Noise-canceling headphones

The audio headset uses a combination of open-loop and feedback loops with adaptive gain control to optimize noise cancellation across frequencies, addressing the challenge of maintaining effective noise reduction in varying conditions.

FR3136308B1Active Publication Date: 2024-11-01DEVIALET
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
FR2022005379
Authority / Receiving Office
FR · FR
Patent Type
Patents
Current Assignee / Owner
Filing Date
2022-06-03
Publication Date
2024-11-01
Estimated Expiration
2042-06-03

AI Technical Summary

Technical Problem

Existing noise-canceling headphones face challenges in maintaining effective noise cancellation at high frequencies while adapting to low frequencies, with existing filters either reducing noise reduction performance or amplifying high-frequency noise due to variability in sound leakage and ear canal shape.

Method used

The audio headset employs a main open-loop anti-noise processing filter connected to a bank of elementary filters with variable gain amplifiers, combined with a feedback loop and adaptive gain control, using error and external noise signals to optimize noise cancellation across various frequencies and conditions.

Benefits of technology

The solution ensures robust noise cancellation performance across a wide frequency range by dynamically adjusting filter gains based on environmental and user-specific factors, minimizing noise amplification and maintaining effective noise reduction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000012_0000
    Figure 00000012_0000
  • Figure 00000013_0000
    Figure 00000013_0000
  • Figure 00000014_0000
    Figure 00000014_0000
Patent Text Reader

Abstract

Noise-canceling headphones The headphones comprise: - an electro-acoustic transducer (16); - at least one noise-canceling processing chain comprising: - external microphones (31) and internal microphones (32); - a processing filter (30) comprising: - mounted in parallel between the external microphone (31) and a summing mixer (54), at least one elementary filter (48A, 48B, 48C) each comprising a recursive filter and an elementary amplifier with variable gain (50A, 50B, 50C); - a unit (70) for controlling the gain of each elementary amplifier (50A, 50B, 50C) as a function of a signal from the external microphone (31) and an error signal (e) from the internal microphone (32). The processing filter comprises a main open-loop filter (44) connected at its input to the external microphone (31), and whose output is connected to the summing junction (54) and to the input of each elementary filter (48A, 48B, 48C). Figure for the abbreviation: 2
Need to check novelty before this filing date? Find Prior Art

Description

Title of the invention: Noise-reducing headphones

[0001] The present invention relates to an audio headset of the type comprising:

[0002] - an electro-acoustic transducer placed in a sound reproduction cavity;

[0003] - at least one anti-noise processing chain comprising:

[0004] - an external microphone for capturing ambient sound outside the cavity;

[0005] - an internal microphone for capturing internal sound in the cavity;

[0006] - a noise-reduction filter for processing the signal from the external microphone to produce an external anti-noise signal, which anti-noise processing filter comprises:

[0007] - mounted in parallel between the external microphone and a summer at the output of which the external anti-noise signal is obtained, at least one elementary filter each comprising at least one recursive filter and each associated in series with an elementary variable gain amplifier;

[0008] - a unit for controlling the gain of each elementary amplifier in function an external anti-noise setpoint signal from the external microphone and an error signal from the internal microphone;

[0009] - amplification means for exciting the electro-acoustic transducer to at least from the external anti-noise signal.

[0010] In noise reduction processing chains using the signal from the external microphone capturing ambient sound outside the cavity, the quality of the noise reduction processing depends greatly on the insulation of the sound reproduction cavity, and in particular on the seal provided between the ear and the mechanical structure of the headset delimiting the cavity.

[0011] To take into account possible leaks resulting from the fitting of the headset, it is known that the noise-cancelling processing filter is adaptive, so that the processing chain is modified over time.

[0012] For this purpose, it is known to use finite impulse response filters, known by the acronym FIR for Finite Impulse Response in English, which are easy to adapt. However, these filters are not implementable in a low latency process as in the case of a noise reduction system, the consequence being reduced noise cancellation performance.

[0013] It is known in this context to use recursive filters or infinite impulse response filters known by the acronym IIR for Infinite Impulse Response in English.

