Generation of a drive signal for a sound transducer

FIELD OF THE INVENTION

The invention relates to a method and apparatus for generating a drive signal for a sound transducer and in particular, but not exclusively, for generating a drive signal for a low frequency loudspeaker.

BACKGROUND OF THE INVENTION

There is a general desire to be able to provide increased and improved sound levels from increasingly smaller sound devices. However, in particular for lower frequencies, high sound levels and small physical dimensions tend to be conflicting requirements. In particular, loudness is related to the amount of air that a loudspeaker displaces. However, the displacement is frequency dependant such that if the sound pressure level is kept constant then the lower the frequency, the bigger the required displacement. For these low frequencies the mechanical power handling of a loudspeaker usually forms the limiting factor instead of the electrical power handling and in order to provide the required sound levels relatively large physical dimensions tend to be needed.

European Patent Application EP04769892.3 discloses a system wherein a given sound pressure level can be achieved by a sound transducer with reduced physical dimensions. In accordance with the proposed system, a low frequency band of a signal is replaced by a fixed single frequency carrier signal with a frequency close to a resonance frequency of a loudspeaker. The amplitude of the carrier follows the amplitude of the signal components falling in the low frequency band. Thus, effectively a low frequency signal component is replaced by a single tone carrier with an amplitude equal to the signal component. Thus, by concentrating the low frequency signal into a single carrier frequency close to the resonance frequency of the loudspeaker, a much higher efficiency of the loudspeaker can be achieved. Furthermore, as the mechanical power handling and air displacement capability of a loudspeaker is highest around the resonance frequency, smaller dimensions of the sound transducer can be achieved by this approach.

However, although the approach can provide substantial advantages in many scenarios it also has some associated disadvantages. In particular, the approach distorts the low frequency sound signal and can result in a degraded sound quality.

Hence, an improved audio system would be advantageous and in particular a system allowing increased flexibility, facilitated implementation, improved audio quality, increased efficiency, reduced physical dimensions of a sound transducer and/or improved performance would be advantageous.

SUMMARY OF THE INVENTION Accordingly, the Invention seeks to preferably mitigate, alleviate or eliminate one or more of the above mentioned disadvantages singly or in any combination.

According to an aspect of the invention there is provided an apparatus for generating a drive signal for a sound transducer , the apparatus comprising: means for receiving an input signal; generating means for generating the drive signal from the input signal; sound level means for determining a sound level indication for the sound transducer , the sound level indication being dependent on the input signal; wherein the generating means comprises frequency compression means arranged to perform a frequency compression of a first frequency interval of the input signal to a second frequency interval corresponding to a resonance frequency of the sound transducer in response to the sound level indication. The invention may allow improved generation of a drive signal for a sound transducer. In particular, the invention may allow an improved trade-off between generated sound levels and signal distortion. The invention may allow reduced dimensions of the sound transducer and may in particular allow increased sound levels from smaller sound transducers. In some embodiments, the frequency compression means may be arranged to generate a first signal corresponding to a signal component of the input signal in the first frequency interval. A second signal having a frequency bandwidth limited to the second frequency interval may be generated from the first signal. The second signal may be generated to have an amplitude, power and/or energy measure corresponding to that of the first signal. Specifically, an amplitude detector may generate an amplitude measure for the first signal and an amplitude of the second signal may be set accordingly.

The sound transducer may be a device for converting an electrical drive signal into an acoustic signal. The sound transducer may specifically be a loudspeaker. It will be appreciated that any suitable means of defining or determining the first and/or second

frequency intervals may be used. For example, an edge of a frequency interval may be determined as a frequency wherein an attenuation of the signal falls below a given threshold.

The sound transducer may for example be a bass reflex loudspeaker system and the resonance frequency may e.g. be a Helmholtz frequency. The sound level indication is directly or indirectly dependent on the input signal. For example, the sound level indication may be calculated as a function of the input signal itself and/or from another signal derived from the input signal, such as the drive signal.

The sound level indication may be an indication of an estimated and/or predicted sound level that will result from feeding a drive signal generated from the input signal to the sound transducer. The sound level indication may for example be determined as a sound level measure resulting from presenting a non-frequency processed drive signal to the sound transducer and/or may be determined as a sound level measure resulting from presenting a frequency compressed drive signal to the sound transducer. In particular, the sound level indication may be indicative of a sound level that would result from feeding a drive signal corresponding to an undistorted input signal to the sound transducer.

