EP3900392A1

EP3900392A1 - Nearfield audio devices with resonant structures

Info

Publication number: EP3900392A1
Application number: EP18825978.2A
Authority: EP
Inventors: Genaro Wölfl
Original assignee: Harman Becker Automotive Systems GmbH
Current assignee: Harman Becker Automotive Systems GmbH
Priority date: 2018-12-18
Filing date: 2018-12-18
Publication date: 2021-10-27
Also published as: KR102562305B1; KR20210104657A; CN113170257A; WO2020125940A1

Abstract

A loudspeaker arrangement comprises at least one first loudspeaker (10, 102) arranged closer to a first position (P1) than to a second position (P2), wherein each of the at least one first loudspeakers (10, 102) is acoustically coupled to a plurality of first tubular structures (30, 301, 302, 303). Each of the first tubular structures (301, 302, 303) comprises an opening (40, 401, 402, 403) that is arranged closer to the second position (P2) than to the first position (P1), and each of the first tubular structures (301, 302, 303) is configured to receive sound from each of the at least one first loudspeakers (10, 102) and to emit sound through the respective opening (40, 401, 402, 403). Each of the first tubular structures (30, 301, 302, 303) has a length (1301, 1302, 1303) that is different from the lengths of each other first tubular structure.

Description

NEARFIELD AUDIO DEVICES WITH RESONANT STRUCTURES

TECHNICAL FIELD

[0001] The disclosure relates to nearfield audio devices with resonant structures, in particular to the mitigation of adverse effects of acoustic resonators and to the application of acoustically resonant structures in nearfield audio devices.

BACKGROUND

[0002] Different devices are known which are worn on a user’s body, e.g., on the user’s head, on the user’s shoulders, or anywhere on the user’s torso. Such devices may contain sound reproduction means or subassemblies for sound reproduction, wherein at least one loudspeaker is arranged close to at least one ear of the user when the device is worn by the user. Due to their audio features, such devices will be referred to as audio devices in the following, although, depending on the device type, these audio features may not be the main features of the device and the device may often therefore not be referred to as an audio device. As such audio devices may be designed to be worn on the body of the user, they may generally be referred to as wearables. However, other devices are known that are not worn on the user’s body but that are still constructed in a way such that at least one loudspeaker contained in such audio devices is positioned close to at least one ear of a user when the audio device is utilized by the user. Within the context of this document, audio devices that place at least one loudspeaker close to an ear of a user (e.g. closer than 0.5m) will be referred as nearfield audio devices. Especially if such nearfield audio devices leave the ears, including the entrance of the ear canal, at least partially open (uncovered), the generation of low frequency sound requires relatively high air volume displacement. While this is generally feasible with adequately sized and constructed loudspeakers, the overall size of audio devices including such loudspeakers may be quite large. If a small size of the audio device is required, different measures may be taken in order to increase sound pressure especially at the low frequency end of the sound frequency spectrum covered by an audio device.

Acoustically resonant structures are known that can increase the sound pressure initially generated by a sound source (e.g. a loudspeaker). For example, such structures may be integrated in the enclosures of loudspeaker systems known as bass reflex loudspeakers, transmission line loudspeakers or horn loudspeakers. Comparable resonant structures may be integrated in nearfield audio devices, placing at least one loudspeaker and at least one output of a resonant structure close to at least one ear of a user. From this close placement, specific beneficial but also disadvantageous characteristics of such near field audio devices with resonant structures may arise. In the following, solutions will be disclosed that allow taking advantage of the beneficial characteristics while mitigating adverse aspects.

SUMMARY

[0003] A loudspeaker arrangement includes at least one first loudspeaker arranged closer to a first position than to a second position, wherein each of the at least one first loudspeakers is acoustically coupled to a plurality of first tubular structures. Each of the first tubular structures includes an opening that is arranged closer to the second position than to the first position, and each of the first tubular structures is configured to receive sound from each of the at least one first loudspeaker and to emit sound through the respective opening. Each of the first tubular structures has a length that is different from the lengths of each other first tubular structure.

[0004] A method includes emitting sound with at least one first loudspeaker that is arranged closer to a first position than to a second position, wherein each of the at least one first loudspeakers is acoustically coupled to a plurality of first tubular structures. The method further includes receiving sound emitted by the at least one first loudspeaker in each of the first tubular structures, and emitting sound from each of the first tubular structures through an opening that is arranged closer to the second position than to the first position. Each of the first tubular structures has a length that is different from the lengths of every other first tubular structure.

[0005] A loudspeaker arrangement includes at least one first loudspeaker arranged closer to a first position than to a second position, wherein the at least one first loudspeaker is acoustically coupled to at least one first tubular structure, and wherein the at least one first tubular structure is configured to receive sound from the at least one first loudspeaker and to emit sound at an opening that is arranged closer to the second position than to the first position. The loudspeaker arrangement further includes at least one second loudspeaker arranged closer to the second position than to the first position, wherein the at least one second loudspeaker is configured to radiate sound that, at the second position, attenuates the level of at least sound emitted by the at least one first tubular structure within at least one first cancellation frequency range.

[0006] A method includes emitting sound with at least one first loudspeaker that is arranged closer to a first position than to a second position, wherein the at least one first loudspeaker is acoustically coupled to at least one first tubular structure. The method further includes receiving sound emitted by the at least one first loudspeaker in each of the at least one first tubular structures, and emitting sound from each of the at least one first tubular structures through an opening that is arranged closer to the second position than to the first position.

The method further includes emitting sound with at least one second loudspeaker that is arranged closer to the second position than to the first position, wherein the sound emitted by the at least one second loudspeaker is configured to, at the second position, attenuate the level of at least sound emitted by the at least one first tubular structure within at least one first cancellation frequency range.

[0007] Other systems, methods, features and advantages will be or will become apparent to one with skill in the art upon examination of the following detailed description and figures. It is intended that all such additional systems, methods, features and advantages be included within this description, be within the scope of the invention and be protected by the following claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The method may be better understood with reference to the following description and drawings. The components in the figures are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention. Moreover, in the figures, like referenced numerals designate corresponding parts throughout the different views.

[0009] Figure 1, including Figures 1 A to 1C, schematically illustrates examples of loudspeaker arrangements. [0010] Figure 2, including Figures 2 A to 2D, schematically illustrates examples of further loudspeaker arrangements.

[0011] Figure 3, including Figures 3A to 3G schematically illustrates examples of even further loudspeaker arrangements. [0012] Figure 4 schematically illustrates exemplary bode plots of transfer functions of components of a loudspeaker arrangement.

[0013] Figure 5, including Figures 5A to 5D, schematically illustrates exemplary loudspeaker arrangements.

[0014] Figure 6 schematically illustrates exemplary magnitude transfer functions of different versions of a component of a loudspeaker arrangement.

[0015] Figure 7 schematically illustrates a loudspeaker arrangement according to one example.

[0016] Figure 8 schematically illustrates a loudspeaker arrangement according to another example. [0017] Figure 9 schematically illustrates a loudspeaker arrangement according to another example.

[0018] Figure 10 schematically illustrates a loudspeaker arrangement according to another example.

[0019] Figure 11 schematically illustrates bode plots of transfer functions of components of an exemplary loudspeaker arrangement with different driving signals for individual loudspeakers.

[0020] Figure 12 schematically illustrates bode plots of transfer functions of components of an exemplary loudspeaker arrangement with different driving signals for individual loudspeakers and relative phase shift between different driving signals. [0021] Figure 13 schematically illustrates a signal flow diagram according to one example. [0022] Figure 14 schematically illustrates a signal flow diagram according to the same example as Figure 13.

[0023] Figure 15 schematically illustrates bode plots of a filter transfer function and a band pass filter amplitude. [0024] Figure 16 schematically illustrates exemplary magnitude transfer functions of components of loudspeaker arrangements with different loudspeaker driving signals.

[0025] Figure 17 schematically illustrates an average amplitude difference between direct and indirect sound for HRTF of 60° and 0° elevation.

[0026] Figure 18 schematically illustrates a signal flow diagram according to one example. [0027] Figure 19 schematically illustrates a signal flow diagram according to another example.

[0028] Figure 20 schematically illustrates a signal flow diagram according to another example.

[0029] Figure 21 schematically illustrates exemplary magnitude transfer functions of a loudspeaker arrangement at two positions for different loudspeaker driving signals.

[0030] Figure 22 schematically illustrates bode plots of a first filter for the generation of loudspeaker driving signals and a magnitude transfer function of a second filter utilized to determine the transfer function of the first filter.

[0031] Figure 23 schematically illustrates a signal flow diagram according to another example.

[0032] Figure 24 schematically illustrates a headphone arrangement worn by a user.

[0033] Figure 25 schematically illustrates a headphone arrangement.

[0034] Figure 26 schematically illustrates a loudspeaker arrangement according to one example. [0035] Figure 27 schematically illustrates an exemplary nearfield audio device that may be worn around the neck of a user.

[0036] Figure 28 schematically illustrates an exemplary nearfield audio device that may be worn around the neck of a user.

DETAILED DESCRIPTION

[0037] A partly enclosed air volume coupled to free (outside) air by an oscillating mass (of air) is often used for loudspeakers in enclosures that are known as ported, vented or bass reflex enclosures. According to one theory of operation of such enclosures, the oscillating mass is the mass of air inside a tubular structure having an arbitrary cross sectional shape that couples the internal enclosure volume to the outside. The internal enclosure volume provides an air spring that resonates with the air mass inside the port. Schematic examples of loudspeaker arrangements comprising bass reflex enclosure structures are shown in Figure 1. A loudspeaker 10 is arranged within an enclosure 20. The loudspeaker 10 comprises a membrane (not illustrated). The loudspeaker 10 is arranged in a wall of the enclosure 20 such that one side of the membrane faces the inside of the enclosure 20 and the other side of the membrane faces the outside. The enclosure 20 comprises a tube 30. The tube 30 may also be referred to as bass reflex port, sound guide, or waveguide. The bass reflex port 30 forms an opening of the enclosure 20. The bass reflex port 30 comprises a sound opening 40, where sound finally exits the bass reflex port 30 into free air. The bass reflex port 30 can be arranged inside the enclosure 20 (see, e.g., Figure 1 A), or may be arranged outside of the enclosure 20 (see, e.g., Figures IB, 1C). The bass reflex port 30 may have a comparably small length 11, as is exemplarily illustrated in Figure IB. It is, however, also possible that the bass reflex port 30 has a comparably large length 12 as compared to the dimensions of the enclosure 20, which would therefore hardly fit into the enclosure 20 (see, e.g., Figure 1C). Generally, the acoustic resonance of an enclosed air volume in an enclosure with an opening is known as Helmholtz-Resonance. If the port 30 protrudes into the enclosure 20, the air within the port 30 may be considered part of the air mass instead of the air spring. Therefore, a long port 30 cannot be integrated into the enclosure 20 without severe reduction of the air volume providing the air spring. [0038] A so-called passive radiator may alternatively provide the oscillating mass. Passive radiators typically comprise a membrane having a certain mass, which is mechanically coupled to an enclosure of an internal air volume by an elastic or springy mechanical structure. Resonant structures for loudspeakers with one or more enclosed air volumes and one or more oscillatory masses for coupling of enclosed air volumes to the outside or to other enclosed air volumes are known as band-pass enclosures of different types. As can be seen in the schematic illustrations of Figure 2, the loudspeakers 10 within band pass enclosures 20 do not radiate directly to free air. In case of 4th order band-pass enclosures (see, e.g., Figures 2A and 2B), the loudspeaker 10 is mounted in one closed chamber 20 and one vented chamber 21. That is, one side of the loudspeaker membrane faces a closed enclosure 20 and the other side of the loudspeaker membrane faces an enclosure 21 having at least one opening 40. In the example illustrated in Figure 2A, a bass reflex port 31 extends into the enclosure 21 and provides an opening 41 to the outside. In the example illustrated in Figure 2B, a bass reflex port 31 is provided that extends outside the enclosure 21 and provides an opening 41 to the outside. 6th order band-pass enclosures (see, e.g., Figures 2C and 2D), for example, provide two ventilated chambers 20, 21. A“ventilated chamber” in this context refers to an enclosure having at least one opening 40, 41. In the example illustrated in Figure 2C, one side of the loudspeaker membrane faces an enclosure 20 having a bass reflex port 30 extending inside the enclosure 20. The other side of the loudspeaker membrane faces a second enclosure 21 having a bass reflex port 31 extending inside the enclosure 21. In the example illustrated in Figure 2D, the bass reflex ports 30, 31 extend outside the respective enclosures 20, 21, wherein the lengths 130, 131 of the bass reflex ports do not necessarily have to be identical.