[0014] The implementation of such filters is described for example in document US 9,549,249. In this document, a bank of IIR filters, each associated with a variable gain amplifier, is provided. These filters are connected in parallel. Their inputs are connected to the external microphone and their outputs are summed to form a signal external noise reduction. A control unit defines the gains applied to each filter, based on the signals measured by the internal microphone and the external microphone.

[0015] At high frequencies, the signal measured at the internal microphone is different from the noise perceived at the eardrum. Consequently, the proposed structure has the disadvantage of amplifying the high-frequency noise at the eardrum when the algorithm minimizes the noise at the internal microphone. This limit can be overcome by reducing the bandwidth of the parallel filters, but the noise reduction is then also limited to this same bandwidth.

[0016] The aim of the invention is to propose an audio headset having a noise-cancelling processing algorithm which allows the adaptation of the external noise-cancelling filters at low frequencies where the optimal filters have the greatest variability, while maintaining noise-cancelling performance at high frequencies.

[0017] To this end, the invention relates to an audio headset of the aforementioned type, characterized in that the anti-noise processing filter comprises a main open-loop anti-noise processing filter connected at the input to the external microphone and the output of which is connected, on the one hand, to the summer at the output of which the external anti-noise signal is obtained, and, on the other hand, to the input of the or each elementary filter.

[0018] According to particular embodiments, the audio headset comprises one or more of the following characteristics:

[0019] - it comprises means for calculating the error signal so that the error signal is a combination of the signal provided by the internal microphone and the signal provided by the external microphone multiplied by a compensation filter taking into account the pressure difference between the pressure at the internal microphone and the eardrum;

[0020] - it has an input for an audio signal to be reproduced connected to the input of the amplification means, and it comprises means for calculating the error signal so that the error signal is a combination of the signal supplied by the external microphone and the audio signal to be reproduced multiplied by a secondary path estimation filter corrected by a secondary path feedback loop if it exists;

[0021] - it further includes a feedback loop to add to the anti-noise signal externally, at the input of the amplification means, an internal anti-noise signal equal to the signal supplied by the internal microphone multiplied by a closed-loop anti-noise processing filter;

[0022] - it includes means for calculating the externally connected anti-noise setpoint signal at the output of the main open-loop anti-noise processing filter comprising a secondary path estimation filter;

[0023] - at least one of the means for calculating the external anti-noise setpoint signal and the means of calculating the error signal includes subsampling means in order to to reduce the computing power required;

[0024] - it comprises means for measuring the energy of the external ambient sound and the unit gain control is suitable for restoring the gains of the various elementary filters to reference gains if the energy of the external ambient sound is lower than a predetermined threshold over a range of upper limit frequencies lower than 2500 Hz;

[0025] - it includes a detector of vocal activity or parasitic vibrations of the wearer of the headset and the gain control unit is capable of blocking the gains of the various elementary filters if vocal activity or parasitic vibrations of the headset wearer are detected;

[0026] - it comprises means for estimating the energy of the measured external ambient sound by the external microphone and the internal sound energy in the cavity measured by the internal microphone and the gain control unit is suitable for restoring the gains of the various elementary filters to reference gains if the internal sound energy is greater than the external ambient sound energy;

[0027] - elementary filters are filters whose frequency response decreases for frequencies above 8000 Hz.

[0028] The invention will be better understood on reading the following description, given solely by way of example and with reference to the drawings in which:

[0029] - [Fig.l] [Fig.l] is a schematic view of an audio headset according to the invention placed in one ear;

[0030] - [Fig.2] [Fig.2] is a diagram of the headset illustrating the signal processing in the helmet ;

[0031] - [Fig.3] [Fig.3] is a combination of the characteristic curves of the amplitude and of the phase depending on the frequency of filters implemented in the headphones;

[0032] - [Fig.4] [Fig.4] is a combination of the characteristic curves of the amplitude and of the phase as a function of the frequency of the Shat transfer function of the secondary path and the anti-noise processing chain; and

[0033] - [Fig.5] [Fig.5] is an amplitude curve of the transfer function of a filter compensation.

