view of the graph shown in FIG. 2a, but showing the latest sound pressure p c At a first point in time a, the listener 40 increases the volume of the audio reproduction arrangement 10, resulting in a new sound pressure p c This increases the acoustic dose D, which is the solid line in Fig. 2b, and at the second point in time b, the actual latest period T c After that, the acoustic dose D exceeds the acoustic dose threshold L, which is the dashed line in FIG. 2b. In the blunt force approach, the listener 40 enters a dose period T at a third time point c. D However, the listener 40 cannot hear any more audio until the latest sound pressure p c Based on the acoustic dose D accumulated at the first time point a and the acoustic dose D accumulated at the first time point a, a latest estimated period T cu In Fig. 2b, the estimation is shown by the dotted line that crosses the dose threshold L at the fourth time point d. ce is the dose period T D Since the time is shorter than the time, the listener 40 can hear the latest sound pressure p c Assuming that the current sound pressure p c Updated sound pressure p u and the latest updated period T cu is the dose period T D, the listener 40 can enjoy the audio at substantially the same volume as before, without risking hearing damage. In FIG. 2b, this is shown by the dashed-dotted line that intersects the acoustic dose threshold L at a third point in time c. cu is the updated sound pressure p u and the latest volume D.

[0054] Based on the description given with reference to Figures 2a and 2b, some further details are provided in relation to Figure 1. In Figure 1, an audio profile 50 associated with a listener 40 is preferably provided. The audio profile 50 includes sound pressure data 52 for the listener 40. The sound pressure data 52 may include one or more cumulative sound pressures in the form of an acoustic dose D. Furthermore, the sound pressure data 52 may include one or more cumulative sound pressures in the form of an acoustic dose D. In addition, the sound pressure data 52 may include one or more cumulative sound pressures in the form of an acoustic dose D. D1,2,…,n Several different acoustic doses D are accumulated during 1,2,…,n The sound pressure data 52 may include data in the form of a cumulative sound dose D, i.e., the most recent sound dose D, but may additionally or alternatively include a series of previous sound pressure data experienced by the listener 40. The series of previous sound pressure data may be time-stamped data or may be averaged data samples indicative of the average sound pressure experienced by the listener 40 during a particular period of time. The sound pressure data 52, in other words, may be presented in many suitable ways and the disclosure is not limited to the particular examples given above.

[0055] The audio profile 50 may further include a hearing profile 54 of the listener 40. The hearing profile 54 may be an audiogram that describes the hearing capabilities of the listener 40 at one or more frequencies. The hearing profile 54 may be based on hearing tests performed by the listener 40 and / or on demographic data regarding, for example, age, gender, music preferences, etc.

[0056] Additionally, in some embodiments, the audio profile 50 further includes a predictive model 56 for the listener 40. The predictive model 56 may be described as an acoustic dose profile prediction 56 or a listening model 56. The predictive model 56 is a model that can be used to predict the listening behavior of the listener 40. The predictive model 56 is described in further detail elsewhere in this disclosure and may be formed in any suitable manner. In some embodiments, the predictive model 56 is based on a prediction of the most recent estimated period T ce , including time derivatives of previous sound pressure data that can be used to estimate . Additionally or alternatively, the predictive model 56 may be an average model. Additionally or alternatively, the predictive model 56 may be a recurrent neural network (RNN) model. Exemplary embodiments of the predictive model 56 may be combined to form a predictive model 56 that includes, for example, an average model and an RNN. It should be noted that the predictive model 56 does not have to be included in the audio profile 50, but may be determined at any time by any of the devices of the present disclosure.

[0057] For example, as described with reference to FIG. 2b, the audio profile 50 in general, and the sound pressure data 52 of the audio profile 50 in particular, are converted into the latest sound pressure p c , as taught herein, to provide a latest estimated time period T that represents the estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L. ce It is possible to calculate

[0058] More specifically, as shown in Figure 1, a controller 100 is provided for controlling the sound dose D experienced by a listener 40 of an audio reproduction arrangement 10. This is achieved by the controller 100 receiving an audio profile 50 associated with the listener 40 and a current sound pressure p from a microphone 17 of the audio reproduction arrangement 10. c This may be achieved by obtaining the audio profile 50 and the latest sound pressure p c Based on this, the controller 100 determines how long the estimated time period T ceThe listener 40 may be configured to determine or estimate the latest sound pressure p c can be continuously applied to itself. ce is the dose period T D If the time t1 is shorter than the time t2, the controller 100 updates the sound pressure p u The updated sound pressure p u is the updated sound pressure p u The latest update period T is estimated based on cu is the dose period T D The controller 100 preferably configures the updated sound pressure p u The latest sound pressure p c The audio playback arrangement 10 may be further configured to provide instructions 60 to enable control of the audio playback arrangement 10.

[0059] It should be mentioned that the audio reproduction arrangement 10 mentioned with reference to Fig. 1 may be any suitable audio reproduction arrangement 10 and is not limited to the audio reproduction arrangements 10 listed herein. Preferably, the audio reproduction arrangement 10 comprises or is operatively connected to an audio transducer 15 and a microphone 17. The one or more microphones 17 are arranged to receive the latest sound pressure p c The device is positioned to measure

[0060] 3a-d, exemplary embodiments of audio playback arrangements 10 according to embodiments of the present invention are presented. These embodiments are non-limiting and one of ordinary skill in the art, after reading this disclosure, will understand that many other shapes, forms or configurations of audio playback arrangements 10 are possible and usable with the teachings of this disclosure.

[0061] In Fig. 3a, the audio reproduction arrangement 10 is shown in the form of a pair of headphones 10. The audio reproduction arrangement 10 comprises at least one audio transducer 15, or transducer 15 for short. The transducer 15 may be any form of speaker 15 arranged to convert electrical signals into vibrations detectable by human ears 42, 44 (see Fig. 3d). The audio reproduction arrangement 10 generally comprises two transducers 15, although embodiments with only one transducer 15 or three or more transducers 15 are also feasible. The audio reproduction arrangement 10 further comprises one or more controllers 100. The controller 100 is arranged to control the audio transducers 15, but may very well perform other control, processing and / or communication related tasks. The headphones 10 comprise one or more microphones 17, preferably arranged at the ears 42, 44 of a user of the headphones 10. The microphones 17 are arranged to measure the current sound pressure p at the ears 42, 44 of the listener 40. c The microphone 17 may very well be configured to perform or assist other functions, such as providing noise measurements at the ears 42, 44 of the listener 40 that can be used for noise cancellation, etc.

[0062] 3b shows another embodiment of the audio reproduction arrangement 10. In this embodiment, the audio reproduction arrangement 10 also includes a transducer 15 and a current sound pressure p c3b, a pair of headphones 10 with a microphone 17 for measuring the audio quality of the audio reproduction arrangement 10, and a controller 100. In this embodiment, the controller 100 comprises or is operatively connected to a transceiver (not shown) configured to communicate with an electronic device 30 via an interface 20. The electronic device may be any suitable rendering device such as a portable electronic device 30, preferably a mobile phone 30 as shown in FIG. 3b, a portable music player, a tablet, a laptop, etc. The controller 100 may be configured to receive an audio stream via the interface 20. Alternatively or additionally, the controller 100 may be configured to transmit and / or receive operational data via the interface 20. The operational data may be any data related to the operation and / or control of the audio reproduction arrangement 10. The interface 20 may be a wired or wireless interface 20 and may further be a digital or analog interface.