As is illustrated in Figure 2D, one bass reflex port 30 may have a length 130 that is shorter than a length 131 of the other bass reflex port 31. This, however, is only an example. The bass reflex ports 30, 31 may also have identical lengths (130 = 131). Similar to bass reflex enclosures, the ported chambers of band-pass enclosures comprise respective Helmholtz resonances mainly defined by the enclosure volumes and port dimensions.

[0039] Acoustic resonances may also develop in tubular or pipe structures. If a tubular structure or pipe with an arbitrary cross sectional shape that is closed at one end is excited acoustically, it resonates at a fundamental resonance frequency with a wave length of four times the length of the tubular structure (quarter wavelength resonance) and at odd integer multiples of the fundamental resonance frequency (odd order harmonics). Loudspeaker enclosures that utilize especially the fundamental resonance in order to amplify the sound pressure level generated by a loudspeaker are widely known as transmission line or quarter wave loudspeaker enclosures. Although an ideal transmission line would not comprise any resonance, practical transmission lines comprise multiple resonances and the lowest resonance is typically utilized for boost of radiated sound energy. Therefore, loudspeaker enclosures or systems relying on the quarter wave length resonance of tubular structures will be referred to as transmission line enclosures or loudspeakers in the following.

[0040] Different exemplary variations of transmission line enclosures are illustrated in Figure 3. A transmission line enclosure may, for example, be implemented as a tube or pipe with a constant cross sectional area over the entire length 120, with one open end (opening) 40 and one closed end, and with a loudspeaker 10 mounted in a wall of the enclosure 20, e.g., at or close to the closed end (see, e.g., Figure 3A). In the example illustrated in Figure 3A, the enclosure 20 comprises a top wall, a bottom wall and a front wall. The top wall and the bottom wall are arranged in parallel to each other, with the front wall arranged perpendicular to the top and bottom walls and closing one end of the enclosure 20. The enclosure 20 also comprises sidewalls which, however, are not specifically illustrated in Figure 3A. The enclosure 20 is open at its second end, that is, the enclosure 20 does not comprise a rear wall. Tapered variants may narrow the cross sectional area towards the open end 40 (see, e.g., Figure 3B) or towards the closed end of the enclosure 20 (the pipe) (see, e.g., Figure 3C). In the example illustrated in Figure 3B, the top and the bottom walls are not arranged in parallel to each other. A distance between the top wall and the bottom wall at the closed end is greater than a distance between the top wall and the bottom wall at the open end 40 of the enclosure 20. In the example illustrated in Figure 3C, the enclosure 20 comprises a top wall and a bottom wall, but does not comprise a front wall. The top and the bottom walls are not arranged in parallel to each other. A distance between the top wall and the bottom wall at the closed end is smaller than a distance between the top wall and the bottom wall at the open end 40 of the enclosure 20. In particular, the top wall and the bottom wall may be directly connected to each other at the closed end of the enclosure 20. The example illustrated in Figure 3D is similar to the example of Figure 3 A, with the loudspeaker 10 arranged in the top wall instead of the front wall. [0041] At least one loudspeaker 10 may be mounted at or near the closed end of the pipe (enclosure) 20 or somewhere along the longitudinal axis of the pipe (see, e.g., Figures 3C and 3D). In the examples illustrated in Figures 3 A and 3B, one loudspeaker 10 is arranged in the front wall of the enclosure 20. In the examples illustrated in Figures 3C and 3D, for example, one loudspeaker 10 is arranged in the top wall of the enclosure 20. These, however, are only examples. At least one loudspeaker 10 may be mounted in a top wall, a bottom wall, a front wall or a sidewall of the enclosure 20.

[0042] Vented pipe geometries provide a constricted output of the transmission line (see, e.g., Figure 3E), thereby also introducing a Helmholtz resonance that is potentially lower than the fundamental quarter wave length resonance of the pipe. In the example of Figure 3E, the enclosure 20 comprises a rear wall with an opening and a bass reflex port 30 that is coupled to the opening of the rear wall of the enclosure 20. Sound may leave the enclosure 20 through an opening 40 of the bass reflex port 30. The same is the case for relatively small coupling chambers in front of the transmission line pipe as illustrated in Figure 3F. While in the example of Figure 3E a length 120 of the enclosure 20 is large as compared to a length 130 of the bass reflex port 30, the length 120 of the enclosure 20 is comparably small as compared to the length 130 of the bass reflex port 30 in the example illustrated in Figure 3F. Figure 3G illustrates a transmission line (enclosure 20) with coupling chamber (bass reflex port 30) and an intermediate chamber (second enclosure 22) that may damp higher order pipe resonances. Figures 1 to 3 very generally illustrate different enclosure principles in a very simplified way. Cross sectional areas of enclosures as well as ports and pipes generally may have any arbitrary shape. For example, enclosures may not comprise specific wall sections but may have at least partly continuous wall shapes (e.g. cylindrical, spherical).

[0043] When comparing the exemplary transmission line enclosures of Figures 3E and 3F with the bass reflex enclosure of Figure 1C, it seems that all of these examples may be described as either transmission line or bass reflex enclosures. The boundaries between transmission line and bass reflex enclosures are generally fluent. As the lowest resonance is typically utilized for low frequency enhancement, one possible differentiation would be the nature of this resonance. A Helmholtz resonance may be seen as characterizing a bass reflex enclosure and a quarter wavelength resonance may be seen as characterizing a transmission line enclosure. Often, at least a quarter wave length resonance may be excluded based on port or pipe lengths or, generally, the largest internal dimension within an enclosure. However, geometries may be chosen such that these resonances overlap. Additionally, both resonance types are affected by the combination of relatively long ports or pipes with coupling volumes. For both resonance types typical resonance frequencies (quarter wavelength or Helmholtz, respectively) may be shifted considerably by either the coupling volume behind the loudspeaker or the long port length, respectively. As a result, standard formulas for resonance frequency calculation may not provide accurate results. Actual resonance frequencies for rear volumes behind a loudspeaker combined with relatively long ports or pipes are typically lower than predicted by standard Helmholtz resonance formulas or the quarter wavelength resonance of the pipe. This may be beneficial for near field audio devices with limited space for long ports or large enclosure volumes. It may, however, complicate the classification of the enclosure as transmission line or bass reflex enclosure.

[0044] All tubular resonances of bass reflex ports as well as the tubular resonances above the fundamental resonance frequency of transmission lines are typically disadvantageous in ported and transmission line loudspeakers as they interfere with the frontal sound emitted by the loudspeaker in the respective enclosure. Furthermore, these resonances may have long decay times, which can degrade sound quality. Finally, these resonances may be excited by subharmonic frequencies and, therefore, may cause harmonic distortion. Hence, multiple techniques are known that reduce or damp these resonances. In the case of transmission line loudspeakers, the fundamental transmission line resonance frequency is typically not damped or not damped as much as the higher resonance frequencies, as it is utilized for sound pressure amplification. For bass reflex loudspeakers, the Helmholtz resonance is typically not damped, while the port resonances on the other hand are damped. Most damping techniques affect all of those resonances while others are limited to certain frequency regions. Among the techniques that damp all resonances, albeit to different degrees, are, for example, the application of damping material in the enclosure and/or port or pipe, cross sectional area variations along the longitudinal axis of the pipe or port as well as bending or folding of the pipe or port along the longitudinal axis. In transmission lines, the positioning of one or more loudspeakers at certain relative positions along the longitudinal axis of the pipe may avoid or damp resonances in certain frequency regions depending on the loudspeaker position. In order to damp specific resonance frequencies, for example, stub pipes, Helmholtz absorbers or other resonant structures may be applied for example at pressure nodes within the port or pipe.

[0045] As an example, Figure 4 illustrates the output (Tube, dotted line) from a 19,5cm long tube (bass reflex port 30) with 4mm diameter coupled to a small volume (enclosure 20) of 4,2cm³ at the rear side of a loudspeaker 10 (see, e.g., Figure 3F), as well as the direct output (SPK, dashed line) from the loudspeaker 10 and the acoustic sum (Sum, continuous line) of both. The local minimum in the loudspeaker output (SPK) at 186Hz indicates that the combination of tube 30 and rear volume 20 acts as a Helmholtz resonator tuned to that frequency. With the given chamber volume and port parameters, the Helmholtz resonance frequency can approximately be calculated using the following equation: f_H = c/(2 * pQ * SQRT( _T) (2.1).

Equation 2.1 results in about 212Hz for the given example with S being the cross sectional area of the tube, L’ being the length 130 of the tube with end correction (L’ = L + 1,7 * r), V being the volume in the speaker chamber 20 and r being the diameter of the tube 30. The measured Helmholtz resonance frequency is about 14% lower than the calculated frequency, which suggests that the end correction does not work properly or is generally not sufficient for resonance frequency calculation with the relatively long bass reflex port. In addition multiple port resonances can be seen in the tube output at approximately f = n * c/ (2 * L) (2.2), where n is a positive integer, c is the sound velocity and L is the length (130) of the tube 30. This means that the tube 30 approximately acts as a cylindrical (tubular) resonator, which is open at both ends. The loudspeaker output comprises local minima at the Helmholtz resonance frequency as well as the fundamental tube resonance frequency (l/2) but merely shows negligible effects of higher order tube resonances. However, the acoustic sum of the tube and loudspeaker output is severely modulated by the first five port resonances, which is a clear disadvantage as compared to closed box loudspeakers and should be remedied for good audio performance.

[0046] Due to phase inversion between the tube output 40 and the direct loudspeaker output, the acoustic sum drops below the tube output level at about 165Hz, rapidly declining with decreasing frequency. This steep low frequency ro 11-off of the magnitude response is another disadvantage of bass reflex loudspeakers as compared to closed box loudspeakers. As will be described below, the steep roll-off can be mitigated due to specific characteristics of close loudspeaker and resonator output positions to the ears. Around the Helmholtz resonance frequency the port output is much higher than the loudspeaker output, which means that the loudspeaker excursion is much lower for the ported loudspeaker and for any given sound pressure level than it would be for a closed box loudspeaker. Therefore, it is desirable to keep the bass level boost around the Helmholtz resonance frequency while at least mitigating the disadvantageous effects of port resonances and steep low frequency amplitude roll-off.

[0047] Port resonances as well as general high frequency leakage through a port may be avoided by the utilization of a passive radiator instead of the port. Passive radiators, however, also have several drawbacks. In order to provide more sound output than the active loudspeaker, the air volume displacement needs to be higher, therefore requiring a larger membrane and/or higher membrane excursion. The resulting dimensions of a passive radiator may therefore be prohibitive, especially for wearable devices. Furthermore, passive radiators often require a relatively high membrane weight in order to achieve the desired resonance frequency and excursion. The impulses of the moving mass may cause a generally

lightweight wearable device to vibrate noticeably for the user. This vibration may be disturbing for the user and may reveal the sound source position, which may be undesirable if virtual sound sources are to be synthesized by appropriate signal processing methods to be perceived by the user as distant to the audio device.

[0048] Transmission line loudspeakers generally suffer from comparable problems as bass reflex ports concerning undesired higher order pipe resonances. Multiple methods for damping such pipe or port resonances are known. Generally, such damping techniques are required to damp the undesired resonances while leaving the desired resonances (e.g.

Helmholtz resonance or quarter wavelength resonance, whichever is lower) mostly unaffected. Most damping techniques affect all of those resonances, while some are limited to certain narrow band frequency regions. The passive damping methods proposed below are generally very effective in damping port or pipe resonances while leaving the Helmholtz resonance virtually unaffected. Therefore, they may be used for any resonator

implementation in which the lowest resonance is a Helmholtz resonance. As quarter wavelength resonances below, e.g. 100Hz, require considerable pipe lengths, the Helmholtz resonance, which can be tuned to lower frequencies with shorter port length, will be the lowest in most practical cases (e.g., lower than the lowest quarter wavelength resonance of the port, pipe or enclosure). Hence, the new damping methods are applicable for the majority of useable resonant structures in wearable nearfield audio devices.