[0034] In [Fig.l] an ear 8 is shown diagrammatically in which an audio headset 10 is partially inserted. In the example considered, this audio headset constitutes an earpiece suitable for being worn on one ear. The user has two earpieces, one for each of his ears. Only one will be described here.

[0035] The earpiece comprises, as is known per se, a tip 12 inserted into the ear canal. This tip delimits a cavity 14 for sound reproduction. A housing 15 extends the cavity outside the ear. This housing receives, as is known per se, the electronic components of the headset.

[0036] An electroacoustic transducer 16 is arranged in this cavity opposite the auditory canal of the ear. This transducer is capable of emitting an anti-noise signal into the cavity 14 and possibly also of ensuring the reproduction of a sound signal such as music or voice.

[0037] Alternatively, the cavity 14 is delimited by a shell covering most of the ear. The two shells are connected by a headband to form an audio headset.

[0038] The transducer 16 is connected for its excitation to an amplifier 18 assumed to have unity gain receiving a digital signal to be reproduced through a digital-analog converter not shown.

[0039] The headset has an input 22 for a musical signal to be reproduced. This input 22 is formed for example by a Bluetooth or Wi-Fi receiver capable of receiving a digital sound signal.

[0040] The headset includes a noise-cancelling filter 30 whose output is connected to the amplifier 18.

[0041] Furthermore, the headset comprises an external microphone 31 capable of capturing the ambient sound FF outside the cavity 14. Alternatively, it comprises several external microphones, the filter 30 is then duplicated for each external microphone.

[0042] The headset further comprises an internal microphone 32 arranged in the cavity 14 for capturing the internal sound FB in the cavity. The internal microphone 32 is arranged in the emission axis of the transducer 16. The two microphones 31, 32 are connected to the noise-cancelling processing filter 30.

[0043] An accelerometer 34 is provided in the earpiece housing, this accelerometer also being connected to the noise-cancelling processing filter 30.

[0044] In [Fig. 2], the elements of the headset illustrated in [Fig. 1] bear the same reference number. Thus, we find the two external microphones 31 and internal 32, the transducer 16 connected to the amplifier 18. The secondary path between the output of the noise-cancelling filter 30 and the internal microphone 32 is designated by the letter S in [Fig. 2],

[0045] Similarly, the primary path from the external microphone 31 to the internal microphone 32 is designated by the letter P, it corresponds to the passive attenuation of the headset.

[0046] The primary and secondary paths are shown in dotted lines, since they do not correspond to a part of the electronic circuit but are part of the signal processing chain shown here.

[0047] The anti-noise processing filter 30 comprises an external open-loop anti-noise processing chain 40 comprising a set of anti-noise filters, some of which are associated with a variable gain amplifier and a control unit 42 for the gain of each elementary variable gain amplifier.

[0048] More precisely, the processing filter 30 comprises a main branch 43 connecting the external microphone 31 to the amplifier 18. This branch 43 comprises a static open-loop main filter 44 whose transfer function is noted HFF.

[0049] At the output of the main filter 44 is provided a bypass branch 46 comprising a bank of parallel IIR elementary filters 48A, 48B, 48C, for example three in number. The filter bank comprises a number of filters equal to or greater than one. Each of the elementary filters is connected at the input to the output of the main filter 44 through a variable gain amplifier 50A, 50B, 50C, with gain noted gb g2 and g3 respectively.

[0050] The outputs of the elementary filters 48A, 48B, 48C are connected to a summer 52, itself connected to the output of the main static filter 44 on the main branch by a summer 54. At the output of the summer 54, an external anti-noise signal denoted s is obtained. ABE*

[0051] Filters 48A to 48C are all filters whose frequency response decreases at high frequencies, typically above 2000 Hz for an earpiece and above 8000 Hz for a headset comprising a shell covering most of the ear. They have a characteristic denoted W1, W2, W3 as illustrated in [Fig. 3] in the case of an earpiece.

[0052] In this figure, characteristic W1 is represented by short dotted lines and characteristic W2 is represented by long dotted lines and characteristic W3 is represented by a solid line.