[0063] 3c shows another embodiment of the audio reproduction arrangement 10. In this embodiment, the transceiver 35 is separate from the microphone 37, which still receives the current sound pressure p c 3c. In this particular embodiment, the microphone 17 is located in a headrest of a seat 18. The seat 18 may be a vehicle seat 18, an office chair, a lounge chair, etc. In the particular embodiment shown in FIG. 3c, the seat 18 is a vehicle seat in the form of a car, and the transducer 35 is an audio speaker of the car. The controller 100 may be a separate controller 100, or may be a controller 100 configured to perform other processing tasks of the car. The controller 100 may comprise an interface 20 that operatively connects the audio playback arrangement 10 to, for example, a remote server (described in other sections of this disclosure).

[0064] In Fig. 3d, yet another embodiment of the audio reproduction arrangement 10 is shown. In this embodiment, the transducer 15 and the microphone 17 are arranged to be operatively connected to the controller 100, rather than being physically connected thereto. The microphone 17 is preferably arranged proximate to the ears 42, 44 of the listener 40, so that the current sound pressure 70 at the ears 42, 44 of the listener 40 can be measured. Each of the transducer 15 and the microphone 17 may be a separate device, the transducer 15 being, for example, a Bluetooth speaker and / or a smart speaker, and the microphone 17 being, for example, part of a hearing aid, a headset or a standalone microphone 17 configured to be worn proximate to the ears 42, 44 of the listener 40. The controller 100 may be a portable electronic device 30, such as a mobile phone, configured to communicate with the microphone 17 and the transducer 15 directly via a local connection (WIFI, Bluetooth, etc.) or indirectly via a cloud server. The controller 100 may be, for example, a connectivity gateway configured to provide smart home functionality, voice control, etc. In some embodiments, the controller 100 is included in a cloud server 210 (see FIG. 4).

[0065] 3a-d, audio reproduction arrangements 10 may be arranged in a number of different ways, and the illustrated embodiments may very well be combined or extended in any suitable manner. One common feature between audio reproduction arrangements 10 is that the final sound pressure p experienced by listener 40 is c, which may be achieved by positioning the microphone 17 proximate to the ears 42, 44 of the listener 40. In some embodiments, for example when the audio reproduction arrangement 10 is a pair of in-ear headphones 10, the microphone 17 may be positioned in the ear cavity or ear canal of the listener 40. The microphone 17 may be positioned very far from the listener 40, for example with directional control to measure the current sound pressure p at the ears 42, 44 of the listener 40. c It should be mentioned that the microphone 17 may be optionally an array of microphones configured to measure the current sound pressure 70 at the ears 42, 44 of the listener 40.

[0066] Embodiments of the present disclosure are generally described with the audio playback arrangement 10 configured to receive an audio stream from, for example, a portable electronic device 30. However, in some embodiments, the audio playback arrangement 10 may be configured for a hear-through mode in which sounds sensed by at least one of the microphones 17 per ear 42, 44 may be further enhanced according to an audio profile 50 and rendered by the headphone transducer 15.

[0067] Most embodiments have one acoustic dose D, one dose threshold L, and one dose duration T. D However, all embodiments may be extended to include more than one acoustic dose D, where each acoustic dose D has a dose threshold L and a dose duration T D It may be associated with.

[0068] With reference to Fig. 4, a system 200 for controlling the sound dose D of a listener 40 is described. The system 200 may comprise a cloud server 210 operatively connected to a storage 220, which may be a remote storage 220. The storage 220 may alternatively or additionally be located partially or entirely in the audio reproduction arrangement 10 and / or the portable electronic device 30 presented herein. The storage 220 may be configured to store an audio profile 50 including sound pressure data 52 of the listener 40. The system 200 further comprises one or more audio reproduction arrangements 10, each of which comprises a transducer 15 and a microphone 17, where the microphone 17 is configured to receive and store the latest sound pressure p experienced by the listener 40. c 4 , the controller 100 is shown as being located on the cloud server 210 and the audio playback arrangement 10. This is only one exemplary embodiment, and it should be understood that the controller 100 may also be standalone or may be located partially or entirely on the portable electronic device 30. In any case, the one or more audio playback arrangements 10 are operably connected to the cloud server 210 and to the portable electronic device 30, if present. The one or more audio playback arrangements 10 may in some embodiments be operably connected to the cloud server 210 via the portable electronic device 30.

[0069] Preferably, the remote storage 220 is configured to store audio profiles 50 of one or more listeners 40. The cloud server 210 may be configured to access and / or control the remote storage 220 to retrieve and / or store audio profiles associated with particular listeners 40. The cloud server may receive, for example, a request from one or more devices of the audio reproduction arrangement 10 to provide an audio profile of a particular listener 40. The audio reproduction arrangement 10 may be configured to identify the listeners 40 such that each particular listener is differentiated from other listeners 40. One such way is that the listener 40 must identify himself / herself in front of the audio reproduction arrangement 10 before he / she can hear the audio. Alternatively or additionally, the audio reproduction arrangement 10 may be configured to, for example, measure the acoustic footprint of the ear canal of the listener 40 and identify the listener 40 based on the acoustic footprint.

[0070] The configuration of the system 200 as described above allows several listeners 40 to share the same audio reproduction configuration 10 and / or the same listener 40 to use different audio reproduction configurations 10, and still allow the acoustic dose D to be greater than or equal to 100% for a given dose period T D This is highly beneficial as it ensures that the dose threshold L is not exceeded over the entire listening period. By way of example, a listener 40 may enjoy music through a pair of headphones 10 that access the listener's audio profile 50 via a cloud server 210. As listening progresses, the headphones 10 update the current sound pressure p c and / or the acoustic dose D may be shared with the cloud server 210. The listener 40 may, over time, change the audio reproduction configuration 10 to a different pair of headphones 10, to a car audio reproduction configuration 10 (FIG. 3c), or to a home theater 10 (FIG. 3d). Any of the audio reproduction configurations 10 may identify the listener 40, access the listener's audio profile 50 via the cloud server 210, and retrieve the latest sound pressure p experienced by the listener. c Based on the latest sound pressure p c so that the latest sound pressure p does not exceed the dose threshold L.c In other words, regardless of which audio reproduction configuration 10 the listener 40 selects to use, the teachings of the present disclosure allow for the acoustic dose D to be controlled such that it does not exceed the dose threshold L.

[0071] As is evident from the description of system 200 in Fig. 4, the division into subcomponents, modules, devices and blocks is not fixed. In other words, one or more logical devices may be included in one or more physical devices. The logical division presented in Fig. 4 is provided for illustrative purposes, and a person skilled in the art, after digesting the teachings of the present disclosure, can build several different systems 200 having different layouts but with an equal purpose, i.e., to control the acoustic dose D of listener 40.

[0072] In one embodiment, the portable electronic device 30 stores or receives audio data that in some embodiments is referred to as the rendering device 30 and is decoded and relayed to the controller 100. The rendering device 30 may further receive and transmit to the cloud server 210 an audio profile 50 for the listener. The audio profile 50 includes sound pressure data 52 of the listener 40, but may very well include further data regarding, for example, the hearing and especially the hearing health of the listener 40. As mentioned above, the rendering device 30 may be a portable electronic device 30, a mobile phone, a car stereo, or any other with the ability to connect to a remote location, receive, unpack and / or decode audio. The unpacked and / or decoded audio may be referred to as a source signal S (see FIG. 5) and may be a mono, stereo or any other form of multi-channel signal.