[0049] The basic concept of the proposed damping techniques is illustrated in Figure 5, which, in Figures 5A to 5D, illustrates four exemplary implementation variants. Instead of a single port or pipe 30 (see, e.g., Figures 1 and 2), a multitude of ports or pipes 301, 302, 303 with different lengths is provided. While the examples of Figures 5 A to 5D all illustrate three ports or pipes 301, 302, 303 per enclosure 20, any number m of ports or pipes may be applied, with m > 2. That is, each enclosure 20 has at least one opening, with one or more bass reflex ports or pipes 301, 302, 303 coupled to each of the openings. Sound, therefore, may leave the enclosure 20 through two or more different ports or pipes 301, 302, 303. Each of the ports or pipes 301, 302, 303 has a length 1301, 1302, 1303, that is different from the length of each of the remaining ports or pipes 301, 302, 303. That is, 1301 ¹ 1302 ¹ 1303. The difference in length 130m may be achieved in different ways. Various mechanical alignments of the multitude of ports or pipes 301, 302, 303 are exemplarily illustrated in Figure 5. Port or pipe inputs as well as port or pipe outputs 401, 402, 403 may be aligned within respective planes and/or positioned in close proximity to each other. They may join at one common input or output or may comprise separate inputs or outputs. Relative placement of port or pipe inputs and port or pipe outputs 401, 402, 403 may affect damping or port or pipe resonances, especially towards higher frequencies. Appropriate placement for the desired damping result in a given loudspeaker system may, for example, be chosen empirically or may be based on acoustic simulation.

[0050] As is exemplarily illustrated in Figure 5 A, different ports or pipes 301, 302, 303 may comprise different curvatures or slopes along their respective longitudinal axis. If the inputs of the ports or pipes 301, 302, 303 are positioned close to each other, e.g., adjacent to each other, and the outputs 401, 402, 403 of the ports or pipes 301, 302, 303 are positioned close to each other, e.g., adjacent to each other, the different curvatures or slopes of the ports or pipes 301, 302, 303 may result in different lengths 1301, 1302, 1303 of the ports or pipes 301, 302, 303. In the example illustrated in Figure 5A, one of the ports or pipes 302 is illustrated as a straight tube. This is for illustration purposes only. Generally, all of the ports or pipes 301, 302, 303 may be curved tubes. The ports or pipes 301, 302, 303 may be integrated in a loudspeaker device such as a headphone arrangement or a loudspeaker device that may be worn anywhere on the upper body of a user, for example, and the curvature of the different ports or pipes 301, 302, 303 may, at least partly, depend on a geometry of the respective loudspeaker device. It is, for example, also possible that the different ports or pipes extend spirally around each other with different curvatures or slopes.

[0051] In the example illustrated in Figure 5B, the outputs 401, 402, 403 of the ports or pipes 301, 302, 303 are arranged adjacent to each other. All ports 301, 302, 303 extend parallel to each other along an axis in a horizontal direction. The ports or pipes 301, 302, 303 being implemented as straight tubes again is only an example. The ports or pipes 301, 302, 303 may alternatively be implemented as curved or angled tubes extending parallel to each other with identical curvatures or slopes. The inputs of the ports or pipes 301, 302, 303 within the enclosure 20, however, are not arranged directly adjacent to each other. The ports or pipes 301, 302, 303 extend into the enclosure 20 to different degrees. This results in different lengths of the ports 301, 302, 303.

[0052] In the example illustrated in Figure 5C, the inputs of the ports or pipes 301, 302, 303 are arranged adjacent to each other. All ports or pipes 301, 302, 303 extend in parallel to each other along an axis in a horizontal direction. The ports or pipes 301, 302, 303 being implemented as straight tubes again is only an example. The ports or pipes 301, 302, 303 may alternatively be implemented as curved or angled tubes extending parallel to each other with identical curvatures or slopes. The outputs 401, 402, 403 of the ports or pipes 301, 302, 303 outside the enclosure 20, however, are not arranged directly adjacent to each other. The ports or pipes 301, 302, 303 extend outside the enclosure 20 to different degrees. This results in different lengths of the ports or pipes 301, 302, 303. [0053] In the example illustrated in Figure 5D, both the inputs as well as the outputs 401,

402, 403 of the ports or pipes 301, 302, 303 are not arranged directly adjacent to each other, resulting in different lengths 1301, 1302, 1303 of the ports or pipes 301, 302, 303. The ports or pipes 301, 302, 303 being implemented as straight tubes again is only an example. The ports or pipes 301, 302, 303 may alternatively be implemented as curved or angled tubes extending parallel to each other with identical curvatures or slopes. Obviously, ports or pipes with different curvatures or slopes may additionally comprise spatially distributed inputs and/or outputs and, thereby, different lengths along their longitudinal axis

[0054] While there are no general limitations to the combination of cross sectional areas and lengths for the multitude of ports or pipes 301, 302, 303, it may be desirable to keep the Helmholtz resonance provided by a single port in combination with a certain rear enclosure volume for a multi-port setup. For example, this enables Helmholtz frequency tuning with a single port and subsequent port resonance damping independent from the Helmholtz frequency. As previously described, certain types of transmission line enclosures as well as bass reflex enclosures may both feature an air volume behind one or more loudspeaker(s), which is coupled to free air by a pipe or port, respectively. This may also be the case for a multi-port or multi-pipe setup comprising a common volume at which multiple ports or pipes join. In the latter case, the lowest resonance of the enclosure may occur at a frequency lower than that which would occur at the lowest possible quarter wavelength frequency based on the dimensions of the enclosure including the port(s) or pipe(s). On the other hand, the lowest resonance frequency of an enclosure structure with a rear volume behind at least one loudspeaker and one or more relatively long port(s) may even be lower than would be the case for a pure Helmholtz resonance, where the air in the port is considered a single oscillating mass. Therefore, none of these specific resonance types may apply perfectly to the actual resonance within an enclosure structure. Nevertheless, the term Helmholtz resonance will be applied in the following for a common resonance of the air volume behind one or more loudspeakers and within all pipes or ports connected thereto. This resonance may be the lowest acoustic resonance of the enclosure structure.

[0055] In order to maintain an approximately equal Helmholtz resonance frequency as obtained with a single port or pipe 30, the (average) port cross section area of each of the multitude of ports or pipes 301, 302, 303 may be equal and the sum thereof may equal the (average) cross section area of the single port or pipe 30. In addition, the average length of the multitude of ports or pipes 301, 302, 303 may equal the length 130 of the single port or pipe 30. As an example, Figure 6 illustrates port resonance damping results with one (D7 L33), three (D4 L28/33/38), and six ports (D3 L28/30/32/34/36/38) of respectively equal cross sectional areas and different lengths 130m. Figure 6 shows measured magnitude responses of the output of the aforementioned port combinations. The sum of the respective cross section areas per port combination as well as the average length of all combined ports per port combination where approximately equal. As can be seen, the magnitude response around the lowest resonance, which in this case was a Helmholtz resonance at about 90Hz, is not affected. Fundamental port resonance at about 485Hz, as well as integer multiples that are present in the single port 30, are damped by both multi-port setups up to the fifth harmonic at about 2425Hz.

[0056] Generally, the range over which the port length 130m varies affects the frequency range around each port resonance that is damped as compared to a single port 30. The number m of ports and the distribution of their lengths 130m over the total length variation range may affect the magnitude ripple in the damped frequency regions. For the example illustrated in Figure 6, the port lengths 130m have been spaced linearly across the total length variation range, which was equal for both multi-port setups. Consequently, the variant with six ports (m = 6) shows lower ripple in the damping regions around the port resonances of the single port 30 (m =1) than the example with three ports (m = 3). For a slightly lower ripple within the damped frequency range, logarithmic port length spacing may be chosen.

Individual port lengths may, for example, vary between ±5% and ±35% of the average port length.

[0057] As the length 130m of transmission line pipes, as well as bass reflex ports 30m, affect the resonance frequency of the respective resonant structure, these lengths 130m may be chosen adequately for a desired lowest resonance frequency. Generally, port or pipe lengths 130m become longer the lower the lowest resonance frequency becomes. In case of bass reflex enclosures, as well as various transmission line enclosures, for example, transmission line enclosures with coupling chambers as illustrated in Figure 3F, the cross sectional area of the port or pipe 30m also affects the lowest resonance frequency if the latter is constituted by a Helmholtz resonance. In these cases, the lowest resonance becomes lower in frequency the smaller the cross sectional area of the bass reflex port or pipe becomes. As a certain minimum cross sectional area will be required for any of those cases in order to avoid excessive distortion and air noise, the required port or transmission line length 130m may be relatively large for a desired resonance frequency.

[0058] Pipes of transmission line loudspeaker enclosures, as well as ports in ported enclosures, are often bent or folded along their longitudinal axis in order to fit into a desired enclosure shape. In both cases, the cross sectional shape and area may be either constant or may vary along the longitudinal axis. In order to integrate a loudspeaker enclosure with a long pipe or port into the shape of an audio device, the port or pipe may be bent or folded within that device. In case of a wearable audio device, which may, for example, be designed to be worn around the neck or on the head of a user, a natural choice would be to bend or fold the port or pipe such that it runs around parts of the user’s neck or head. This can be implemented such that any loudspeaker 10 driving the resonant structure, as well as the port or pipe output 40m, are located close to the same ear. As a result, a typical bass reflex or transmission line system is obtained that mainly supplies sound to a single ear. This is exemplarily illustrated in Figure 7, which shows a folded tubular structure 30 that couples a back volume within enclosure 20 behind a loudspeaker 10 to free air. The opening 40 of the tube 30 towards free air is located close to or adjacent to the loudspeaker 10. Both the loudspeaker 10 and the tube opening 40 towards free air mainly supply sound to a first position PI, which is closer to the loudspeaker 10 and the opening 40 than a second position P2. Although the dimensions in Figure 7 are not to scale and although Figure 7 merely principally illustrates a 3-dimensional arrangement by means of a 2-dimensional drawing, the distances dl 1 and dl2 symbolically illustrate this. The distance dl 1 between the loudspeaker 10 and the first position PI or between the output 40 of the tube 30 and the first position PI is significantly shorter than the distance dl2 between the loudspeaker 10 or the output 40 of the tube 30 and the second position P2. The distance between the output 40 of the tube 30 and the first position PI and the distance between the loudspeaker 10 and the first position PI do not necessarily need to be equal. However, both distances may be considerably shorter than the respective distance to the second position P2. The first and the second position PI, P2 may be typical positions of the user’s ears when the user is wearing an audio device incorporating the enclosure structure illustrated in Figure 7. [0059] According to another example, the at least one loudspeaker 10 may be arranged close to a first ear of the user (first position PI) and the port or pipe output 40 may be arranged close to the second ear of the user (second position P2) as is illustrated in Figure 8. That is, the opening 40 of the port 30 is arranged distant from the loudspeaker 10. A distance dl 1 between the loudspeaker 10 and the first position PI may be significantly shorter than a distance dl2 between the loudspeaker 10 and the second position P2. A distance d22 between the output 40 of the port 30 and the second position P2 may be significantly shorter than a distance d21 between the output 40 of the port 30 and the first position PI . This may provide benefits regarding mechanical integration of a relatively long tubular structure, especially into a wearable audio device, because folding of the tube 30 may not be required. Additionally, the steep low frequency magnitude roll-off of typical bass-reflex and transmission line enclosures may be avoided as the out of phase signals of the port and the loudspeakers do not cancel acoustically at the respective ears. Instead, these signals are perceived separately by the respective ears and are combined in the user’s auditory system. Left and right ear loudness levels may add up favorably to a higher binaural loudness level as received by the user. However, phase inversion between left and right ear may negatively affect perceived sound quality for the user.

[0060] In addition to a ported loudspeaker with distant speaker 10 and port output 40 as illustrated in Figure 8, a second sound source 14, e.g., a closed box loudspeaker, may be added close to the port output 40 as is exemplarily illustrated in Figure 9. The embodiment of Figure 9 is essentially similar to the embodiment illustrated in Figure 8. However, the arrangement of Figure 9 comprises a second enclosure 24 with the second loudspeaker 14 mounted therein. The second enclosure 24 does not comprise any openings. The second loudspeaker 14 radiates sound from a position adjacent to the opening 40 of the port 30. The opening 40 of the port 30 and the second loudspeaker 14 are arranged closer to the second position P2 than to the first position PI . Because the phase of the tube output 40 and the additional second loudspeaker 14 may be controlled individually, the loudspeaker signal and the port output may be controlled to sum up acoustically well below the lowest resonance of the ported loudspeaker 10. Furthermore, this setup provides potentially better control of port or pipe resonances as the second loudspeaker 14 may be controlled to provide a

compensation or cancellation signal for those resonances. The latter is not possible for the ported loudspeaker 10 alone because any equalizing that suppresses the resonance peaks at the port output 40 will result in undesirable notches in the loudspeaker output. This is not the case with the separate second loudspeaker 14. Furthermore, the arrangement of Figure 9 also allows playback of stereo signals at least over the frequency range covered by the second loudspeaker 14. There is, however, the drawback of port leakage from the ported loudspeaker 10, which will increase crosstalk from this loudspeaker 10 to the other ear P2 and, thereby, reduce stereo separation. Phase inversion between left and right ear PI, P2 may also negatively affect perceived sound quality for the user.