[0053] Thus in the example considered, the cut-off frequencies are respectively 100 Hz, 500 Hz and 1000 Hz for the elementary filters 48B, 48C and 48A.

[0054] The three elementary filters are designed to act in different frequency ranges, so that the total gain of the external anti-noise processing chain 40 grouping the combined filters providing the anti-noise processing can modulate the static main filter 44 over the entire frequency range, potentially in a different manner depending on the frequencies.

[0055] The elementary filters are designed so that a linear combination of the static main filter 44 and the elementary filters 48A, 48B, 48C models all the transfer functions of the optimal external anti-noise filter which depend on P and S, which can vary for the different operating conditions which may occur.

[0056] The filter formed by the external noise processing chain 40 depends on the secondary path S and the primary path P. Both the secondary path S and the primary path P can vary depending on the placement of the earpiece in the ear, the shape of the ear canal, the level of leakage, etc. During the design, for each combination of the primary path P and the secondary path S the optimal filter is calculated. The elementary filters are designed so that the difference between this optimal filter and the main filter 44 can be described with the elementary filters.

[0057] An internal noise correction feedback loop 55 connects the microphone internal 32 to the output of the anti-noise filter 30 via a summer 56. This loop includes a static internal anti-noise filter 58 whose transfer function is noted HFB. This filter as known per se is chosen to maximize the attenuation at the eardrum (by maximizing its amplitude) while ensuring that the feedback loop remains stable. At the output of the internal anti-noise filter 58 is obtained an internal anti-noise signal noted sAB1.

[0058] This filter is a Feedback type noise filter (feedback loop) which makes it possible to increase the attenuation perceived by the eardrum. Alternatively, the internal noise correction feedback loop 55 does not exist.

[0059] The control unit 42 receives, as input, the raw external anti-noise signal noted sABEB obtained at the output of the main filter 44. This signal is sub-sampled by a sub-sampler 62 to go from a sampling frequency of, for example, 384kHz to 8kHz.

[0060] The signal thus subsampled is addressed in a filter 64 whose transfer function noted Shatest is an estimate of the transfer function between the signal which comes out of the adder 96 and the signal measured by the external microphone 32.

[0061] This transfer function Shat is an estimate of the transfer function of the secondary path S with the internal anti-noise correction feedback loop 55.

[0062] It is defined by Shat = S / (1-SHFB) where S is the secondary path and HFB is the transfer function of the internal noise filter.

[0063] A Shat estimate is obtained by fitting a series of biquad circuits to the above expression using transfer functions measured at the acoustic cavities as illustrated in [Fig.4] representing the amplitude and phase of the transfer function while the static internal noise filter 58 is active.

[0064] In this figure, the transfer function of the secondary path with the closed-loop anti-noise filter S / (1-SHFB) is represented in short dotted lines and the Shat estimate retained for the implementation in long dotted lines.

[0065] The signal obtained at the output of the filter 64 forms an external anti-noise reference signal denoted sRAHF. This signal is addressed in 3 filters 68A, 68B, 68C identical to the filters 48A, 48B, 48C. The external anti-noise reference signal thus filtered denoted xb x2, x3 is addressed at the input of an adaptive gain calculation unit 70.

[0066] The adaptive gain calculation unit 70 also receives as input an error signal e from the internal microphone 32.

[0067] For this purpose, the microphone 32 is connected to a sub-sampler 80 whose output is connected to the calculation unit 70 via a summer 82.

[0068] This adder receives on its other input a correction signal c from the external microphone 31 and making it possible to transform the internal microphone 32 into a virtual microphone placed at the level of the eardrum.

[0069] For this purpose, the microphone 31 is connected through a sub-sampler 90 to a compensation filter 92 with a transfer function noted Hcomp.

[0070] The compensation filter 92 is suitable for compensating the difference between the internal microphone 32 and the eardrum. Its transfer function Hcomp is for example as illustrated in [Fig.5].

[0071] The target curve noted Target is measured experimentally and is approximated by the transfer function Hcomp.