[0073] The controller 100 may be configured to process an audio stream, preferably a stream having a source signal S coming from the rendering device 30 together with a microphone signal 17' (see FIG. 5) coming from the microphone 17. The controller 100 may process the source signal S using processing configured based on one or more of the microphone signal 17', the data of the audio profile 50, and / or any other suitable processing parameters. The processing may be mono or stereo. The processed source signal 15' (see FIG. 4) is sent to the transducer device 15. The processed source signal 15' may be digital or in analog form, and if in analog form, the transducer 15 is preferably configured to convert the digital audio signal 15' into an analog audio signal and output the analog signal to one or more physical transducers 15. The transducer 15 may be mono, e.g. a Bluetooth speaker, or stereo, e.g. a pair of headphones. In a car stereo or home theater, there may be multiple transducers 15.

[0074] The recording device 17 consists of at least one microphone 17, which, as previously mentioned, is preferably located at an ear reference (ER) point, i.e., physically close to the ears 42, 44 of the listener 40. In a pair of headphones 10, there are preferably at least two microphones 17, each recording audio at the left and right ER points respectively.

[0075] The cloud server 210 of the system 200 of Fig. 4 may be referred to as a server 210, a remote server 210, or a cloud service 210. The cloud server 210 may be configured to receive hearing health data, as appropriate, which is preferably included in an audio profile 50 for the listener 40. The hearing health data, or hearing data for short, may be included in the audio profile 50 and calculated based on usage by the listener 40, e.g., listening level and duration. That is, the server 210 may receive the acoustic dose D, the latest sound pressure p c , and / or any other sound pressure data 52. The cloud service 210 may be configured to store in the storage 220 all or part of the received hearing data associated with the listener 40. The server 210 may further be configured to perform an analysis of the hearing data and to store the results of the analysis, preferably in the storage 220. The results of the analysis may be made accessible to the listener 40 so that the listener 40 can access and view the hearing health, i.e. audio profile 50, associated with the listener 40. The server 210 may be configured to provide an updated hearing health model and an acoustic dose D budget, i.e. how much cumulative sound pressure the listener 40 can be exposed to without exceeding the acoustic dose threshold L. All of this may be included in the listener's audio profile 50. The hearing health model is preferably included in the audio profile 50 and may be provided to the rendering device 30 and / or audio reproduction arrangement 10, for example according to a predefined scheme or upon request from the rendering device 30 and / or audio reproduction arrangement 10. The predefined scheme may be instantiated, for example, at the first connection of the day, etc.

[0076] In Fig. 5, an exemplary signaling block diagram can be used, for example, with any of the embodiments of the controller 100, system 200 and / or audio playback arrangement 10 of the present disclosure. The specific details and embodiments presented with reference to Fig. 5 can be used with any embodiment of the present disclosure and are not limited to this specific example. For illustrative purposes, as a non-limiting example, assume that the block diagram of Fig. 5 is used on the system 200, where the first frame N of a sample is p is conveyed from the rendering device 30 to the transducer device 15 via a processing device, which may be a controller 100. Furthermore, sample N r is transferred from recording device 17 to processing block 410.

[0077] The sample rates and frame sizes N of the various signals in the system 200 p and N r may vary depending on, for example, hardware capabilities, signal quality and / or power consumption requirements. Sample rate alignment and sample alignment may be easily obtained using well-known resampling techniques, for example based on anti-aliasing filtering and sinc interpolation. Without loss of generality, N p =N r Both are typical operations for sample alignment and resampling operations.

[0078] The system 200 of Figure 2 is shown using two transducers 15 and two microphones 17. It is assumed that the system 200 includes an audio reproduction arrangement 10 as shown in Figure 3b. It is assumed that the system 200 is configured for stereo or two-channel reproduction. One channel, or mono reproduction, is a special case where the source signal S is split into two identical stereo channels before being fed to the transducers 15.

[0079] At the top of Figure 5, a source signal S may be received, for example from a rendering device 30 (not shown in Figure 5), and processed by enhancement in processing block 410. The enhancement may be time and frequency adaptive and may be controlled by processing control block 415. Once the source signal S has been enhanced, the resulting audio signal 15' is relayed to transducer 15 for playback.

[0080] Each channel of the source signal S may be processed separately. The processed source signal may be relayed to one or more transducers 15, e.g., in the case of a pair of headphones 10, a mono source signal S is split into two channels that are processed separately and rendered on corresponding left and right headphone transducers 15. The same is true for transducers 15 located near each ear 42, 44, e.g., in a car seat 18.

[0081] At points in the system where no further processing is applied to the digital transducer signal 15', i.e., only DA conversion remains, the transducer signal 15' may be provided to the processing control block 415 as a signal reference. The signal reference can be used, for example, for signal separation using linear echo cancellation.

[0082] Microphones 15, preferably at least one microphone 17 per ear 42, 44, sense (e.g., record, measure, etc.) sound reaching each ear 42, 44. For headphones 10 using active noise control (ANC), the microphones 17 are generally positioned completely within the ear headphone cavity, adjacent to the transducer 15. In an automobile, the preferred location for each microphone is in the head support of the seat 18, preferably one microphone 17 on each side of the listener's 40 head.

[0083] Each microphone signal 17' represents a signal describing the sound received by one ear 42, 44. Using the microphone signals 17', an acoustic dose D is calculated, preferably separately for each ear 42, 44, in an acoustic dose calculation block 420, as described elsewhere in this disclosure. The calculated acoustic dose D is stored. The acoustic dose D may be stored locally in a local storage 220 of the audio reproduction device 10 and / or the rendering device 30. Preferably, the calculated acoustic dose D is additionally stored (i.e. backed up, mirrored, replicated, etc.) in a server 210 or a remote storage 220 operatively connected to the cloud service 210. The calculated acoustic dose D may be transferred by the rendering device 30, i.e. the portable electronic device 30, to the cloud service 210. As mentioned above, the acoustic dose D is calculated over several different dose periods T D The rate of sampling of the microphone signal 17′, the amount of available storage, the available bandwidth for transmitting the acoustic dose D, etc., determine the minimum period T over which the acoustic dose is preferably calculated in order to accurately and precisely control the enhancement of the source signal S. m It may be useful in determining the minimum period T m An exemplary, non-limiting setting for is 1 minute.

[0084] In one embodiment, the calculation of the acoustic dose D is performed for each audio frame and has a minimum duration T m Minimum acoustic dose D per m The doses are then aggregated to form another dose period T D The acoustic dose D for is the minimum acoustic dose D m It is calculated as the sum of the number of dose periods T D is 1 hour, and the minimum period T m If is 1 minute, the acoustic dose of sound per hour is

number

[0085] According to the aforementioned ITU-T H.870 entitled "Guidelines for safe listening devices / systems", the recommended acoustic dose D R is the 7-day dose period T D For adults, it is given by D per week. R7 The recommended acoustic dose per day (7 days) is D R7 =1.6Pa 2 This may be calculated as: This is usually known as the reference dose. R7 corresponds to an average SPL of 80 dBA during 40 hours of listening.

number

number

number

[0086] The acoustic dose control block 430 may be configured to upload the acoustic dose D to the server 210 or the cloud service 210 at a specific time, e.g. periodically, or upon request, e.g. from the rendering device 30 or the storage 220. The acoustic dose control block 430 may further be configured to request, determine, and / or directly receive a model for the acoustic dose profile prediction 56, i.e. a prediction model 56. This model 56 may be included in the audio profile 50 as described above or may be determined based on historical and / or latest sound pressure data 52 for the listener 40. The acoustic dose control block 430 may further be configured to request, determine, and / or directly receive an updated prediction model 56.