The sound level means may be arranged to determine the sound level indication for the sound transducer when driven by the drive signal.

In accordance with some embodiments, there may be provided an apparatus for generating a drive signal for a sound transducer, the apparatus comprising: means for receiving an input signal; means for generating a first signal corresponding to the input signal in a first frequency interval; means for generating a second signal by a frequency compression of the first signal to a second frequency interval being smaller than the first frequency interval, the second frequency interval comprising a resonance frequency of the sound transducer; means for generating the drive signal, the drive signal being at least partly represented by the second signal in the first frequency interval; means for determining a sound level indication for the sound transducer; wherein the frequency compression is dependent on the sound level indication. Specifically, the frequency compression may be dependent on the sound level indication by a bandwidth of the second frequency interval being dependent on the sound level indication. The second signal may be generated to have an amplitude, power and/or energy measure corresponding to that of the first signal.

Specifically, an amplitude detector may generate an amplitude measure for the first signal and an amplitude of the second signal may be set accordingly.

In accordance with an optional feature of the invention, the generating means is arranged to perform frequency compression only if the sound level indication is indicative of a sound level above a threshold.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. In particular, the feature may allow no distortion to be introduced by the processing at low sound levels while allowing a sound transducer to be driven substantially at the resonance frequency for sound levels which cannot mechanically be supported by the sound transducer over a larger frequency range. The threshold may be a function of other parameters and characteristics, such as e.g. characteristics of the input signal and/or drive signal.

In accordance with an optional feature of the invention, the generating means is arranged to adapt the second frequency interval in response to the sound level indication.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. For example, for increasing sound levels, the concentration around the resonance frequency of the sound transducer may be increased.

In accordance with an optional feature of the invention, the frequency compression means comprises: means for generating a first signal corresponding to the input signal in the first frequency interval; an amplitude detector for generating an amplitude signal for the first signal; a frequency generator for generating a carrier signal in the second frequency interval; means for generating a frequency compressed signal by modulating the carrier signal by the amplitude signal, and means for generating the drive signal in response to the frequency compressed signal. This may allow particularly advantageous performance and/or facilitated operation. The approach may allow the sound transducer to be driven very close to the resonance frequency thereby increasing sound level output for given mechanical and/or physical characteristics. The feature may alternatively or additionally allow low complexity frequency compression which specifically may result in a highly concentrated frequency spectrum with has power and/or amplitude characteristics corresponding to the characteristics of the first signal.

The drive signal may be generated such that it substantially corresponds to the frequency compressed signal in the first frequency interval. The amplitude signal may

specifically be substantially limited to frequencies below 5 Hz. The second frequency interval may specifically have a lower limit above 10 Hz and an upper limit below 250 Hz.

In some embodiments the carrier signal may have a fixed frequency which specifically may correspond to the resonance frequency. Alternatively or alternatively, the carrier signal may have a dynamically varying frequency, e.g. dependent on the input signal and/or the first signal.

In accordance with an optional feature of the invention, the frequency generator is arranged to detect a first frequency of the first signal and to set a carrier frequency of the carrier signal in response to the first frequency. This may in many embodiments provide an improved sound quality. In particular, it may allow a psychoacoustic perception of the resulting acoustic signal to more closely resemble one from a non-distorted signal.

Any algorithm for detecting a frequency from a signal may be used. The frequency may for example be a dominant or an average frequency of the first signal. In accordance with an optional feature of the invention, the frequency generator is arranged to set the carrier frequency as a function of the first frequency, the function being dependent on the sound level indication.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation.

In accordance with an optional feature of the invention, the function defines a frequency reduction of a first frequency difference relative to a second frequency difference in response to the sound level indication, the first frequency difference being a frequency difference between the carrier frequency and the resonance frequency and the second frequency difference being a frequency difference between the first frequency and the resonance frequency.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation.

In accordance with an optional feature of the invention, the function defines the carrier frequency as a frequency having an offset to the resonance frequency, the offset

being determined as a continuous function of the second frequency difference and the sound level indication.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation.

In particular, a continuous dependency may allow an improved audio quality and may e.g. reduce or avoid audio artifacts relating to the sound level indication dependency to be introduced. The continuous function may specifically define the offset as the second frequency difference scaled by a scale factor (smaller than one) which is dependent on the sound level indication.