[0072] It can be seen from this curve that the compensation is significant above 800 hertz. At low frequencies, the compensation is very small since the acoustic pressure at the internal microphone 32 and at the eardrum are very similar. Thus, minimizing the error at the internal microphone 32 is equivalent to minimizing it at the eardrum.

[0073] In the absence of a reproduced musical signal, the error e is expressed in the form of e = s FB "t Hcomp Spp,

[0074] where sfb is the signal from the internal microphone 32;

[0075] Hcomp is the transfer function of the compensation filter 92;

[0076] Spp is the signal from the external microphone 31.

[0077] The input 22 for the audio signal to be reproduced is connected by a summer 96 to the main branch at the output of the anti-noise processing filter 30 so that the signal to be reproduced, the external anti-noise signal sAbe and the internal anti-noise signal sAbi are added to form the excitation signal denoted sE applied to the transducer 16 after amplification by the amplifier 18.

[0078] Advantageously, but optionally, an echo cancellation filter 120 with a transfer function equal to Shat receives as input the signal to be reproduced from the input 22 and is connected to the output of the microphone 32 by a summer 122 ensuring the sum of the internal signal sFB and the signal to be reproduced filtered by the filter 120 before being applied to the adaptive gain calculation unit 70.

[0079] The unit 70 is connected to the variable gain amplifiers 50A, 50B, 50C to ensure the fixing of their gain gb g2, g3 by implementing a least mean squares algorithm known by the acronym LMS for Least Mean Squares in English.

[0080] The equations for updating the gains gb g2, g3 are given below:

[0081] g=g+e.^- i L z xjnm

[0082] g + 2 r xjnin

[0083] g^g^ei^- 3 3 2 A;”* x.mm

[0084] where is the step of the least mean squares algorithm;

[0085] x> is the external anti-noise setpoint signal filtered by the characteristic filter W; ;

[0086] Pxs is the power of signal xi;

[0087] Pxjnîn is a fixed constant; and

[0088] e is the error signal from the adder 82.

[0089] Alternatively, and for faster convergence, a recursive least squares algorithm known by the acronym RLS for Recursive Least Square (RLS) algorithm in English is implemented by unit 70.

[0090] Furthermore, the external microphone 31, the internal microphone 32 and the accelerometer 34 are each connected to an input of an energy calculation unit 130 suitable for calculating energies allowing the control of a convergence unit 140 itself connected to the adaptive gain calculation unit 70 to stop the adaptive calculation of the gains as a function of the values ​​of the calculated energies.

[0091] The energy calculation unit 130 is capable of calculating the energy of the external ambient sound picked up by the microphone 31 in the frequency range of 50 to 2000 hertz. This energy is denoted PFF. Similarly, over the same frequency range, it is capable of calculating an energy of the internal sound denoted PFB corresponding to the energy of the signal picked up by the internal microphone 32.

[0092] Finally, it is suitable for calculating the acceleration energy noted Pacc corresponding to the energy of the signal measured by the accelerometer 34 in the frequency band from 70 to 1500 hertz.

[0093] From these calculated energies, the convergence control unit 140 is able to control the gain calculation unit 70.

[0094] For this purpose, the energy of the external ambient sound PFF is compared to a threshold. If the value of the energy of the external ambient sound PFF is lower than this threshold for a predetermined duration, for example one second, thus indicating that the external ambient sound is low, then the gains gi, g2, g3 are set to predetermined reference values, without taking into account the gain evolution algorithm implemented by the adaptive gain calculation unit 70.

[0095] Similarly, if the energy of the internal sound PFB is much greater than the energy of the external ambient sound PFF, for example by a ratio greater than 10, the gains gb g2, g3 are restored to initial reference values. The algorithm for defining the gains gb g2, g3 is restarted only when the energy of the internal sound PFB is no longer much greater than the energy of the external ambient sound PFF.

[0096] Finally, if the acceleration energy Pacc is greater than a threshold value, thus reflecting vibrations of the user's skin representative of words spoken by the user or of parasitic vibrations, the gains of the different elementary filters are blocked at the current gains and the algorithm for defining the gains is stopped until the acceleration energy Pacc is less than the predetermined threshold value.