[0087] Using the predictive model 56, the acoustic dose control block 430 determines the dose duration T D It may be configured to check whether the latest accumulated acoustic dose D during a dose period T results in the exceedance of an associated dose threshold L, i.e. the recommended maximum dose. D Or, as mentioned above, predict whether the current estimated period T ce is the dose period T D Predict whether the field is shorter than the dose threshold L (T ce >T D ), no action needs to be taken. If the prediction result exceeds the dose threshold L (T ce <T D ), latest sound pressure p c Actions may be taken to lower

[0088] Functions, blocks, and features are described herein with reference to, for example, processing, control, calculation, etc., and may be implemented in hardware, software, or a combination thereof. The processing block 410, the processing control block 415, the acoustic dose calculation block 420, the acoustic dose control block 430, and / or the storage 220 may be included in the controller 100. The blocks may form part of a distributed system, or alternatively or additionally, some functions or blocks may be fully or partially included within the controller 100.

[0089] FIG. 6 shows a non-limiting example presented for further disclosure of the working of the embodiments of the present disclosure. In FIG. D is 168 hours (7 days), and the dose period T D The maximum recommended dose during radiation exposure, i.e., the dose threshold L, is 1.6 Pa. 2 h. The lower dashed curve in Figure 6 represents the dose period T D As can be seen, this dashed and dotted curve It does not reach the maximum critical dose threshold L. The long curve above the dashed and dotted line shows the dose profile during the dose period T DThis is an example of a hearing behavior in which the acoustic dose D exceeds the dose threshold L before the time has elapsed. In the long term, this can lead to hearing damage.

[0090] Continuing to refer to FIG. 6, to illustrate the prediction of acoustic dose D, the solid curve shows the acoustic dose D predicted by a prediction algorithm according to the teachings of the present disclosure at a current time T c The evaluation result is shown by the upper dotted curve in Figure 6. c The estimated acoustic dose D based on the evaluation in the dose period T D The dose threshold L is clearly exceeded before the end of the ce <T D ) In response, the adaptive frequency transfer function of processing block 410 is adjusted, i.e., a new estimated acoustic dose D is calculated based on the adjusted frequency transfer function, and the resulting hearing profile is shown in the dashed curve of FIG. 6. As can be seen in FIG. 6, the adjustment is performed such that the dashed curve is adjusted for a dose period T D The dose threshold L is not exceeded within the ce >T D ), thereby reducing the risk of damaging the hearing of the listener 40. The frequency transfer function may further be adjusted to maximize the intelligibility of the audio stream, for example taking into account background noise, the user's hearing impairment, etc.

[0091] The adjusted frequency transfer function is the latest sound pressure p c It should be noted that the processing block 410 may not only include attenuation across all frequencies that effectively reduces the attenuation of the signal S. Alternatively, the processing block 410 may be configured to calculate an appropriate adaptive processing, i.e., enhancement, based on the source signal S, the background noise, the hearing impairment of the listener 40, the cumulative acoustic dose D, and / or processing parameters. The processing parameters may include, but are not limited to, an audiogram describing the hearing capabilities of the listener 40 and / or an intelligibility mode. The processing parameters may be included in an audio profile 50 for the listener 40. The intelligibility mode parameters may be configured to indicate which intelligibility measures should be improved, the intelligibility measures being presented in more detail elsewhere in this disclosure.

[0092] In the following, some additional implementation details are given that can be used in any of the embodiments presented herein. Assuming a standard calibration of the microphone 17, the latest sound pressure p c may be calculated based on the signal from the microphone 17. Thus, the dose time T m Acoustic dose in D m teeth,

number

[0093] Dose Period T D The prediction of the acoustic dose profile in the current time period is based on the historical acoustic dose data, a feature set, e.g., time, date, volume index, and device identification, and the most recent dose period T D The acoustic dose D may be based on one or more of the current and past values ​​of the acoustic dose D, and combinations and equivalents thereof.

[0094] One exemplary predictor is the average of several previous dose profiles, i.e.

number

[0095] A second exemplary predictor is the maximum-based predictor

number

[0096] In one embodiment using a predictor, the dose period T D The acoustic dose D at the end of

number

[0097] A third exemplary predictor is a regression model. The regression model may be a recurrent neural network (RNN), a Markov chain, or an autoregressive process. The RNN may be based on a long short-term memory (LSTM) or a gated recurrent unit (GRU). The RNN with LSTM is suitable for multivariate multi-stage time series prediction.

[0098] In general, the vectors of the RNN and the output vectors may be selected in a number of ways, and the two choices, the size of the RNN, along with the training data quality and the training of the RNN, determine the performance of the RNN and its ability to output desired output data given input (feature) data.

[0099] The feature set is, as described above, DThe input may include the past and current values ​​of the acoustic dose D in the audio device 10, the date and time, a volume index, and a device identification, e.g., an identifier for the audio device 10. The input of the acoustic dose D is beneficial for the work of the predictive model, in addition, the date and time may affect the dynamic behavior, since an increase in the acoustic dose D may be time-dependent, e.g., measured only during office hours, on weekends, etc.

[0100] In addition, prolonged exposure to sound may saturate and / or temporarily impair the hearing of the listener 40. As a result, the listener 40 may be inclined to increase the volume of the audio reproduction arrangement 10. Furthermore, depending on the time of day, the listener 40 may be located in different locations with more or less background noise, for example visiting a loud cafeteria during lunch. Therefore, it may be beneficial to use one or more of the hour, day and / or volume index to predict the dose profile.

[0101] Predicted dose profile

number

number

number

number

[0102] The calculation of the gain G(k) is performed at regular intervals, e.g., a minimum period T m Considerations for the gain G(k) may include, for example, the update rate versus the stability of a frequency transform of the source signal S. Large gain changes in an audio signal relatively quickly can result in degradation of audio quality.

[0103] In some embodiments, the gain

number

number

[0104] Gain G(k) and the subsequent maximum SPL L spl Additional embodiments of may include, for example, rules describing whether applying the gain G(k) results in a reduction in sound pressure. In some embodiments, the latest sound pressure p c The limit for the acoustic dose D is the latest sound pressure p c and the acoustic dose threshold L. In a further embodiment, the controller 100 may be invoked only if a limiting threshold between p c may be configured to limit the latest sound pressure p by attenuating higher frequency components of the source signal S etc. cFor those skilled in the art, a set of rules controlling the sound pressure limit can be easily established after digesting the teachings of this disclosure.

[0105] In some embodiments, the acoustic dose profile prediction 56 is D In these embodiments, the latest sound pressure p c may be performed at a certain time index k where the listener 40 typically experiences a higher average sound pressure. Additionally or alternatively, the latest sound pressure p c The reduction of may be performed by adjusting the frequency weighting of the source signal S such that unimportant frequency components and / or frequency components that cannot be heard by the listener 40 are removed or attenuated, so that unimportant frequency components and frequency components outside the range that the listener 40 can hear do not contribute to the acoustic dose. By doing this, the risk that the acoustic dose D will exceed the dose threshold L and thereby damage the hearing of the listener 40 is reduced without compromising the audio experience of the listener 40.