[0061] Finally, the port of the first ported enclosure 20 may end at the loudspeaker 14 of a second ported enclosure 24 and vice versa, as is schematically illustrated in Figure 10. The loudspeaker 10 of the first enclosure 20 and the port 44 of the second enclosure 24 may be located close to a first position PI and the loudspeaker 14 of the second enclosure 24 and the port 40 of the first enclosure 20 may be located close to a second position P2. As in the examples of Figure 7 to 9, positions PI and P2 may be typical positions of a user’s ears during use of a nearfield audio device including the loudspeaker system of Figure 10. With the loudspeaker arrangement of Figure 10, the relative phase of the loudspeaker signals and the respective nearby port output signal may be controlled for better summation below the lowest enclosure resonance. Contrary, however, to the arrangement of Figure 9, the arrangement of Figure 10 comprises cross-coupling of loudspeakers 10, 14 and ports 30, 34, which means that any phase change on a loudspeaker signal includes a phase change of the port output 40, 44 of the respective enclosure 20, 24. Therefore, a phase change on a single loudspeaker affects signal summation of respective loudspeaker and port output signals at both locations (PI and P2). This may impose restrictions on relative phase manipulation between loudspeakers.

[0062] In Figures 7 to 10, only one port or pipe 30, 34 is illustrated for each of the enclosures 20, 24. This, however, is only an example. It is also possible that each enclosure 20, 24 is implemented with two or more ports or pipes 301, 302, 303, 341, 342, 343, as has been described with respect to Figure 5 above.

[0063] One benefit of the cross-coupling resonant enclosure arrangement of Figure 10 is the general availability of equal dynamic headroom or maximum sound pressure levels at both positons (PI and P2), provided that both loudspeakers and enclosures are essentially acoustically identical. In principle, this allows the generation of equal sound signal levels at both positons PI, P2, although relative phase changes between the loudspeakers 10, 14 of both enclosures 20, 24 may result in an unequal acoustic summation of the loudspeaker and port output signal on both sides. Furthermore, similar to the arrangement of Figure 9, the arrangement of Figure 10 also provides the potential for stereo playback and a potentially better control of port or pipe resonances as the second loudspeaker 14 may be controlled to provide a compensation or cancellation signal for those resonances. However, port resonances, port leakage and phase inversion may negatively affect perceived sound quality for the user if they are not remedied or at least mitigated by appropriate measures.

[0064] For bass reflex or transmission line loudspeakers that are arranged in the far field of a user, the distance between the port or pipe output 40 and the loudspeaker membrane is typically much smaller than the distance of the port or pipe output to the ears of the user or the distance between the loudspeaker membrane and the ears of the user. As a result, the user may experience the acoustic sum of the direct loudspeaker signal and the port output signal at both ears. Such an acoustic sum signal having the aforementioned characteristics is exemplarily illustrated in Figure 4. However, if a sound device places loudspeakers 10, 14 and port or pipe outputs 40, 44 close to the ears of a user, the port or pipe output 40, 44 may be closer to one ear and the corresponding loudspeaker(s) 10, 14 may be closer to the other ear of the user. Hence, the distance between the port or pipe output 40, 44 and the

corresponding loudspeaker(s) 10, 14 may be considerably larger than the distance between the port or pipe output 40, 44 and the respective ear or the distance between the loudspeakers 10, 14 and the respective ear. This means that the user may mainly experiences the direct loudspeaker output from a loudspeaker 10, 14 in an enclosure 20, 24 at one ear while receiving essentially the output of a port or pipe 30, 34 connected to the respective enclosure 20, 24 at the other ear. Both signals are exemplarily illustrated in Figure 4for a loudspeaker arrangement with a single port or pipe per enclosure. In this case, the steep low frequency roll-off below the Helmholtz resonance frequency region of typical bass reflex loudspeakers or below the lowest resonance frequency region of transmission line loudspeakers (e.g. l/4 resonance), as can be seen in the acoustic sum (Sum) illustrated in Figure 4, may not occur at the ears of the user. Instead, there is essentially a 180° phase shift between the direct loudspeaker sound and the tube output sound received at the ears of the user in this frequency range (below about 100Hz in Figure 4). Furthermore, sound level differences between the port or pipe output and the direct loudspeaker output may exist especially close to the lowest enclosure resonance (e.g. Helmholtz or l/4).

[0065] The phase inversion affects sound perception by the user who is located in the near field of the device as well as sound pressure levels in the far field of the device. In the far field of the device, the inverse phase signals from port or pipe output 40, 44 and loudspeaker 10, 14 cancel each other if sound pressure levels are equal, thereby reducing sound levels externally to the device within the frequency range below the lowest resonator resonance. Although this could be considered as an advantage because the lower sound level in the far field of the device may reduce disturbance to other persons in the proximity of the user, it should be noted that this phase inversion is also possible for closed box loudspeaker systems, which position individual loudspeakers close to both respective ears of the user and is therefore of no specific advantage in vented loudspeaker systems. Furthermore, as is exemplarily illustrated in Figure 4, the relative phase between the port or pipe output 40, 44 and the direct loudspeaker output may vary over frequency, which limits the frequency range for which the phase is actually inversed. In contrast, closed box loudspeakers allow relative phase inversion over the complete frequency spectrum of the audio device with minimum signal processing effort. In that regard, vented enclosure types may therefore be seen as disadvantageous concerning their potential for far field sound level reduction.

[0066] For the user of an audio device presenting inverse phase signals to the respective ears of the user, the perceived sound image in the concerned frequency range will be affected such that it is at least partly externalized (perceived as being outside of the user’s head). Sound may be perceived as originating from two respective sound sources, each one close to one ear. In addition, a sensation of vibration may be felt on the auricles that is caused by the inverse phase of bass notes and that disappears when the ear signals are essentially in phase. Furthermore, bass notes may be perceived as resonant or lacking a clear definition in time. These effects may be disliked by users, especially by skilled listeners, but may actually remain unnoticed by other, e.g., unexperienced, users. Depending on the type of audio device, the relative distances between port or pipe output 40, 44, as well as corresponding

loudspeakers 10, 14 from the respective ears of the user, may either be essentially fixed or stable (e.g. head mounted device) or may vary with head movements (e.g. shoulder mounted device). If these distances vary with head movements of the user, the sound pressure level at the ears may vary considerably, depending on the position of the user’s head relative to the port or pipe outputs 40, 44 and to the corresponding loudspeakers 10, 14 (due to head movements), such that the low frequency end of the transmission range of the audio device may vanish completely at certain ear positions (position of user’s ears with regard to port or pipe outputs 40, 44 and to corresponding loudspeakers 10, 14). Especially if sound is radiated out of phase on two respective sides of the head (e.g. on the shoulders), these out of phase signals will cancel each other at positions between the radiating loudspeakers and ports or pipes on the respective side of the head (e.g. in front, behind and above the head with the user facing in a forward direction perpendicular to an axis between the shoulders).

[0067] These negative impacts of relative phase inversion between the ear positons or the typical ear positions PI, P2 can be mitigated or completely compensated with cross-coupling loudspeaker systems like the arrangement illustrated in Figure 10. If both loudspeaker systems 10, 40 / 14, 44 are essentially identical and driven with an identical signal, as well as positioned symmetrically with respect to positions PI and P2 (ear positions or typical ear positions), the resulting signals at both positions PI and P2 resemble an acoustic summation signal of loudspeaker 10, 14 and port output 40, 44 from a single enclosure 20, 24 (e.g., as in position PI of Figure 7, or as exemplarily illustrated in Figure 4 as sum signal) and are thus in phase. This, however, will also bring back the steep low frequency ro 11-off below the lowest resonance of the loudspeaker enclosures. Basically, there is a tradeoff between negative effects of relative phase shift between left and right ear of the user or at least the typical ear positions PI and P2, and the low frequency extension of the audio device.

[0068] Figure 11 is based on measured impulse responses of an exemplary single

loudspeaker system with resonant structure and remote port output as illustrated in Figure 8. Measurements were taken at the ear positions PI, P2 of a dummy head with a loudspeaker 10 positioned close to the left ear PI and the corresponding port output 40 close to the right ear P2. In order to simulate a completely symmetrical cross-coupled loudspeaker system as illustrated in Figure 10 with symmetrical placement regarding positons PI and P2, a time domain summation of the measured and, optionally, phase shifted or inverted signals was carried out to receive simulated sum signals for both positions PI, P2. Figure 11, therefore, shows magnitude and phase plots of measured or simulated impulse responses at a first position PI (e.g. left ear of a user) of individual output signals of a first loudspeaker 10 (dashed line) driving a first resonant enclosure structure with a remote first port opening 40, a port output (dotted line) of a second identical resonant enclosure structure driven by an identical remote second loudspeaker 14 as well as the time domain sum of these first two plotted signals when both speakers 10, 14 are driven with identical signals (bold solid line) or inverted signals with identical amplitude (thin solid line). The plots obtained by the described measurement and simulation are identical for both positions (PI and P2), and are therefore only shown for one of the positions PI (e.g. left ear position of a user) in Figure 11.

[0069] Although the plots illustrate the frequency range up to 1kHz in order to put the lower frequency range into a larger context, the frequency range below 120Hz is of special interest for the following considerations. The measured loudspeaker system shows a Helmholtz resonance close to 80Hz and a distributed half wavelength port resonance around 400Hz. The latter is achieved by the combination of three ports 301, 302, 303 with different lengths but equal cross sectional areas, similar to the arrangements described above with reference to Figure 5. Concerning low frequency extension, the plots illustrate that an acoustic summation of loudspeaker and port output below about 80Hz results in higher amplitude levels if the loudspeaker signals are out of phase (180° phase shift, phase or signal inversion), while an acoustic summation yields higher output levels between 80Hz and about 450Hz if the driving signals are in phase. For phase inversion (exact 180° phase shift) between left and right loudspeaker driving signals, the acoustic summation at first and second positions PI and P2 is generally identical and, therefore, only illustrated for one of the positions PI (left ear) in Figure 11. However, for relative phase shifts between the loudspeaker driving signals and, therefore, essentially also between the left and right ear signals of other than 180°, acoustic summation on ear positions PI and P2 may actually lead to different signal amplitudes on these respective positions PI, P2, depending on the phase shift applied to the port output signals by the resonant structures. The reason for this is the aforementioned cross-coupling between loudspeakers 10, 14 and respective port outputs 40, 44, which results in different relative phase shift between respective co-located speaker 10, 14 and port outputs 40, 44 close to the respective positions PI, P2. A relevant exception is the situation where the resonant structures apply no further phase shift to the respective rear side loudspeaker signals. This is essentially the case below the effective frequency region of the lowest enclosure resonance. Either identical or out of phase driving signals (180° relative phase shift btw. inverted) with equal amplitude on the respective loudspeakers generally yield the highest possible equal amplitude of the acoustic sum for both positions if the phase shift of the resonant structures is not equal to 90°. For a 90° phase shift of the resonant structures, in phase and out of phase loudspeaker driving signals will result in equal acoustic sum amplitude at both positions PI, P2.

[0070] Figure 12 illustrates effects of different relative phase shifts (undermost plot) between the loudspeaker driving signals on acoustic summation of loudspeaker and port output on positions PI (uppermost plot) and P2 (middle plot). Based on the aforementioned, the higher level magnitude of either of the sum signals for in phase loudspeaker driving signals ( Left Mic Sum, respectively Right Mic Sum (bold solid line) in Figure 12) or the sum signal for the inverted driving signals ( Left Mic Sum Fullrange Inversion, respectively Right Mic Sum Fullrange Inversion (thin solid line) in Figure 12) marks the highest amplitude that may be obtained with equal levels on both positions PI, P2.

[0071] If one position (e.g., PI) exceeds the highest possible equal amplitude level at any frequency, the other position (e.g., P2) will have an amplitude level at that frequency which is lower than the highest possible equal amplitude level. It can also be shown that the relative phase shift between the loudspeaker driving signals that yields the highest sum of the absolute signal at both positions (PI and P2) always results in equal signal level on both positions PI, P2. Assuming that binaural loudness summation within the human auditory system is comparable to a summation of the absolute signal level on left and right ear irrespective of relative phase, the sum of the absolute signal at both positions PI, P2 would be relevant for the loudness of the device as perceived by the user. Therefore, absolute signal levels on both positions may preferably be equal. This is also desirable for symmetry between left and right ear loudness in order to avoid a shift of sound image location perceived by the user to the side with higher loudness.