In accordance with an optional feature of the invention, the amplitude detector is arranged to generate the amplitude signal in response to a correlation of the first signal and the carrier signal.

This may allow improved audio quality and may in particular reduce the likelihood of stochastic noise being translated into an audible signal around the resonance frequency.

In accordance with an optional feature of the invention, the sound level means is arranged to generate the sound level indication in response to the first signal.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation. For example, an amplitude level may be determined for the first signal and the sound level indication may be determined as a function thereof. The function may e.g. be defined as a look-up table relating a frequency compression algorithm and/or characteristics to the characteristic of the first signal.

The feature may in particular allow the frequency compression to be controlled by characteristics of the input signal which are most likely to impact the performance of the sound transducer (namely that of the first frequency interval).

In accordance with an optional feature of the invention, the sound level means is arranged to generate the sound level indication in response to at least one of the input signal and the drive signal.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation. For example, an amplitude level may be determined for the input signal and the sound level indication may be determined as a function thereof. The function may e.g. be defined as a look-up table relating a frequency compression algorithm and/or characteristic to the characteristic of the input signal.

Using the input signal may in particular facilitate operation as this signal is readily available and in particular is available independently of and prior to the frequency compression being applied to the signal.

Using the drive signal itself may provide an increased accuracy as a closer correspondence to the actual performance of the sound transducer can be achieved.

In accordance with an optional feature of the invention, the sound level means is arranged to generate the sound level indication in response to a volume setting for the drive signal.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation.

In particular, the feature may in many embodiments allow a very low complexity determination of the appropriate frequency processing without requiring any complex processing of signals to derive the sound level indication. In particular, no actual signal characteristics need to be estimated or calculated. In accordance with an optional feature of the invention, the apparatus further comprises a detector for detecting an operational characteristic of the sound transducer and wherein the sound level means is arranged to generate the sound level indication in response to the operational characteristic.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer. Alternatively or additionally, the approach may facilitate implementation and/or operation.

In particular, the feature may allow an improved adaptation of the frequency compression to the actual performance and/or conditions experienced by the sound

transducer. Thus, a closer correspondence to the actual performance of the sound transducer can be achieved.

The operational characteristic may for example be a characteristic of the audio signal being generated by the sound transducer and/or may e.g. be a physical/mechanical characteristic of the sound transducer itself.

In accordance with an optional feature of the invention, a frequency of variations in the sound level indication is below 1 Hz.

This may allow improved performance and in particular may allow an improved trade-off between distortion, sound level capability and physical dimensions of the sound transducer.

According to another aspect of the invention there is provided a method of generating a drive signal for a sound transducer , the method comprising: receiving an input signal; determining a sound level indication for the sound transducer, the sound level indication being dependent on the input signal; and generating the drive signal from the input signal; wherein generating the drive signal comprises performing a frequency compression of a first frequency interval of the input signal to a second frequency interval corresponding to a resonance frequency of the sound transducer in response to the sound level indication.

These and other aspects, features and advantages of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which

FIG. 1 illustrates an example of a sound driver device in accordance with some embodiments of the invention;

FIG. 2 illustrates an example of a sound pressure output and cone displacement as a function of frequency;

FIG. 3 illustrates an example of a sound driver processor in accordance with some embodiments of the invention; FIG. 4 illustrates an example of a frequency compressor for a sound driver in accordance with some embodiments of the invention;

FIG. 5 illustrates an example of a frequency compressor for a sound driver in accordance with some embodiments of the invention; and

FIG. 6 illustrates an example of a flowchart of a method of generating a drive signal for a sound transducer in accordance with some embodiments of the invention.

DETAILED DESCRIPTION OF SOME EMBODIMENTS OF THE INVENTION The following description focuses on embodiments of the invention applicable to driving a small loudspeaker in response to a music signal. However, it will be appreciated that the invention is not limited to this application but may be applied to many other sound transducers and signals.

FIG. 1 illustrates and example of a sound driver device in accordance with some embodiments of the invention.

The sound driver device comprises an input processor or circuit 101 which receives an input signal that in the specific example is a music signal. The input signal may have a bandwidth from e.g. around 10Hz to around 3-10 kHz.

The input processor 101 is coupled to a drive processor 103 which is further coupled to a sound transducer which in the specific example is a loudspeaker 105. The drive processor 103 generates a drive signal from the input signal and feeds it to the loudspeaker 105 which in response generates a sound signal.