[0106] As shown in FIG. 5, the SDC 430, which may be included in the controller, preferably performs the acoustic dose profile prediction 56 and asserts that the prediction model 56 is correct. When data is sent to the server 210 or cloud service 210, the acoustic dose profile prediction 56 may be updated remotely and downloaded to the SDC 430 via the rendering device 30. Depending on the type of prediction model 56, the update or construction of the prediction model 56 may additionally or alternatively be performed locally, for example, by the processing block 430 or the rendering device 30. The average prediction model exemplified above may be updated recursively and may require less processing and storage when updated locally. The RNN prediction model 56 generally requires more resources and is preferably updated remotely. The prediction model update may be triggered, for example, periodically at set time intervals or when set stability conditions are not met. The predictor model 56 is preferably associated with the listener 40 and, as a result, changes when the listener 40 changes.

[0107] Further implementation details below are provided that can be used in any of the embodiments presented herein. As mentioned elsewhere in this disclosure, the server 210 or cloud service 210 may be configured for remote storage of user data and acoustic dose data, i.e., audio profile 50 for listener 40. The predictive model 56 may be updated and / or retrieved by the portable electronic device 30. The predictive model 56 may be transmitted as parameters, e.g., parameters of an RNN or an autoregressive process. The server 210 or cloud service 210 may be described in some embodiments as part of the system 200 for providing storage and / or computation of the predictive model 56.

[0108] The processing control block 415 may estimate the stability of the acoustic environment for each microphone signal 17' using the transducer signal 15' and the microphone signal 17'. This can be accomplished in a number of ways. In one embodiment, this is done by evaluating the short-term power spectrum change per time unit per frequency band. In another embodiment, this is done by comparing the SPL averaged over a short period of time with the SPL averaged over a relatively long period of time, whereby equal amounts indicate at least a semi-stationary acoustic environment.

[0109] The processing control block 415 may further apply an auditory model, preferably an auditory profile 54 of the listener 40, to the microphone signal 17' to model how the listener 40 perceived the sound. Parameters for enhancing the source signal S may be calculated based on the perceived sound and / or the auditory model 54.

[0110] The microphone signals 17' are preferably transformed from the time domain to the frequency domain by a discrete frequency transform (DFT), i.e. using an FFT implementation of the DFT. For each 1 / 3 octave frequency band, adjacent frequency bins of the DFT are amplitude mean squared accumulated such that a resulting 1 / 3 octave frequency representation is obtained for all signals.

[0111] The estimation of acoustic propagation may be performed in several ways. In some embodiments, the transducer signal 15' is input to a linear echo cancellation function, for example based on least mean squares (LMS) adaptation. This may be usable to estimate the echo path, i.e. the acoustic path between the transducer 15 and the microphone 17. In some embodiments, the acoustic path is estimated as a secondary path used in ANC applications. In some embodiments, the acoustic path is measured using acoustic measurement equipment, modeled as a frequency function or filter, and stored for use during operation. The latter is a simpler arrangement compared to linear echo cancellation or ANC, but is less flexible and may require one model for each acoustic path considered under normal operation.

[0112] To estimate the background noise, the microphone signal 17' may in some embodiments be split into propagated source signals (echoes) and background noise via signal separation. Signal separation may be performed, for example, using an echo controller, and the propagated source signals may be removed using signal subspace decomposition, for example the Karhunen-Loeve transform, which is suitable since the covariance matrix of the source signals is known and the approximate echo path is obtained from acoustic propagation calculations.

[0113] In some embodiments, background noise may be measured as the microphone signal 17' when the source signal S is zero or when the source signal S is below a set threshold. If the source signal S is speech, there are natural pauses in the speech signal, but if it is music, this may not be the case. In the case of music, the estimation of the background noise is preferably performed by removing the propagating source signal S from the microphone signal 17'. The latter feature requires more resources than using pauses in the source signal S, but can also be used for voice signals.

[0114] The hearing model 54 is preferably configured to receive as input the microphone signal 17'. The microphone signal 17' may be converted to an equivalent sound pressure value, for example using a conversion function from the calculation of the acoustic dose D. Then, an audiogram of the listener 40, i.e., a frequency and level dependent gain that models how the listener 40 perceives the source signal S depending on the hearing model 54, may be calculated. For each 1 / 3 octave frequency band, the band pass level may be measured and a gain may be calculated that represents the attenuation of the signal that the listener 40 would experience due to hearing impairment. In some embodiments, the gain H(b) per frequency band may be calculated as shown below:

number

[0115] The microphone signal 17' may be converted to an equivalent sound pressure value, for example using a conversion function from the calculation of the acoustic dose D.

[0116] There are many hearing models 54 that can be used in this disclosure, and the teachings herein are not limited to the listed models. The hearing model 54 may be configured to model hearing impairments that consider either or both conductive and sensorineural hearing loss. Conductive impairments are generally the result of, for example, blockage of the ear canal and are manifested as a loss of sound volume equal to the degree of impairment. Conductive hearing loss is often successfully treated by medical or surgical means. Sensorineural hearing loss is generally due to a decrease in sensory capacity in the middle ear, and more commonly due to a decrease in sensory capacity in the inner ear due to defects in the cochlea. In general, this class accounts for most cases of hearing loss. Causes of sensorineural hearing loss are, for example, genetic factors, disease, high levels of noise or sound exposure, and the aging (presbycusis) process.

[0117] The auditory model 54 presented above approximates loudness supplemented hearing loss. One skilled in the art, after considering this disclosure, could replace and / or extend the auditory model 54 with additional considerations, such as reduced amplitude sensitivity, reduced frequency range, loss of dynamic range, loss of distinction of spectral detail, and / or any combination of the foregoing.

[0118] There are several methods in the art that can be used to estimate speech intelligibility, such as signal-to-noise ratio (SNR), speech intelligibility index (SII) or coherence signal intelligibility index (CSII), short-term objective intelligibility (STOI), etc. Most of the intelligibility methods use a similarity measure, i.e., correlation or coherence between the source signal S (reference) and the processed signal (signal in noise or otherwise degraded source signal). The choice of the method for intelligibility may affect most of the processing blocks 100, 410, 415, 420, 430, but for those skilled in the art, this is within the scope of their expertise. In this disclosure, a correlation-based intelligibility measure is presented, but those skilled in the art can easily substitute intelligibility measures.

[0119] For speech, the correlation coefficient in the frequency domain between the source signal S and the hearing-impaired microphone signal 17' is considered as an intelligibility measure. A relatively high correlation generally corresponds to a high intelligibility, and a relatively low correlation generally corresponds to a low intelligibility. Therefore, a relatively high correlation is desired for each frequency band. For music, a correlation-based intelligibility measure is less obvious and may lose components in the objective intelligibility method. However, the SNR-based method is intuitively appropriate, and the application of this specification does not depend on absolute values, so a correlation measure is also appropriate. The processing component may be composed of frequency-selective and adaptive enhancement subcomponents that use, for example, enhancement parameters for configuration.