[0072] A relative phase shift between the simulated loudspeaker driving signals was induced by all pass filters applied to those signals. As can be seen in the undermost plot of Figure 12, a variable relative phase shift over frequency that was applied with the maximum phase shift reached between 40Hz and 50Hz, being either about 90° ( RelPhase90 in Figure 12 undermost plot, dash-dot line), about 120° ( RelPhasel20 in Figure 12, undermost plot, dotted line), or about 180° {RelPhaselSO in Figure 12, undermost plot, dashed line). The corresponding time domain sum signal amplitudes of loudspeaker 10, 14 and port output 40, 44 for positions PI or left mic (uppermost plot) and P2 or right mic (undermost plot) have been plotted with the same line style in Figure 12 as the respective plots of relative phase of loudspeaker driving signals and are labeled according to the maximum phase shift of the corresponding phase shift plots (e.g., Left Mic Sum 90Deg Phase Shift corresponds to RelPhase90). As can be seen in Figure 12, for the simulated loudspeaker setup the relative phase shift curve with about 180° maximum shift (. RelPhaselSO , dashed line) results in a sum signal over frequency for which the amplitude level runs closely below the maximum level possible (for equal level on both positions) for the first position (left btw. PI) and at or slightly above this level (at 80Hz) for the second position (right or P2). As the optimum relative phase for the loudspeaker driving signals regarding maximized output level would be 180° below about 80Hz and 0° above the same frequency, the transitions frequency range of the relative phase between 0° and 180° leads to a non-optimum signal summation. In order to improve signal summation, the transition frequency range for the phase could be narrower with a steep slope in the relative phase plot. Actual implementations of such steep relative phase slopes, however, may result in other objectionable effects like excessive resonant behavior of filters applied for phase control. Ultimately, there will always be a narrow frequency region in which the summation of direct loudspeaker output and tube output is not optimal, such that signal levels at positions PI and P2 will be different. Independent of the extend of the affected frequency region, such signal level differences may, for example, be compensated by means of inverse level compensation of the loudspeaker signals such that both positions PI, P2 or both ears of a user receive equal sound levels. For this purpose, a relative level difference between the loudspeaker signals may be induced that is either constant or variable over frequency.

Because signal level variations between the amplitude plots of Figure 12 are completely induced by relative phase shift between the loudspeaker driving signals, the maximum sound pressure that can be reproduced with the loudspeaker system varies accordingly. Therefore, any difference in ear signal level (e.g. left mic and right mic signal in Figure 12) means a difference in maximum possible sound pressure level at these positions, provided that the first and second loudspeakers and enclosures are essentially identical and positioned symmetrically with respect to both (typical) ear positions PI, P2 of the user.

[0073] As mentioned above, there are also negative effects of a phase inversion between the left and right ear positons or more general between first and second positions PI and P2, which may be the actual ear positons or merely the typical or expected ear positions of a user of the nearfield audio device. Hence, less relative phase shift between the loudspeaker driving signals and, ultimately, the first and second positions PI and P2 may be preferable for some applications. Relative maximum phase shifts of 120° and 90°, therefore, have also been simulated and the results are plotted in Figure 12. Signals with a 90° relative phase shift and an identical level add up to a +3dB relative signal level, while signals with a 120° relative phase shift add up to the same level of the initial signals. Nearfield audio devices with cross- coupled loudspeaker systems (e.g., as illustrated in Figure 10) may, for example, position loudspeakers 10, 14 and port outputs 40, 44 on the shoulders of a user below the ears. When the user turns his head, the relative position of the user’s ears and the respective loudspeakers and port outputs may vary. If the ear signals at the typical ear positions of the user’s ears are out of phase (inverted or 180° phase shift) and at equal signal levels, they may cancel each other completely at any position with equal free air distance from both typical ear positions. The actual degree of cancellation at any ear position distant from the typical ear positions may, among other factors, depend on the distance of the loudspeakers and port outputs from that position, the signal frequency and the size and shape of the user’s head. However, at any signal frequency for which the wavelength is large as compared to the dimensions of the user’s head, the signal level at the typical ear positions will be multiple times higher than along the intersection of the user’s head with a median plane with the user facing a horizontal direction in the front. The median plane, also called midsagittal plane, crosses the user’s head midway between the user’s ears, thereby dividing the head into essentially mirror- symmetrical left and right half sides. It is is perpendicular to a horizontal plane, which is also called transverse plane and which and divides the user’s head in an upper part and a lower part and is therefore, parallel to the ground surface. Due to the above, the signal level at the actual ear position of the user may decrease rapidly when the user turns his head while the audio device remains in the same position. If the relative phase at the ear positions is 120° or 90°, there is no cancellation of these signals. Therefore, signal level variation resulting from head movements may be much lower. A relative phase shift of, for example, 90° instead of 180° will also have less effect on the sound image perceived by the user and, therefore, may be preferred at least by some users. Generally, effects on sound image will reduce with relative phase shift.

[0074] Due to potential advantages of a lower relative phase shift between the ears of the user, maximum phase shifts between the loudspeaker driving signals of about 120° and 90° have been simulated and the resulting amplitudes of time domain signal summation of the respective direct loudspeaker and port output have been plotted in Figure 12. As can be seen in Figure 12, the more the relative phase shift of the loudspeaker driving signals deviates from 180°, the lower the sum signal amplitude and the higher the level difference between the ear positions below 80Hz becomes. However, as compared to the maximum possible level improvement over the case without any relative phase shift, especially the case with 120° maximum phase shift still provides substantial level increase even on the left ear position (left mic or PI) that generally shows lower levels than the right ear position (right mic or P2) due to the direction of relative phase shifts between loudspeaker signals (e.g. positive or negative relative phase shift). Even the 90° phase shift preserves more than 50% of the maximum level increase at the left ear. Therefore, both alternatives may be considered advantageous, depending on the priority of the dependent characteristics of a nearfield audio device for a given application. Level differences between positions PI and P2 or the typical positions of the ears of a user may, for example, be compensated by different driving signal levels applied to the loudspeakers 10, 14 closer to the respective positions PI, P2. The relative level difference between the loudspeaker driving signals may be either constant or variable over frequency.

[0075] In near field audio devices which provide individually controllable loudspeakers close to both ears (e.g., arrangements of Figures 9 and 10), active cancellation of detrimental port or pipe resonances may be applied. Although the driving signal of a loudspeaker in a resonant enclosure with a port or pipe (e.g., Figure 8) may be filtered such that any port resonances are equalized at the port output, the direct loudspeaker output signal will comprise notches corresponding to the equalized peaks in the port output. In order to avoid such adverse effects of port resonance equalizing, the port resonances may instead be cancelled acoustically by a cancellation signal radiated by a second independent loudspeaker. [0076] Port resonance cancellation for one port 40, 44 of the examples of Figures 9 and 10 may, for example, be applied based on the signal flow as illustrated in Figure 13. The transfer functions H in Figure 13 represent the transfer functions from the first loudspeaker 10 to the first positon (Hu to PI) and to the second position (H12 to P2) as well as from the second loudspeaker to the first (H21 to PI) and second position (H22 to P2). These transfer functions may include the loudspeaker transfer functions or merely the acoustic transfer functions from the loudspeakers 10, 14 towards the respective positions PI and P2 if the transfer functions of both loudspeakers 10, 14 are equal. Microphone symbols at the positions PI and P2 represent acoustic summation points. If at least the upper (first) loudspeaker 10 of Figure 13 is positioned close to the first position PI and mounted in a resonant enclosure structure 20, 30 with a port output 40 close to the second position P2 as exemplarily illustrated in Figures 9 and 10, the transfer function Hu is mainly affected by the direct signal from the first loudspeaker 10 located close to the first position PI towards the first position PI, while the transfer function H12 from the first loudspeaker 10 located close to the first position PI towards the second position P2 is primarily defined by the loudspeaker 10 and the acoustic transfer function of the enclosure 20, 30 of which the port opening is close to the second position P2. H12 may therefore comprise port resonances.

[0077] In order to cancel the port output at least partly in a frequency range containing port resonances, a cancellation signal may be radiated through the second loudspeaker 14, which is the lower loudspeaker in the exemplary signal flow of Figure 13. The cancellation signal may be derived from the input signal SI, which in the case of the signal flow shown in Figure 13 is also the input signal of the first loudspeaker 10, for example, by a filter with the transfer function C12, which generates a crossfeed signal applied to the second loudspeaker 14.

Therefore, the following equation (Equation 6.1) may be derived from the signal flow of Figure 13 with HBS representing a band stop filter with a certain attenuation in a limited frequency band, which suppresses the transfer function H12 in the respective frequency band.

C12 * H₂₂ + H₁₂ = H₁₂ * H_BS (6.1).

Equation 6.1 solved for the crossfeed transfer function C12, results in Equation 6.2:

A signal flow described by this equation 6.2 is illustrated in Figure 14.

[0078] In a practical implementation H12/H22 may, for example, be provided by a digital FIR (finite impulse response) filter while HB_S may be performed by a digital HR (infinite impulse response) filter. The filter providing the HB_S transfer function may be tuned for the desired suppression of at least one frequency region comprising port resonances. It should, however, be noted, that the phase response of HB_S will also affect the remaining crosstalk signal H12 * HB_S. Therefore, care needs to be taken to ensure that the phase of HB_S is essentially 0° at least in the frequency region around the Helmholtz resonance of the enclosure of the first loudspeaker 10. Otherwise, the second loudspeaker 14 may be charged with a high cancellation signal amplitude that merely causes a phase shift in the high output signal level of the port 40 of the enclosure 20, 30 of the first loudspeaker 10.

[0079] As an example for the cancellation of a frequency region comprising port resonances, a simulation based on the signal flow of Figure 13 has been conducted. In order to obtain a completely symmetrical cross-coupled loudspeaker system as is illustrated in Figure 10, measured impulse responses of an exemplary loudspeaker arrangement with resonant structure 30 and remote port output 40 as illustrated in Figure 8 were utilized. Measurements were taken at the ear positions of a dummy head with a loudspeaker 10 positioned close to the left ear, represented by PI of Figure 13 and the corresponding port output 40 close to the right ear, represented by P2 of Figure 13. The impulse response measured at PI (left ear), corresponding to Hu of Figure 13 and at P2 (right ear), corresponding to H12 of Figure 13, were also used for the corresponding transfer functions of the second loudspeaker 14 (H22 = H1 1 and H21 = H12), which leads to the aforementioned simulation of a completely

symmetrical cross-coupled loudspeaker system.

[0080] The crossfeed transfer function C12 has been approximated as a FIR filter with an FxLMS (Filtered x Least Mean Square) algorithm according to Equation 6.3 with HBP representing a band pass function.

The signal inversion represented by the initial negative unity factor of Equation 6.3 has not been part of the approximation and, therefore, is not comprised in the phase plot of Figure 15 of the resulting crossfeed filter (XF Filter, bold solid line). Within the simulation of the port resonance cancellation, the inversion was carried out as an independent processing step.

[0081] Figure 15 further illustrates the amplitude response of the approximated crossfeed filter {XF Filter, bold solid line) as well as the amplitude response of the band pass according to HBP of Equation 6.3. The maximum band pass amplitude has been chosen to be below unity in order to limit cancellation of the concerned frequency range to avoid a deep notch in the port transfer function when cancellation is applied. The FxFMS algorithm for approximation of C12 according to Equation 6.3 has been set up such that the target phase of the band pass function HBP is essentially zero, which avoids unnecessary phase changes on the remaining cross-coupling signal. Measured amplitude transfer functions of loudspeaker (corresponding to Hu of Figure 13) and port {Right Mic Port, bold solid line, corresponding to H12 of Figure 13) without port resonance cancellation are illustrated in Figure 16, in comparison to simulated signals for the same measurement positions {Right Mic Port with Cancellation, bold dashed line and Left Mic SPK with Cancellation, thin dashed line) with port resonance cancellation applied. In this example, the frequency region around the distributed half wavelength port resonance around 400Hz is attenuated moderately. As has been mentioned before, the distributed port resonance was achieved through passive damping by the combination of three ports 301, 302, 303 with different lengths but equal cross sectional areas, similarly as has been described previously with reference to Figure 6.