Claims

Claims

1. Headphones comprising: - an electro-acoustic transducer (16) placed in a sound reproduction cavity (14); - at least one anti-noise processing chain comprising: - an external microphone (31) for capturing ambient sound outside the cavity (14); - an internal microphone (32) for capturing sound internal to the cavity (14); - a filter (30) for anti-noise processing of the signal from the external microphone (31) to produce an external anti-noise signal (sABE), which anti-noise processing filter (30) comprises: - mounted in parallel between the external microphone (31) and a summer (54) at the output of which the external anti-noise signal (sABE) is obtained, at least one elementary filter (48A, 48B, 48C) each comprising at least one recursive filter and each associated in series with an elementary variable gain amplifier (50A, 50B, 50C);- a unit (70) for controlling the gain of each elementary amplifier (50A, 50B, 50C) as a function of an external anti-noise setpoint signal (srabe) from the external microphone (31) and an error signal (e) from the internal microphone (32); - amplification means (18) for exciting the electro-acoustic transducer (16) from at least the external anti-noise signal (s abe)" characterized in that the anti-noise processing filter comprises a main open-loop anti-noise processing filter (44) connected at the input to the external microphone (31) and the output of which is connected, on the one hand, to the adder (54) at the output of which the external anti-noise signal (s ABE) is obtained, and, on the other hand, to the input of the or each elementary filter (48A, 48B, 48C).;

2. Headphones according to claim 1, characterized in that it comprises means (82) for calculating the error signal (e) so that the error signal is a combination of the signal supplied by the internal microphone (32) and the signal supplied by the external microphone (31) multiplied by a compensation filter (92) taking into account the pressure difference between the pressure at the internal microphone (32) and the eardrum.

3. Headphones according to claim 1 or 2, characterized in that it comprises an input (22) for an audio signal to be reproduced connected to the input of the amplification means, and in that it comprises means (122) for calculating the error signal (e) so that the error signal (e) is a combination of the signal supplied by the external microphone (31) and the audio signal to be reproduced multiplied by a secondary path estimation filter (120) corrected by a secondary path return loop (55) if it exists.

4. Headphones according to any one of the preceding claims, characterized in that it further comprises a feedback loop (55) for adding to the external anti-noise signal (sABEX at the input of the amplification means (18), an internal anti-noise signal (sABi) equal to the signal supplied by the internal microphone (32) multiplied by a closed-loop anti-noise processing filter (58).

5. Headphones according to any one of the preceding claims, characterized in that it comprises means for calculating the external anti-noise setpoint signal (sràee) connected to the output of the main open-loop anti-noise processing filter (44) comprising a secondary path estimation filter (Shat).

6. Headphones according to any one of the preceding claims, characterized in that at least one of the means for calculating the external anti-noise setpoint signal (sràee) and the means for calculating the error signal (e) comprises sub-sampling means (62, 80, 90) in order to reduce the computing power required.

7. Headphones according to any one of the preceding claims, characterized in that it comprises means for measuring the energy (PFF) of the external ambient sound and the gain control unit (70) is capable of restoring the gains of the different elementary filters to reference gains if the energy (PFF) of the external ambient sound is lower than a predetermined threshold over a range of frequencies with an upper limit lower than 2500 Hz.

8. Headphones according to any one of the preceding claims, characterized in that it comprises a detector of vocal activity or parasitic vibrations (Pacc) of the wearer of the headphones and the gain control unit (70) is capable of blocking the gains of the different elementary filters if vocal activity or parasitic vibrations (Pacc) of the wearer of the headphones are detected.

9. Headphones according to any one of the preceding claims, characterized in that it comprises means for estimating the energy (P FF) of the external ambient sound measured by the external microphone (31) and the energy (PFB) of the internal sound in the cavity measured by the internal microphone (32) and the gain control unit (70) is capable of restoring the gains of the different elementary filters to reference gains if the energy (PFB) of the internal sound is greater than the energy of the external ambient sound (PFF).

10. Headphones according to any one of the preceding claims, characterized in that the elementary filters (48A, 48B, 48C) are filters whose frequency response decreases for frequencies above 8000 Hz.