[0120] In the following, the desired gain per frequency band is calculated as an exemplary embodiment that can be used with other teachings of the present disclosure. The desired gain per frequency band should be applied to the source signal S using a finite impulse response (FIR) filter with linear phase or minimum phase. Furthermore, it may be a parameterized equalization consisting of a set of infinite impulse response filters, for example. However, the latter may cause relatively high audio distortion depending on the design. Therefore, the FIR filter is the preferred choice. The following calculation is described for one channel for simplicity, but the extension to two channels is obvious to those skilled in the art. For each frame of samples of the source signal S and the microphone signal 17', a calculation of new weights per frequency band W(b) is preferably performed. Let the microphone signal 17' representing the sound pressure level per frequency band be M(b) and the source signal S representing the sound pressure level per frequency band be X(b). Then M(b)=E(b)H(b)W(b)X(b)+H(b)V(b), where E(b) is the echo path frequency transfer function, where H(b) is the attenuation due to the hearing impairment of the listener 40 (contained in the hearing profile 54), W(b) is the (adaptive) frequency gain applied by the processing block 410, and V(b) is the background noise received by the microphone 17. In matrix form, the same equation is M=EHWX+HV, where M, X, and V are vectors with each element corresponding to a band b, and E, H, and W are diagonal matrices with all elements zero except for the diagonal. In summary, E is a matrix describing the echo path such that EX is equal to the source signal S measured by the microphone 17.

[0121] As an example of a measure of intelligibility, the sum of the diagonals of the correlation matrix between the source signal and the microphone signal is used, i.e., the intelligibility is expressed as I=tr(T x R XM ) in which R XM =E(X T M)=EHWR XX +HR XV and R XVis the correlation matrix between the source signal and the background noise, and T x is the weight matrix. R XX is the correlation matrix of the source signal S, which is not normalized so that the diagonal elements correspond to the squared sound pressure per frequency band. Using this intelligibility measure, the source signal S can be frequency-adjusted using W, but not R. XV is unaffected. In other words, W is a frequency gain matrix, which is preferably applied to source signal S in processing block 410.

[0122] To optimize the selected intelligibility and take into account hearing impairment, background noise and acoustic dose, and to XV Since is independent of the weight W, we can formulate the optimization as follows: w=arg max I=arg max tr(T x R XM )=arg max tr(T x EHWR XX )

[0123] By adding constraints on the weights and amplifications, the optimization may be formulated as a second order cone program, where the constraints change depending on whether the sound pressure is limited or not.

[0124] When sound pressure is limited:

number

[0125] Or, if sound pressure is not limited:

number

[0126] The amplification constraint Γ may be set by configuration, may be part of the processing parameters, and may be a condition where an increase in level is expected when adjusting the hearing profile 54 of the listener 40. T v ,T xis a weighting matrix used to emphasize some frequency bands relative to the rest. It is typically used in speech applications where speech frequencies have different importance to intelligibility. Noteworthy weighting matrices T v ,T x elements (frequency bands) are set to 1, other less important elements range from 1 to 0. Typically the important frequency bands are 100-4000 Hz, 4-8 kHz is important for comfort and partially for intelligibility, 8-12 kHz provides some detail but little intelligibility, and content above 12 kHz is reserved for study.

[0127] In some embodiments, a limit to the maximum gain may be added:

number

[0128] The frequency gain W may then be converted to an FIR filter using frequency sampling methods. Since the desired frequency response of the filter is given by W, other design methods or equivalent filter configurations are possible.

[0129] Preferably, the adaptive processing block 410 includes an FIR filter configured to process frames of the source signal S.

[0130] 7, a method 300 for controlling sound dose for a listener 40 according to an embodiment of the present disclosure is presented. The method 300 may be extended to further comprise any suitable features enumerated within the present disclosure. The method includes a step 310 of obtaining a previously presented audio profile 50 including sound pressure data 52 for the listener 40. The audio profile 50 may be obtained from a remote or local storage 220. If an audio profile 50 is not available, the step 310 of obtaining the audio profile 50 may include creating the audio profile 50. The method 300 includes a step of obtaining a previously presented audio profile 50 including sound pressure data 52 for the listener 40. The audio profile 50 may be obtained from a remote or local storage 220. If an audio profile 50 is not available, the step of obtaining the audio profile 50 may include creating the audio profile 50. The method 300 includes a step of obtaining a previously presented audio profile 50 including a sound pressure data 52 for the listener 40. c The method further includes a step 320 of acquiring the sound pressure data 52 and the latest sound pressure p. This may be performed by any means disclosed within this disclosure or any other suitable means available to a person skilled in the art when performing the method 300. c Based on this, the method 300 calculates the latest estimated period T ce The latest estimated period T ce represents the estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L. Typically, as taught within this disclosure, multiple acoustic dose thresholds L may be used. Each acoustic dose threshold L corresponds to one distinct dose period T D and / or may be associated with one separate set of sound pressure data 52. As taught, the latest estimated period T ce The estimation 330 of may optionally be based on one or more predictive models 56 , and the estimation 330 may include determining 335 the predictive models 56 .

[0131] Latest estimated period T ce is the dose period T associated with the acoustic dose threshold L. D If the sound pressure is below the threshold, the method 300 uses the sound pressure data 52 and the latest sound pressure p c Based on the updated sound pressure p u As described in this disclosure, the updated sound pressure p u is the latest updated period T cu is the dose period (T D) or greater. The method 300 calculates the updated sound pressure p cu The latest sound pressure p c The method may further include providing 350 instructions 60 for controlling the

[0132] Updated sound pressure p cu The latest sound pressure p c It should be mentioned that the instructions 60 for controlling may be any suitable instructions 60 that can be used directly by the audio reproduction arrangement 10 to reduce the signal level provided to the transducer 15. In one embodiment, the instructions 60 are preferably provided to the audio reproduction arrangement 10, and in particular to the controller 100 of the audio reproduction arrangement 10. Additionally or alternatively, the instructions 60 may be an instruction or prompting to the listener to reduce the playback volume of the audio reproduction arrangement 10. The instructions 10 may be an audible instruction reproducible by the audio reproduction arrangement 10 and / or a visual instruction presented by, for example, the portable electronic device 30.

[0133] The resulting audio profile 50 and sound pressure data 52 usable in the method 300 may be in any suitable shape or form and may follow any embodiment or feature presented within this disclosure.

[0134] In some embodiments of the method 300, the audio profile 50 may include a hearing profile 54 of the listener 40. In such embodiments, the latest estimated period T ce Further, based on the hearing profile 54 of the listener 40, the latest sound pressure p u may be based on the hearing profile 54 by a frequency weighting step 325. The frequency weighting step 325 is carried out based on the latest sound pressure p c This may be performed as part of step 320 of obtaining

[0135] The method optionally includes: cThe method may include a step 360 of updating the sound pressure data 52 of the audio profile 50 for the listener 40 based on the

[0136] In Fig. 8a-e a controller 100 for controlling the audio reproduction arrangement 10 is shown. The controller 100 may be configured to cause, for example, the method 300 described with reference to Fig. 7 to be performed. Generally, the controller 100 may be configured to cause (e.g. execute) one or more steps of the method 300 described with reference to Fig. 7 according to any embodiment. This means, for example, that the actual obtaining, provisioning, calculation etc. may not be performed by the controller 100, but the controller may be configured to cause other devices, modules or blocks to perform the actual operations. However, the controller 100 may very well be configured to cause one or more steps of the method 300 described with reference to Fig. 7. In a preferred embodiment, the controller 100 is configured to cause one or more steps of the method 300 described with reference to Fig. 7 to be performed, and to cause the remaining steps of the method 300 described with reference to Fig. 7 to be performed.