Contrary to mere equalizing of port resonances, the direct loudspeaker signal {Left Mic SPK with Cancellation, thin dashed line) does not show any distinct notch at the cancellation frequency range. Instead, cross coupling mainly through the port 44 of the second loudspeaker 14 results in slight modulation of the direct loudspeaker amplitude response, which can simply be compensated by equalizing filters on the input signal SI of Figure 13 upstream of the crossfeed signal tap. Outside the desired cancellation frequency range around the lowest port resonances, only insignificant cancellation effects occur in the present exemplary simulation. [0082] Figure 16 exemplarily illustrates measured port (Right Mic Port, bold solid line) and loudspeaker (Left Mic SPK, thin solid line) amplitude transfer functions and simulated versions of these amplitude transfer functions (Right Mic Port with Cancellation, bold dashed line and Left Mic SPK with Cancellation, thin dashed line) with port resonance cancellation applied.

[0083] As noted above, moderate port resonance cancellation was deliberately chosen for the simulated example. A cancellation frequency range around at least one port resonance may be attenuated by port resonance cancellation. For example, a frequency range of f_res/k to f_res*k, with f_es being the port resonance frequency, and with k>l .1 , or k>l .2. If desired, the port resonance frequency region may be cancelled to a lower residual level than shown in Figure 16. The cancellation range may also be extended for band limited crosstalk cancellation.

[0084] Port resonances and general leakage of the loudspeaker signals through the ports of the respective enclosures may drastically increase crosstalk between left and right ear positions of the user for a cross-coupled loudspeaker arrangement, as is exemplarily illustrated in Figure 10, when compared to loudspeaker arrangements without cross-coupling resonant structures. Several known methods for sound image control rely on binaural localization cues. Such methods may support anything from a simple stereophonic effect over the control of the width and distance of a perceived stereo image to synthesis of multiple virtual sound sources as perceived by the user for surround sound or 3D audio applications. Practical applications of such methods, therefore, include mere stereo sound playback with nearfield audio devices, for example, worn on the shoulders or the head of the user or integrated in head rests of car seats. Stereo playback may be enhanced by the extemalization of the sound image, for example, towards a position in front of the user, instead of the sound image being perceived at a position inside the user’s head. Multiple virtual sound sources may be synthesized that are perceived by the user at different locations which are not coincident with physical loudspeaker positions. Different channels of surround sound formats may be reproduced over those virtual sound sources. Virtual reality headsets may provide binaural audio signals including binaural localization cues that are to be preserved by a nearfield audio device. [0085] Binaural localization cues utilized by sound image control methods are interaural time differences (ITD) and interaural level differences (ILD). ITD refers to the time difference between the arrival of sound at each of the respective ears. ILD refers to the difference in sound pressure level at each of the respective ears. Both, ILD and ITD are part of the so called head-related transfer functions (HRTF) that describe the transfer function from any source position around a person towards both ears of the person and are thus frequency dependent. In order to synthesize virtual sound sources around a user of a nearfield audio device, control over ILD and ITD is required. Excessive crosstalk between left and right ear positions induces detrimental binaural cues as any signal reproduced at one ear generates a crosstalk signal with a certain relative amplitude response and time delay at the other ear. These spurious binaural cues interfere with the intentionally provided cues of sound image control methods and, therefore, render such methods partly or completely ineffective. This may severely limit application of nearfield audio devices with cross-coupling resonant structures.

[0086] As an example, Figure 17 illustrates an average level difference (ILD) between the direct sound to the ear on the source side (ipsilateral ear) and the indirect sound to the ear on the other side (contralateral ear). The level difference that is illustrated in Figure 17 was averaged over individually measured HRTF for a sound source from 60° azimuth and 0° elevation of 50 persons. Below about 70Hz the measurements were compromised by measurement noise and are thus not reliable. It is apparent that nearfield audio devices with a higher crosstalk amplitude than this average IFD curve will not be able to provide realistic directional cues, at least not for the corresponding source direction. However, even the synthesis of source directions with less IFD in the respective HRTF may suffer from the so- called precedence effect. The perceived spatial direction may be dominated by the first sound that arrives at the ears, even if subsequently a higher sound level is received. Therefore, nearfield audio devices with certain crosstalk delays may compromise or even prevent synthesis of directions that require longer interaural time delays (ITD), even if the interaural level difference for the respective direction is lower than the crosstalk amplitude of the system. The precedence effect gets weaker the lower the amplitude of the first arriving sound as compared to subsequent sounds is. This means that any reduction in crosstalk of the nearfield audio device improves potential virtual source synthesis performance of the device. [0087] Crosstalk may, for example, be cancelled with a signal flow as schematically illustrated in Figure 18. Starting from the single input signal flow of Figure 13, a second input S2 and crossfeed transfer function C21 have been added to the system of Figure 18. As previously described with respect to Figure 13, the transfer functions H describe the transfer functions from the respective loudspeakers 10, 14 to the positions PI and P2, which may, for example, be the ear positions of a user of a nearfield audio device comprising a loudspeaker arrangement according to Figure 10. Transfer functions H may include loudspeaker transfer functions, including respective enclosures 20, 24, although this is not necessary if

loudspeakers 10, 14 and enclosures 20, 24 are essentially identical. Microphone symbols at PI and P2 positions symbolize acoustic summation points at these respective positions.

Microphones may be placed at these positions to measure the corresponding transfer functions and convert the acoustic signals to electrical signals. Loudspeakers 10, 14 illustrated in Figure 18 also stand for summation points and transducers into the acoustic domain.

[0088] In order to cancel crosstalk from SI input to P2 position the following equation may apply.

C L2 * H₂2 + H₁₂ = 0 (7.1).

If equation 7.1 is solved for Cn, this results in (7.2).

Accordingly, crosstalk from S2 input to the first position PI may be cancelled if:

C21 * lu T 7^21— 0 (7.3).

Therefore: [0089] Crosstalk in the frequency region of the lowest enclosure resonance is inherent to the previously described principles of operation of cross-coupling resonant structures for increase of low frequency output levels. Fow frequency crosstalk is acceptable for many if not most spatial audio applications because bass is often presented monophonically. Therefore, low frequency crosstalk may be allowed, but should also be limited to the lowest possible frequency range for any spatial audio applications, including stereophonic sound.

Furthermore, crosstalk for a given nearfield audio device with cross-coupling resonant structures may partly be caused by enclosure structures (e.g. bass reflex port), and partly by free air crosstalk from the loudspeaker to the opposite ear. Especially for a higher frequency region, crosstalk may be dominated by the free air path, as resonant structures may have a low pass characteristic. The transfer function of such a free air path may vary dramatically, for example, with head movements of a user or if foreign objects (e.g. the user’s hand, clothing etc.) are located within that path. Therefore, it may also be beneficial to limit the crosstalk cancellation range towards higher frequencies. In order to control the frequency range in which crosstalk cancellation is applied, a control transfer function HSHP may be introduced, such that:

(7.5),

(7.6).

[0090] HSHP may, for example, comprise a band pass characteristic concerning the magnitude response. If applied according to Equations 7.5 and 7.6 in a signal flow as illustrated in Figure 18, HSHP must be zero phase and, therefore, may be understood as a variable control factor over frequency. It may, however, be linear phase if corresponding delays are inserted between the crossfeed signal taps and the loudspeakers 10, 14, as is illustrated in Figure 19 with delay units Du and D22. Alternatively, HSHP may have any phase response, as long as it is compensated by transfer functions Du and D22, which may generally have an all pass characteristic with an equal phase response as HSHP. These, however, are merely examples for signal flow topologies suitable for crosstalk cancellation.

[0091] Another exemplary signal flow that supports crosstalk cancellation is illustrated in Figure 20. In this arrangement, the crossfeed signals for crosstalk cancellation are applied in a recursive structure. This can compensate for higher order crosstalk, which may otherwise alter the direct transfer function of the loudspeaker system as the crosstalk cancellation signal itself also exhibits crosstalk. Equations 7.1 to 7.6 also apply for this signal flow of Figure 20. Due to the recursive nature of that signal flow, the magnitude of C12 and C21 generally needs to be at or below 1. Otherwise, the recursive crossfeed structure may become instable. This means that, according to Equations 7.2 and 7.4, crosstalk cancellation may only be applied for those frequencies where the magnitudes of the direct transfer functions of the audio system are lower than the magnitudes of the corresponding indirect transfer functions. With the control function HSHP of Equations. 7.5 and 7.6, the crossfeed transfer functions C12 and C21 may be controlled, such that their magnitude response is below 1. This may be required at resonance frequencies of the cross-coupling resonant enclosure structures. As mentioned before, the lowest resonance frequency is beneficial for bass improvement and, therefore, typically does not require cancellation. Half wavelength tube resonances, however, may require passive damping to avoid magnitude peaks above the direct signal path magnitude. The passive resonance damping methods described above may be applied in order to enable recursive crosstalk cancellation in such cases.

[0092] If a near field audio device comprises multiple loudspeakers for each ear of a user, crosstalk cancellation may be applied for some or all of these loudspeakers, either

individually or in combination. Generally, the loudspeaker symbols 10, 14 in Figures 13 and 18 to 20 may each represent a group of loudspeakers that are ultimately driven by a single source signal SI, S2, which is the input signal or the sum of the input signals of the respective loudspeaker symbols 10, 14. Individual loudspeakers forming a group of loudspeakers may be connected in parallel or in series to each other or to individual driving amplifiers.

Loudspeakers that are arranged in groups may be driven by identical or individual loudspeaker input signals. For example, crossover filters may control the frequency range supplied to individual loudspeakers of a loudspeaker group. Loudspeakers may be located at different positions within a resonant enclosure structure. If an arrangement comprises a group of loudspeakers, the transfer functions H are to be understood as transfer function from the input of the group of loudspeakers to the respective positions PI, P2.

[0093] Similarly, the microphone symbols in Figures 13 and 18 to 20 may each represent a group of microphones for which the output signals are combined in a suitable way during the measurement of the loudspeaker transfer functions H. Individual microphones in a group of microphones may be located at different positions in order to enable spatial averaging during the transfer function measurements. Therefore, positions PI, P2 may be understood as average positions of a multitude of positions. Crosstalk cancellation signals that are derived from spatially averaged transfer functions may provide better cancellation over a larger space than filters derived from transfer functions to single locations.

[0094] For practical implementation, units C12 and/or C21 of Figures 13 and 18 to 20 may be implemented as FIR filters for which the filter coefficients may be derived analytically according to the required transfer functions. Another option is to approximate the coefficients of such filters for a desired transfer function or cancellation range. For a band-limited crosstalk cancellation example, coefficients of crossfeed filters according to Figure 18 as well as Equations. 7.5 and 7.6 have been approximated with an FxLMS algorithm for measured transfer functions H of a cross-coupling loudspeaker arrangement as illustrated in Figure 10. Crossfeed filters have been implemented as FIR filters. The measured loudspeaker arrangement comprised closely located loudspeakers 10, 14 and port outputs 40, 44 on the shoulders of a manikin below each ear. Furthermore, passive port resonance damping by means of three ports 301, 302, 303 with different lengths as described with reference to Figure 5 avoided peaks above direct path magnitude in the crosstalk paths. Transfer functions were measured with single microphones at the ear positions of the manikin.

[0095] Measured crosstalk cancellation performance for this exemplary loudspeaker and control signal flow arrangement is illustrated in Figure 21 for one input signal (e.g. SI). Figure 21 illustrates the direct (bold line) and crosstalk (thin line) signal magnitude without (upper plot) and with (lower plot) crosstalk cancellation applied. In this example, reduced crosstalk between about 120Hz and 6kHz was achieved. In the given example, crosstalk cancellation reduces crosstalk to less than -lOdB between 200Hz and 20kHz. Without crosstalk cancellation, the only contiguous frequency range spanning more than an octave where crosstalk is below - lOdB is between 5kHz and 20kHz. This is, however, merely an example. Crosstalk for a device incorporating any combination of the disclosed active (signal processing) and passive (acoustic) measures for crosstalk control below - lOdB may be between 300Hz and 3kHz or between 200Hz and 16kHz. Effects of crosstalk cancellation on the direct transfer function of the loudspeaker were not compensated. Such a compensation may generally be achieved by applying simple equalizing filters on the input signal (e.g. SI).

[0096] Bode plots of the transfer function (C12) of the applied crossfeed filter are illustrated in Figure 22. Besides the magnitude and phase (solid line) of the crossfeed filter for crosstalk cancellation, the magnitude (dashed line) for a shape filter (HSHP) is illustrated that was deployed during crossfeed filter coefficient approximation. Naturally, the approximated crossfeed filter does not implement the ideal transfer function according to Equation 7.5 accurately, but it comes sufficiently close, as the measured crosstalk cancellation shows.