The characteristics and performance of sound transducers depend on the physical properties of the specific sound transducer. In particular, the air displacement characteristics are dependent on the physical characteristics and accordingly the sound level that can be produced by a speaker without mechanical distortion is dependent on the physical characteristics. Typically, larger physical dimensions are required for increasing sound levels and lower frequencies as the amount of air that needs to be displaced increases. Accordingly, a trade-off is typically required between the low frequency sound level capabilities and the physical dimensions.

Furthermore, sound transducers typically have one or more resonance frequencies wherein the physical characteristics provide a maximum sensitivity of the sound transducer. Furthermore, at these resonance frequencies the speaker cone or membrane movement or excursions is minimized for a given output sound level. Thus, at these frequencies an increasing sound level can be produced before the cone excursion become so large that the mechanical limitations of the sound transducer start to introduce distortions.

For example, for a conventional speaker in bass reflex enclosure, a typical sound pressure output and cone displacement as a function of frequency is illustrated in FIG. 2. Specifically, FIG. 2 illustrates the port output 201, the cone output 203, the total summed

pressure of port and loudspeaker cone 205 and the cone excursion 207 as a function of frequency.

As can be seen, the cone excursion has a minimum at the system resonance frequency, also known as the Helmholtz frequency, while the acoustic output of the total system is still high. At this frequency, the output of the cone is minimal while the bass reflex port is radiating sound. Thus at this frequency the generated sound level of the system may be maximized for a given allowable cone excursion.

The driver device of FIG. 1 is arranged to exploit these characteristics to provide an improved performance at a low frequencies. Specifically, the drive processor 103 comprises a frequency compressor which is arranged to compress a frequency band/interval/range of the input signal into a narrower more concentrated frequency band/ interval/range located around the resonance frequency. Specifically, a low frequency band may be compressed to a narrow band around the resonance frequency thereby allowing a higher sound level to be generated at low frequencies for a given size of the loudspeaker or equivalently a smaller speaker may be used for a given desired sound level.

Although such a frequency compression may introduce a distortion to the generated signal, this distortion is often acceptable in return for the increased sound level capability. In particular for low frequencies, the psychoacoustic impact of replacing a given frequency band with a more narrow frequency band with a corresponding power is acceptable in many embodiments.

Thus, the sound device drive processor 103 may be arranged to adapt a frequency interval of the input audio signal and may specifically comprise means for detecting first signal components in a first audio frequency interval and means for generating second signal components in a second audio frequency interval wherein the second audio frequency interval is (substantially) narrower than the first audio frequency interval. The drive processor 103 may further comprise means for controlling the amplitude of the second signal components in response to the amplitude of the first signal components. Furthermore, the second audio frequency interval is such that it comprises a maximum sensitivity resonance frequency for the sound transducerlO5. By generating second signal components in a second audio frequency interval which is substantially narrower than the first frequency interval, the amplitude of the second signal components being controlled in response to the amplitude of the first signal components, the energy of the audio signal is concentrated in the second frequency interval. As a result, the bandwidth of the first frequency interval can effectively be reduced and the

energy of the audio signal is concentrated in a substantially narrower (second) interval. This has the advantage that the energy of the audio signal can be concentrated in a interval in which a transducer is particularly effective and can produce increasing sound levels.

Thus, the described approach is based upon an insight that concentrating the signal components of a frequency interval in a relatively narrow band where sound transducers are most efficient allows a more effective use of the energy of the audio signals. For example, it may allow higher sound levels to be produced.

The sensitivity of the transducer is preferably the voltage sensitivity, that is, the ratio of the (output) sound pressure and the (input) voltage, although other measures are also possible, such as the efficiency, which may be defined as the ratio of the (output) acoustical power and the (input) electrical power.

The bandwidth reduction is especially effective at relatively low frequencies, as it allows low-frequency transducers to be used which are particularly efficient in a narrow frequency range. It is therefore preferred in many embodiments that the first frequency range has an upper boundary of not exceeding 200 Hz, preferably not exceeding 150 Hz, more preferably approximately 120 Hz.