[0137] For example, as shown in Figure 8b, the controller 100 may be configurable, e.g. included, in a first device 10. The first device may be an audio reproduction arrangement 10, e.g. any of the audio reproduction arrangements 10 presented with reference to Figures 3a-d.

[0138] Alternatively or additionally, the controller 100 may be configurable, e.g. included, in an electronic device 30, as shown in Fig. 8c. The electronic device 30 may be the portable electronic device 30 of Fig. 3b, Fig. 4 and / or the rendering device 30 described with reference to Fig. 5. The electronic device 30 may be any electronic device 30 suitable for controlling the acoustic dose D of a listener 40.

[0139] Alternatively or additionally, the controller 100 may be configurable, for example included, in a server 210 or a cloud server 210, as shown in Fig. 8d. The cloud server 210 may be the cloud server 210 of Fig. 4. The cloud server 210 may be any server 210 suitable for controlling the acoustic dose D of the listeners 40.

[0140] Alternatively or additionally, the controller 100 may be configurable, for example included, in a system 200. The system 200 may be any system 200 suitable for controlling an acoustic dose D of a listener 40, for example the system 200 presented with reference to Figures 4 and 5.

[0141] As one skilled in the art will understand after contemplating the teachings of the present disclosure, the described embodiments and their equivalents may be implemented in software or hardware, or a combination thereof. The embodiments may be performed by general-purpose circuitry. Examples of general-purpose circuitry include digital signal processors (DSPs), central processing units (CPUs), co-processing units, field programmable gate arrays (FPGAs), and other programmable hardware. Alternatively or additionally, the embodiments may be performed by dedicated circuitry, such as application specific integrated circuits (ASICs). The general-purpose circuitry and / or the dedicated circuitry may be associated with or included in audio device 10, for example.

[0142] According to some embodiments, as shown in FIG. 8, a computer program 600 is stored on a non-transitory computer readable medium 510 to form a computer program product 500. The computer program product 500 comprises one or more non-transitory computer readable media 510 and one or more computer programs 600. The non-transitory computer readable medium may be any suitable non-transitory computer readable medium, exemplified by, but not limited to, a Universal Serial Bus (USB) memory, a Compact Disk (CD) ROM, a plug-in card, an embedded drive, or a Read Only Memory (ROM). FIG. 9a shows an exemplary computer readable medium in the form of a vintage 5,25 inch floppy disk 510. The computer readable medium 510 stores a computer program 600 including program instructions 610. The computer program 600 can be loaded into the controller 100 (see FIG. 9b), which may be included in, for example, an electronic device (see FIG. 9c) and / or an audio playback arrangement 10 (see FIG. 9d). When loaded into the controller 100, the computer program 600 may be stored in a memory 110 associated with or included in the controller 100. According to some embodiments, the computer program 600 when loaded into and executed by the controller 100 may cause the execution of method steps, for example according to the method 300 illustrated in Figure 7 or otherwise described herein.

[0143] Modifications and other variations of the described embodiments will occur to those skilled in the art having the benefit of the teachings presented in the foregoing description and the associated drawings. It should therefore be understood that the embodiments are not limited to the specific exemplary embodiments described in this disclosure, and that modifications and other variations are intended to be included within the scope of the present disclosure. Furthermore, although certain terms may be used herein, they are used in a generic and descriptive sense only and not for purposes of limitation. Thus, those skilled in the art will recognize numerous variations to the described embodiments that will still fall within the scope of the appended claims. Furthermore, although individual features may be included in different claims (or embodiments), these may in some cases be advantageously combined, and the inclusion of different claims (or embodiments) does not imply that a combination of features is not feasible and / or advantageous. Furthermore, a reference in the singular does not exclude a plural. Finally, reference signs in the claims are provided merely as a clarifying example and should not be construed as limiting the scope of the claims in any way. EXAMPLES