[0097] Previously described signal processing steps may, for example, be combined in a signal flow as illustrated in Figure 23. The signal flow illustrated in Figure 23 is divided into sections B1 to B4, each of the sections B1 to B4 serving a different purpose. Sections B3 and B4 are identical to the arrangement illustrated in Figure 18 and have been described with reference to Figure 18 above. Within the first section Bl, input signals SI and S2 may be distributed over the outgoing signal lines. Transfer functions Fn and F22 in the forward paths may, for example, possess high-pass characteristics, while transfer functions F12 and F21 may comprise a low-pass behavior. These high- and low-pass transfer functions may be complementary to each other such that they sum up to unity gain if connected in parallel. In this case, input signals below a certain frequency may be fed forward according to factors d and cross-coupled according to factors i. For d = i = 0.5, a low frequency part of the respective input signals SI, S2 may be distributed evenly over the outgoing signal lines. Depending on the signal processing in the following sections of the signal flow of Figure 23, this may lead to an essentially equal distribution of the low frequency signal content over the loudspeakers 10, 14 and, ultimately, to equal signal levels at the first and the second position PI, P2. If d = 1 and i = 0, the complete input signals may be fed forward and, again depending on downstream signal processing, may mainly be applied to the respective loudspeaker 10, 14 in the forward path. If the acoustic crosstalk transfer functions in section B4 (H12, H21) show a higher magnitude in a lower frequency range (e.g. around an enclosure resonance frequency) than the direct transfer functions (Hu, H22), a lower frequency part of input signals SI and S2 may mainly be cross-coupled to the opposing positions (SI to P2 and S2 to PI). This, for example, is the case in the examples described with reference to Figures 16 and 21, which show higher port output signal levels than direct loudspeaker signal levels due to the Helmholtz resonance of the enclosure. If factors d = 0 and i = 1, the complete low frequency part of the input signals SI, S2 is cross-coupled within section B 1. In this case, subsequent acoustic crosstalk within the loudspeaker arrangement may partly be compensated by the reverse cross-coupling in Bl, such that the higher signal levels in the frequency region around the enclosure’s Helmholtz resonance are radiated close to the position corresponding to the respective input signal (SI close to PI and S2 close to P2), at which also the high frequency part of the input signals SI, S2 is radiated. For this purpose, the crossover frequency between high-pass transfer functions Fn and F22 and low-pass transfer functions F12 and F21 may be chosen as the lowest frequency above the lowest enclosure resonance where the magnitude of the direct signal and the crosstalk signal are equal (e.g. 140Hz for Figure 16 and 115Hz for Figure 21). As a result, merely a narrow frequency region around the crossover frequency may comprise high crosstalk levels. Above and below that frequency range, crosstalk may be relatively low despite cross-coupling enclosure structures.

[0098] Within signal flow section B2, transfer functions Pn and P22 may apply phase shift, such that a desired relative phase shift between the two signal lines is achieved. This may improve signal summation between direct loudspeaker 10, 14 and cross-coupled port output 40, 44 as described with reference to Figures 11 and 12. If relative phase shift is between (not at) 0° and 180°, resulting different signal levels at the positions PI and P2 or the typical positions of the ears of a user, may be compensated by different factors d, i in section Bl .

This means that factors d or i are different for signals added to the output of the Fn or F22 transfer functions or processing blocks in Figure 23. Or, in other words, F12 may be equal to F21 but Fn*d may be different from F21 *d because factors d in the different signal flow branches are different. The same may apply for factors i, such that the B2 section receives different low frequency signal levels on both respective inputs. In section B3, crosstalk cancellation signals may be derived from the direct signals as previously described with reference to Figures 18 to 22, in order to cancel acoustic crosstalk represented by transfer functions H12 and H21 of section B4 over a controlled frequency range.

[0099] Nearfield audio devices, which may comprise cross-coupled loudspeaker

arrangements as described above, may include, but are not limited to, wearable devices that may be worn on the body of a user 2. For example, such devices may be mounted on the head, neck or shoulders of a user 2. Head mounted devices may, for example, comprise virtual reality headsets, headphones or any other device that may be worn on the head and that positions loudspeakers somewhere close to the ears of a user 2. An exemplary open headphone is illustrated in Figures 24 and 25.

[00100] The exemplary headphone of Figures 24 and 25 is open in the sense that the user’s ears are not covered laterally. Ring shaped ear cups 52 merely encircle the ears instead of fully enclosing them. Therefore, the right ear of the user 2 can be seen in the side view of Figure 24. The ear cups 52 may be coupled to a headband 54. Other types of open headphones, however, may be open towards directions in front, behind, above and/or below the user 2. Because open headphones do not provide enclosed chambers around the ears, the generation of low frequency sound with sufficient sound pressure levels is challenging for devices with small form factors, and the solutions for low frequency sound pressure improvement disclosed herein may, therefore, be beneficial for such devices.

[00101] In the schematic illustration of an open headphone in Figure 25, two loudspeakers 10F, 10R each comprising an oval membrane shape are illustrated for the right ear cup 52. Only one ear cup 52 is illustrated in more detail, as the other ear cup may be essentially symmetrical to the right ear cup with respect to the median plane if worn by a user as illustrated in Figure 24. Furthermore, a roughly rectangular port output 44 is illustrated above the frontal loudspeaker 10F. Dashed lines in Figure 25 outline an exemplary internal loudspeaker chamber 20 and port 30 routing. Both loudspeakers 10F, 10R in the right ear cup 52 may share a common rear chamber 20 or have individual rear chambers, each rear chamber connected to a shared port 30. Although only a single port 30 is illustrated, the port 30 may comprise multiple tubular structures of equal or different length and/or cross section areas or shapes. The port 30 may run across the user’s head through the head band 54 connecting the ear cups 52. An equally configured port coming from the left ear cup may comprise an opening 44 above the frontal loudspeaker 10F of the right ear cup 52. Both ports may flip position from a backwards oriented part of the head band 54 at the connection to the respective rear loudspeaker chamber to a frontal oriented part of the head band 54 in order to exit in a frontal region of the respective ear cup 52. In Figure 25, this positional change is illustrated in a central part of the head band 54 above the user’s head. [00102] Exemplary cross-coupled resonant enclosure structures that may be implemented in headphones as illustrated in Figures 24 and 25 are illustrated in Figure 26. In Figure 26 a), two loudspeakers 10, 102, 14, 142 per side (or ear cup 52) share a common rear chamber 20, 24 with a port 30, 34 routed to the opposing side (other ear cup 52). As previously noted the ports 30, 34 may comprise a single tubular structure or a multitude of tubular structures with different lengths, like exemplarily illustrated in Figure 5.

[00103] Figure 26 b) illustrates two loudspeakers 10, 102, 14, 142 per side (or ear cup 52) within individual rear chambers 20, 202, 24, 242 coupled to a common port 30, 34 at different positions along the longitudinal axis of the port 30, 34. In this case, too, the ports 30, 34 that exit at respective openings 40, 44 at the opposing sides (or ear cups 52) may each comprise a multitude of tubular structures with different lengths (see, e.g., Figure 5) or merely a single tubular structure. With one or more loudspeakers coupled to each port 30, 34 along the longitudinal axis of the port 30, 34, there are multiple longitudinal dimensions that may be understood as the length of the port 30, 34. On the one hand, this may be the total or maximum length of the port (e.g. from the openings 40, 44 to the respective closed end of the ports in Figure 26 b). On the other hand, this may be a length between the respective opening 40, 44 of the port and any position at which a loudspeaker or loudspeaker enclosure is coupled to the port 30, 34. One or more of the described lengths may be varied over multiple different ports connected to the same loudspeakers. It is, however, important to note that any specifications for length or length variation (e.g. the total length variation range or the distribution of lengths over the total length variation range) apply to at least one uniformly defined length for all ports coupled to the same loudspeakers. Comparable enclosure structures like illustrated in Figures 24 to 26 may be integrated in any kind of head mounted device to provide a cross-coupled loudspeaker arrangement.

[00104] Cross-coupled loudspeaker arrangements may also be integrated in nearfield audio devices that may be worn around the neck and/or on the shoulders of a user 2. An exemplary device is illustrated in Figure 27. The device may be U-shaped, having a first end and a second end that are coupled by a middle section. The first end may be arranged close to one ear of a user and the second end may be arranged close to the other ear of the user. The device may comprise one loudspeaker 10, 14 below each ear, each loudspeaker 10, 14 being located next to a port output 40, 44. That is, one loudspeaker 10 may be arranged at the first end of the device and the second loudspeaker 14 may be arranged at the second end of the device. The ports 30, 34 may run through the device such that a port opening 40, 44 is arranged at the end opposite its corresponding loudspeaker 10, 14. Although only one loudspeaker 10, 14 per side of the user is illustrated, such a device may generally comprise multiple loudspeakers per side. For example, one loudspeaker 10, 14 per side located in front and one per side located behind the respective pinna if viewed from above, like in the perspective of the middle drawing of Figure 27. When worn by a user 2, the device may surround part of the neck and rest on the shoulders.

[00105] Figure 28 exemplarily illustrates cross-coupled ports 30, 34 running from respective rear or coupling chambers 20, 24 behind each loudspeaker 10, 14 to the loudspeaker on the other side, where they exit into free air through an opening 40, 44. These ports 30, 34 may again be understood as single ports or as a multitude of ports with individual dimensions (e.g., three ports 301, 302, 303, as shown in Figure 5, not illustrated in Figure 28). The enclosure structures in the illustrated device may resemble that of Figure 10, although other resonant enclosure arrangements may also be integrated in such devices, for example, resonant enclosure arrangements as shown in Figure 26.

[00106] It may be understood, that the illustrated nearfield audio devices are merely examples. Other nearfield audio devices may, for example, be integrated to headrests of seats in cars or other transportation vehicles. For example, lounge chairs or similar seating arrangements may also include nearfield audio devices with cross-coupled enclosure structures. Nearfield audio devices may further comprise additional loudspeakers with or without sound guides. For example, additional high frequency loudspeakers may be employed in nearfield audio devices.

[00107] While various embodiments of the invention have been described, it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible within the scope of the invention. In particular, the skilled person will recognize the interchangeability of various features from different embodiments. Although these techniques and systems have been disclosed in the context of certain embodiments and examples, it will be understood that these techniques and systems may be extended beyond the specifically disclosed embodiments to other embodiments and/or uses and obvious modifications thereof. Accordingly, the invention is not to be restricted except in light of the attached claims and their equivalents.

[00108] The description of embodiments has been presented for purposes of illustration and description. Suitable modifications and variations to the embodiments may be performed in light of the above description or may be acquired from practicing the methods. The described arrangements are exemplary in nature, and may include additional elements and/or omit elements. As used in this application, an element recited in the singular and proceeded with the word“a” or“an” should be understood as not excluding plural of said elements, unless such exclusion is stated. Furthermore, references to“one embodiment” or“one example” of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features. The terms“first,” “second,” and“third,” etc. are used merely as labels, and are not intended to impose numerical requirements or a particular positional order on their objects. The described systems are exemplary in nature, and may include additional elements and/or omit elements. The subject matter of the present disclosure includes all novel and non-obvious combinations and sub-combinations of the various systems and configurations, and other features, functions, and/or properties disclosed. The following claims particularly point out subject matter from the above disclosure that is regarded as novel and non-obvious.

Claims

1. A loudspeaker arrangement comprising:

at least one first loudspeaker (10, 102) arranged closer to a first position (PI) than to a second position (P2),

wherein each of the at least one first loudspeakers (10, 102) is acoustically coupled to a plurality of first tubular structures (30, 301, 302, 303),

wherein each of the first tubular structures (301, 302, 303) comprises an opening (40, 401, 402, 403) that is arranged closer to the second position (P2) than to the first position (PI), and each of the first tubular structures (301, 302, 303) is configured to receive sound from each of the at least one first loudspeakers (10, 102) and to emit sound through the respective opening (40, 401, 402, 403), and

wherein each of the first tubular structures (30, 301, 302, 303) has a length (1301, 1302, 1303) that is different from the lengths of each other first tubular structure (30, 301, 302, 303).

2. The loudspeaker arrangement of claim 1, further comprising at least one second loudspeaker

(14, 142) arranged closer to the second position (P2) than to the first position (PI).

3. The loudspeaker arrangement of claim 2,

wherein each of the at least one second loudspeakers (14, 142) is acoustically coupled to a plurality of second tubular structures (34),

wherein each of the second tubular structures (341, 342, 343) comprises an opening (44) that is arranged closer to the first position (PI) than to the second position (P2), and wherein each of the second tubular structures (34) is configured to receive sound from the at least one second loudspeaker (14, 142) and to emit sound through the respective opening (44), and

wherein each of the second tubular structures (341, 342, 343) has a length (134) that is different from the lengths of each other second tubular structure (341, 342, 343).