[0144] Below are numbered specific implementation examples, all of which are within the scope of this disclosure. Example 1 A method 300 for controlling an acoustic dose D of a listener 40, said method 300 comprising: obtaining 310 an audio profile 50 including sound pressure data 52 for said listener 40; The latest sound pressure p experienced by the listener 40 c and a step 320 of obtaining The sound pressure data 52 of the listener 40 and the latest sound pressure p c Based on the above, a latest estimated period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. ce and a step 330 of estimating The latest estimated period T ce is a dose period T associated with the acoustic dose threshold L. D In response to falling below The sound pressure data 52 and the latest sound pressure p c Based on the above, a last updated time period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. cu is the dose period T D Updated sound pressure p that meets or exceeds u A step 340 of calculating The updated sound pressure p cu The latest sound pressure p c and providing instructions 60 for controlling the The method 300 includes: Example 2 2. The method 300 of example 1, wherein the sound pressure data 52 of the audio profile 50 includes historical sound pressure data 52 including previous sound pressures experienced by the listener 40. Example 3 3. The method 300 according to any one of the preceding claims, wherein the sound pressure data 52 of the audio profile 50 comprises a cumulative sound pressure 52 experienced by the listener 40. Example 4 The method 300 according to any one of the first to third embodiments, wherein the sound pressure data 52 of the audio profile 50 includes a trend in sound pressure experienced by the listener 40. Example 5 The latest sound pressure p c is acquired 320 from a microphone 17 placed at the ears 42, 44 of the listener 40, and the method 300 according to any one of the first to fourth embodiments. Example 6 The audio profile 50 further includes a hearing profile 54 of the listener 40, and the latest estimated time T ce is calculated based on the hearing profile 54 of the listener 40. c The method 300 according to any one of Examples 1 to 5, further based on the hearing profile 54 of the listener 40 by frequency weighting 325 the Example 7 A current estimated time period T representing the estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L.ce The method 300 according to any one of the first to sixth embodiments, wherein the step of estimating 330 includes determining 335 a predictive model 56. Example 8 8. The method of claim 7, wherein the predictive model comprises a mean model. Example 9 The method 300 of any one of Examples 7 to 8, wherein the predictive model 56 comprises a recurrent neural network (RNN) model. Example 10 The latest sound pressure p c The method 300 according to any one of Examples 1 to 9, further comprising a step 360 of updating the sound pressure data 52 of the audio profile 50 for the listener 40 based on the sound pressure data 52 of the audio profile 50 for the listener 40. Example 11 The method 300 according to any one of the first to tenth embodiments, wherein the audio profile 60 is obtained from a remote storage 220. Example 12 The method 300 according to any one of Examples 1 to 11, wherein the method 300 is performed for the left ear 42 and / or the right ear 44 of the listener 40, and the audio profile 50 includes separate sound pressure data 52 for each of the left ear 42 and the right ear 44 of the listener 40. Example 13 A controller 100 for controlling an audio reproduction arrangement 10 comprising an audio transducer 15 and a current sound pressure p experienced by a listener 40 of the audio reproduction arrangement 10. c and a microphone 17 arranged to measure the signal strength of the sensor 100, the controller 100 comprising a control circuit, the control circuit comprising: obtaining 310 an audio profile 50 including sound pressure data 52 for said listener 40; The latest sound pressure p experienced by the listener 40 c and a step 320 of obtaining The sound pressure data 52 and the latest sound pressure p cBased on the above, a latest estimated period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. ce and a step 330 of estimating The latest estimated period T ce is a dose period T associated with the acoustic dose threshold L. D In response to falling below The sound pressure data 52 and the latest sound pressure p c Based on the above, a last updated time period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. cu is the dose period T D Updated sound pressure p that meets or exceeds u A step 340 of calculating The updated sound pressure p cu The latest sound pressure p c and providing instructions 60 for controlling the The controller 100 is configured to: Example 14 An audio playback arrangement 10 comprising a controller 100 as described in Example 13. Example 15 The current sound pressure p experienced by the audio transducer 15 and the listener 40 of the audio reproduction arrangement 10 c The audio reproduction configuration 10 of Example 14, further comprising a microphone 17 arranged to measure. Example 16 The controller 100 is obtaining an audio profile 50 including sound pressure data 52 for said listener 40; The latest sound pressure p experienced by the listener 40 c and obtaining The sound pressure data 52 and the latest sound pressure p c Based on the above, a latest estimated period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. ce and estimating The latest estimated period T ceis a dose period T associated with the acoustic dose threshold L. D In response to falling below The sound pressure data 52 and the latest sound pressure p c Based on the above, a last updated time period T representing an estimated time until the acoustic dose D of the listener 40 exceeds the acoustic dose threshold L is calculated. cu is the dose period T D Updated sound pressure p that meets or exceeds u A step 340 of calculating The updated sound pressure p cu The latest sound pressure p c and providing instructions 60 for controlling the The audio playback configuration 10 of Example 15 is configured to perform the following. (Example 17) The audio reproduction arrangement 10 according to Example 16, wherein the controller 100 is further configured to execute the method 300 according to any one of Examples 2 to 12. (Example 18) 18. The audio reproduction arrangement 10 of Example 16 or 17, wherein the controller 100 is further configured to control the audio transducer 10 to reproduce a source signal S received by the controller 100. (Example 19) The controller 100 is further configured to process the source signal S to adjust the intelligibility of the source signal S, and optionally, the audio reproduction configuration 10 further comprises an external microphone 17, and the intelligibility of the source signal S is adjusted based on background noise obtained from the external microphone 17. (Example 20) An audio reproduction configuration 10 as described in Examples 18 or 19, wherein the audio profile 50 further includes a hearing profile 54 describing a hearing impairment of the listener 40, and the controller 100 is further configured to process the source signal S based on the hearing profile 54. Example 21 The controller 100 calculates the updated sound pressure p u The latest sound pressure p c The audio reproduction arrangement 10 of any one of Examples 16 to 20, further configured to control the transducer 15 based on the instructions 60 for controlling the transducer 15. Example 22 A system 200 for controlling an acoustic dose D of a listener 40, said system 200 comprising: a cloud server 210 operatively connected to a remote storage 220 configured to store an audio profile 50 including sound pressure data 52 for said listener 40; The transducer 15, the controller 100, and the current sound pressure p experienced by the listener 40 c one or more audio reproduction arrangements 10 comprising a microphone 17 arranged to measure Equipped with The system 200, wherein the one or more audio playback configurations 10 are operatively connected to the cloud server 210, and the cloud server 210 and / or the one or more audio playback configurations 10 are equipped with a controller as described in Example 13. (Example 23) The remote storage 220 is configured to store a plurality of audio profiles 50, each of which includes sound pressure data 52 associated with an individual user 40 of the system 200; the one or more audio reproduction arrangements 10 are configured to determine an identity of the listener 40; The system 200 of Example 22, wherein the system 200 is further configured to associate the identity with an individual user 40 of the system 200. Example 24 A computer program product 500 comprising a non-transitory computer readable medium 510, the non-transitory computer readable medium 510 having a computer program thereon, the computer program 600 including program instructions 600, the computer program 600 being loadable into a controller 100, the computer program product being configured to cause the controller 100 to perform a method according to any one of the embodiments 1 to 12 when the computer program 600 is executed by the controller 100.

Claims

1. A system (200) for controlling an acoustic dose (D) of a listener (40), said system (200) comprising: a cloud server (210) operatively connected to a remote storage (220) configured to store a plurality of audio profiles (50), each of the plurality of audio profiles (50) including sound pressure data (52) and a hearing profile (54) associated with a particular user (40) of the system (200); A transducer (15), a controller (100), and the current sound pressure (p ｃ one or more audio playback arrangements (10) comprising a microphone (17) positioned to measure a frequency band (Fb) of the listener (40), the one or more audio playback arrangements (10) being operatively connected to the cloud server (210) and configured to determine an identity of the listener (40), the system (200) being configured to associate the identity with a particular user (40) of the system (200); Equipped with The cloud server (210) and / or the controller (100) of the one or more audio playback arrangements (10) may comprise a control circuit, the control circuit comprising: obtaining (310) an audio profile (50) comprising sound pressure data (52) and a hearing profile (54) for said listener (40); The latest sound pressure (p ｃ ) (320); The sound pressure data (52), the hearing profile (54) and the latest sound pressure (p ｃ ), a latest estimated time period (T) is calculated based on the acoustic dose (D) of the listener (40) and represents an estimated time until the acoustic dose threshold (L) is exceeded. ｃｅ ) based on the hearing profile (54) of the listener (40), ｃ ) by frequency weighting (325), The latest estimated period (T ｃｅ ) is the dose period (T Ｄ ) in response to The sound pressure data (52) and the latest sound pressure (p ｃ ), a last updated time period (T) representing an estimated time until the acoustic dose (D) of the listener (40) exceeds the acoustic dose threshold (L). ｃｕ ) is the dose period (T Ｄ ) that meets or exceeds the updated sound pressure (p ｕ ) (340); The updated sound pressure (p ｃｕ ) so as to satisfy the latest sound pressure (p ｃ providing (350) instructions (60) for controlling The system (200) is configured to:

2. 2. The system (200) of claim 1, wherein the sound pressure data (52) of the audio profile (50) comprises historical sound pressure data (52) comprising previous sound pressures experienced by the listener (40).

3. 3. The system (200) of claim 1 or 2, wherein the sound pressure data (52) of the audio profile (50) comprises a cumulative sound pressure (52) experienced by the listener (40).

4. The system (200) of any one of claims 1 to 3, wherein the sound pressure data (52) of the audio profile (50) comprises a trend in sound pressure experienced by the listener (40).

5. The latest sound pressure (p ｃ ) is acquired (320) from microphones (17) placed at the ears (42, 44) of the listener (40).

6. A latest estimated time period (T) representing the estimated time until the acoustic dose (D) of the listener (40) exceeds the acoustic dose threshold (L). ｃｅ The system (200) of any one of claims 1 to 5, wherein estimating (330) the metric (metric) comprises determining (335) a predictive model (56).

7. The system of claim 6 , wherein the predictive model comprises an average model.

8. The system (200) of claim 5 or 7, wherein the predictive model (56) comprises a recurrent neural network (RNN) model.

9. The control circuit determines the latest sound pressure (p ｃ 9. The system (200) of claim 1, further configured to: update (360) the sound pressure data (52) of the audio profile (50) for the listener (40) based on a

10. The control circuitry controls the latest sound pressure (p ｃ 10. The system (200) of claim 1, configured to cause a step (350) of providing instructions (60) for controlling a left ear (42) and a right ear (44) of the listener (40), the audio profile (50) comprising separate sound pressure data (52) for each of the left ear (42) and the right ear (44) of the listener (40).

Images