4. The loudspeaker arrangement claim 2 or 3, wherein the at least one second loudspeaker (14,

142) is configured to radiate sound that, at the second position (P2), attenuates the level of sound emitted by the first tubular structures (301, 302, 303) within at least one first cancellation frequency range.

5. The loudspeaker arrangement of claim 3 or 4, wherein the at least one first loudspeaker (14, 142) is configured to radiate sound that, at the first position (PI), attenuates the level of sound emitted by the second tubular structures (341, 342, 343) within at least one second cancellation frequency range.

6. The loudspeaker arrangement of any of claims 1 to 5, wherein the greatest length (1301, 1302, 1303) of all of the first tubular structures (301, 302, 303) and the shortest length (1301, 1302, 1303) of all of the first tubular structures (301, 302, 303) each deviate from an average length of all first tubular structures (301, 302, 303) by more than 5% and less than 35% of the average length of all first tubular structures (301, 302, 303).

7. The loudspeaker arrangement of any of claims 1 to 6, wherein the respective lengths (1301, 1302, 1303) of all of the first tubular structures (301, 302, 303) are distributed evenly on either a linear or logarithmic scale.

8. The loudspeaker arrangement of any of the preceding claims, wherein the average area of all cross sections of each respective first tubular structure (301, 302, 303) perpendicular to a longitudinal axis of the respective first tubular structure (301, 302, 303) is equal for all of the first tubular structures (301, 302, 303).

9. The loudspeaker arrangement of any of the preceding claims, comprising at least three first tubular structures (301, 302, 303) or at least five first tubular structures (301, 302, 303) configured to receive sound from the at least one first loudspeaker (10, 102) and to emit sound closer to the second position (P2) than to the first position (PI).

10. The loudspeaker arrangement of any of the preceding claims, wherein each of the first tubular structures (301, 302, 303) comprises at least one acoustic tube resonance for which the resonance frequency is different than for each other first tubular structure (301, 302, 303).

11. The loudspeaker arrangement of any of the preceding claims, wherein the first tubular structures (301, 302, 303) are configured to emit sound at adjacent positions that are arranged closer to the second position (P2) than to the first position (PI).

12. The loudspeaker arrangement of any of the preceding claims, wherein each of the at least one first loudspeaker (10, 102) is configured to radiate sound with inverted polarity from two respective sides of the loudspeaker (10, 102), and wherein the loudspeaker arrangement further comprises:

at least one first enclosure (20, 202), wherein each of the at least one first loudspeakers (10,

102) is mounted in a wall of one of the at least one first enclosure (20, 202) such that it radiates sound inside and outside the enclosure (20, 202) with inverted respective polarity; and

each of the first tubular structures (301 , 302, 303) is configured to guide sound that is generated on the inside of each of the at least one first enclosure (20, 202) to the outside through the opening (40, 44) arranged at one end of the respective first tubular structure (301, 302, 303).

13. The loudspeaker arrangement of any of claims 2 to 12, wherein the first tubular structures (301, 302, 303) are configured to emit sound at positions adjacent to the at least one second loudspeaker (14, 142).

14. The loudspeaker arrangement of any of claims 4 to 13, wherein a contiguous frequency range over which the level of sound emitted by the first tubular structures (301, 302, 303) is attenuated at the second position (P2) by sound radiated by the at least one second loudspeaker (14, 142) comprises at least one acoustic tube resonance frequency of each of the first tubular structures (301, 302, 303).

15. The loudspeaker arrangement of any of claims 2 to 14,

wherein a first signal (SI) is processed with at least one first filter (Cl 2) to receive at least one first cancellation signal,

wherein the first signal (SI) is either directly provided to the at least one first loudspeaker (10,

102), or delayed and/or filtered and afterwards provided to the at least one first loudspeaker (10, 102), thereby becoming at least a component of a loudspeaker input signal of the at least one first loudspeaker (10, 102), and

wherein the at least one first cancellation signal is provided to at least one of the at least one second loudspeaker (14, 142), the first cancellation signal thereby becoming at least a component of a loudspeaker input signal of at least one of the at least one second loudspeaker (14, 142).

16. The loudspeaker arrangement of any of claims 2 to 14, wherein an input signal of the at least one first loudspeaker (10, 102) is processed with at least one first filter (Cl 2) to receive at least one first cancellation signal, and the at least one first cancellation signal is provided to at least one of the at least one second loudspeaker (14, 142), the first cancellation signal thereby becoming at least a component of a loudspeaker input signal of at least one of the at least one second loudspeaker (14, 142).

17. The loudspeaker arrangement of any of claims 5 to 13, wherein the at least one first cancellation frequency range substantially equals the at least one second cancellation frequency range.

18. The loudspeaker arrangement of any of claims 2 to 17, wherein at least over one frequency range the at least one first loudspeaker (10, 102) and the at least one second loudspeaker (14,

142) are configured to radiate sound with a relative phase shift of at least 90°.

19. The loudspeaker arrangement of any of claims 2 to 18, wherein at least over one frequency range the at least one first loudspeaker (10, 102) and the at least one second loudspeaker (14, 142) are configured to radiate sound with a relative phase shift between 90° and 180°, and an either constant or frequency-dependent relative level difference, wherein the relative level difference at least partly compensates for sound level differences between the first position (PI) and the second position (P2) induced by the relative phase shift.

20. The loudspeaker arrangement of any of the preceding claims, wherein the loudspeaker arrangement is integrated to a device which is at least one of a headphone, a virtual reality or augmented reality headset, a personal audio device worn on the body of a user (2) and a headrest.

21. The loudspeaker arrangement of any of the preceding claims, wherein the first position (PI) corresponds to the position of a first ear of a user (2) of the loudspeaker arrangement and the second position (P2) corresponds to the position of the second ear of the user (2).

22. The loudspeaker arrangement of any of the preceding claims, wherein the loudspeaker arrangement is integrated in a nearfield device, and wherein the nearfield device is configured to be arranged on or close to the head, neck or shoulders of a user (2) of the nearfield device.

23. A loudspeaker arrangement comprising:

wherein the at least one first loudspeaker (10, 102) is acoustically coupled to at least one first tubular structure (30, 301, 302, 303),

wherein the at least one first tubular structure (30, 301 , 302, 303) is configured to receive sound from the at least one first loudspeaker (10, 102) and to emit sound at an opening (40) that is arranged closer to the second position (P2) than to the first position (PI), wherein the loudspeaker arrangement further comprises at least one second loudspeaker (14,

142) arranged closer to the second position (P2) than to the first position (PI), and wherein the at least one second loudspeaker (14, 142) is configured to radiate sound that, at the second position (P2), attenuates the level of at least sound emitted by the at least one first tubular structure (30, 301, 302, 303) within at least one first cancellation frequency range.

24. The loudspeaker arrangement of claim 23,

wherein the at least one second loudspeaker (14, 142) is acoustically coupled to at least one second tubular structure, and

wherein the at least one second tubular structure is configured to receive sound from the at least one second loudspeaker (14, 142) and to emit sound through an opening (44) that is arranged closer to the first position (PI) than to the second position (P2).

25. The loudspeaker arrangement of claim 24, wherein the at least one first loudspeaker (10, 102) is configured to radiate sound that, at the first position (PI), attenuates the level of sound emitted by the at least one second tubular structure (341, 342, 343) within at least one second cancellation frequency range.

26. The loudspeaker arrangement of any of claims 23 to 25,

27. The loudspeaker arrangement of any of claims 23 to 25, wherein an input signal of the at least one first loudspeaker (10, 102) is processed with at least one first filter (Cl 2) to receive at least one first cancellation signal, and the at least one first cancellation signal is provided to at least one of the at least one second loudspeaker (14, 142), the first cancellation signal thereby becoming at least a component of a loudspeaker input signal of at least one of the at least one second loudspeaker (14, 142).

28. The loudspeaker arrangement of any of claims 25 to 27, wherein the at least one first cancellation frequency range substantially equals the at least one second cancellation frequency range.

29. The loudspeaker arrangement of any of claims 23 to 28, wherein each of the at least one first loudspeakers (10, 102) is configured to radiate sound with inverted polarity from two respective sides of the loudspeaker (10, 102), and wherein the loudspeaker arrangement further comprises: at least one first enclosure (20, 202), wherein each of the at least one first loudspeaker (10, 102) is mounted in a wall of one of the at least one first enclosure (20, 202) such that it radiates sound inside and outside the enclosure (20, 202) with inverted respective polarity; and

each of the at least one first tubular structure (301, 302, 303) is configured to guide sound that is generated on the inside of each of the at least one first enclosure (20, 202) to the outside through at least one opening (40, 44) at one end of the tubular structure (301, 302, 303).

30. The loudspeaker arrangement of any of claims 23 to 29, wherein each of the at least one first tubular structure (301, 302, 303) is configured to emit sound at a position adjacent to the at least one second loudspeaker (14, 142).

31. The loudspeaker arrangement of any of claims 23 to 30, wherein at least over one frequency range the at least one first loudspeaker (10, 102) and the at least one second loudspeaker (14, 142) are configured to radiate sound with a relative phase shift of at least 90°.

32. The loudspeaker arrangement of any of claims 23 to 31 , wherein at least over one frequency range the at least one first loudspeaker (10, 102) and the at least one second loudspeaker (14, 142) are configured to radiate sound with a relative phase shift between 90° and 180°, and an either constant or frequency-dependent relative level difference, wherein the relative level difference at least partly compensates for sound level differences between the first position (PI) and the second position (P2) induced by the relative phase shift.

33. The loudspeaker arrangement of any of claims 23 to 32, wherein the loudspeaker arrangement is integrated to a device which is at least one of a headphone, a virtual reality or augmented reality headset, a personal audio device worn on the body of a user (2) and a headrest.

34. The loudspeaker arrangement of any of claims 23 to 33, wherein the first position (PI) corresponds to the position of a first ear of a user (2) of the loudspeaker arrangement and the second position (P2) corresponds to the position of the second ear of the user (2).

35. The loudspeaker arrangement of any of claims 23 to 34, wherein the loudspeaker arrangement is integrated into a nearfield device, and wherein the nearfield device is configured to be arranged on or close to the head, neck or shoulders of a user (2) of the nearfield device.

36. A method comprising:

emitting sound with at least one first loudspeaker (10, 102) that is arranged closer to a first position (PI) than to a second position (P2), wherein each of the at least one first loudspeakers (10, 102) is acoustically coupled to a plurality of first tubular structures (30, 301, 302, 303),

receiving sound emitted by the at least one first loudspeaker in each of the first tubular structures (30, 301, 302, 303), and emitting sound from each of the first tubular structures (30, 301, 302, 303) through an opening (40, 401, 402, 403) that is arranged closer to the second position (P2) than to the first position (PI), and

wherein each of the first tubular structures (301, 302, 303) has a length (1301, 1302, 1303) that is different from the lengths of each other first tubular structure (301, 302, 303).

37. The method of claim 36 further comprising:

emitting sound with at least one second loudspeaker (14, 142) that is arranged closer to the second position (P2) than to the first position (PI), wherein the sound emitted by the at least one second loudspeaker (14, 142) is configured to attenuate, at the second position (P2), the level of at least sound emitted by the first tubular structures (301, 302, 303) within at least one first cancellation frequency range.

38. A method comprising:

emitting sound with at least one first loudspeaker (10, 102) that is arranged closer to a first position (PI) than to a second position (P2), wherein the at least one first loudspeaker (10, 102) is acoustically coupled to at least one first tubular structure (301, 302, 303), receiving sound emitted by the at least one first loudspeaker (10, 102) in each of the at least one first tubular structures (301, 302, 303), and emitting sound from each of the at least one first tubular structures (301, 302, 303) through an opening (40) that is arranged closer to the second position (P2) than to the first position (PI), emitting sound with at least one second loudspeaker (14, 142) that is arranged closer to the second position (P2) than to the first position (PI), wherein the sound emitted by the at least one second loudspeaker (14, 142) is configured to, at the second position (P2), attenuate the level of at least sound emitted by the at least one first tubular structure (301, 302, 303) within at least one first cancellation frequency range.

39. The method of any of claims 37 to 38, wherein at least one of the at least one first cancellation frequency range comprises a contiguous frequency range around at least one acoustic tube resonance frequency (f_res) of at least one first tubular structure, and the contiguous frequency range extends from f_res/k to f_res*k, with k>l .1 or k>l .2.

40. The method of any of claims 36 to 39, wherein the first position (PI) corresponds to the position of a first ear of a user (2) and the second position (P2) corresponds to the position of the second ear of the user (2).