In a typical embodiment, the second audio frequency interval is comprised in the first frequency interval. This implies that the second audio frequency interval is located within the first audio frequency interval and that no frequencies are generated outside the original (first) audio frequency interval. It effectively means that the second interval is a sub- interval of the first interval. Although the beneficial effect of the present invention is already attained when the second interval is a little narrower than the first interval, for example 10% (that is, has a bandwidth which is reduced by 10%), it is preferred that the second interval is substantially narrower, for example 50% or even more. Depending on the type of transducer being used, the second interval may be very narrow and may have a bandwidth of only a few hertz.

Accordingly, in many embodiments, advantageous performance can be achieved when the second audio frequency range spans less than 50 Hz, preferably less than 10 Hz, more preferably less than 5 Hz. The second frequency range may even comprise only a single frequency, for example the resonance frequency of a transducer.

Thus, in the system, a first frequency interval is mapped onto a second, smaller frequency interval which is preferably contained in the first frequency interval. E.g. the first frequency interval may be the interval from 20 Hz to 120 Hz, while the second interval may be the interval around 60 Hz, for example 55-65 Hz. This first interval

substantially covers the "low-frequency" part of an audio signal, whereas the second interval is chosen so as to correspond with a particular transducer, such as a loudspeaker, and will depend on the characteristics of the transducer. Specifically, the second interval is selected to include a resonance frequency of the transducer. The driver device of FIG. 1 comprises a sound level processor 107 which is coupled to the drive processor 103 and which is arranged to determine a sound level indication for the sound transducer. The sound level indication is directly or indirectly dependent on the input signal and provides an indication of the sound level that is desired from the sound transducer. The sound level indication may for example be an amplitude measure derived by low pass filtering the input signal or the drive signal or may e.g. be determined as a current volume or gain setting.

In the system of FIG. 1, the sound level indication is fed to the drive processor 103 which is arranged to perform the frequency compression in response to the sound level indication. For example, one or more characteristics of the frequency compression may be set in response to the sound level indication.

As a low complexity example, the frequency compression of the first frequency range into a second narrower frequency range may only be performed if the sound level indication meets a given criterion. Specifically, the frequency compression may only be performed if the sound level indication is indicative of a sound level above a given threshold. Thus, for low sound levels no frequency compression is performed whereas for higher sound levels frequency compression is applied thereby exploiting the increased sound level capability of the transducer around the resonance frequency.

The driver device of FIG. 1 may accordingly dynamically adapt its operation to the specific dynamic conditions. Specifically, for a low sound level frequency compression is typically not needed as the transducer is likely to be able to generate the required sound level with no distortion incurred by the physical characteristics of the transducer. However, at higher sound levels, the physical characteristics are likely to limit the performance and may e.g. introduce distortion or may not be able to generate the required sound level across the spectrum. Accordingly, frequency compression may be introduced to allow the sound level to be generated or to reduce the distortion introduced by the transducer 105.

For example, for a conventional bass reflex speaker, the cone displacement is modest for low sound levels resulting in the loudspeaker being effective over the entire frequency range and accordingly no frequency compression is required. However, at high sound levels the frequency compression is used to concentrate the signal energy at the

Helmholtz resonance frequency in order to reduce the excursion of the loudspeaker for a given sound level.

The frequency compression may inherently introduce a distortion to the drive signal. However, at high sound levels, this distortion may be acceptable in view of the reduced distortion of the transducer itself. Furthermore, the introduced distortion may be acceptable as it allows reduced dimensions for the transducer for a given required sound level. Thus, the approach may allow an improved trade-off between signal distortion, transducer distortion, sound level capability and the physical dimensions of the sound transducer. Furthermore, such an improved trade-off can be achieved with negligible impact on the performance achieved at lower sound levels.

Furthermore, the specific approach of using frequency compression to drive a transducer around a resonance frequency has been found to provide a particularly advantageous approach. In particular, the audio quality perception resulting from the frequency compression distortion has been found to be small. In particular for low frequencies it has been found that the psycho-acoustic impact of concentrating signal energy in a narrow frequency band around a resonance frequency is very low.

FIG. 3 illustrates an example of the drive processor 103 in more detail.

The drive processor 103 comprises a bandpass processor 301 which receives the input signal and which is arranged to generate a first signal corresponding to the input signal in the first frequency interval. Specifically, the bandpass processor 301 can comprise a bandpass or low pass filter which filters the input signal to generate the first signal. The first signal may e.g. be the result of a bandpass filtering by a 10Hz to 120Hz bandpass filter.

The drive processor 301 furthermore comprises a high pass processor 303 which generates a high frequency signal corresponding to the input signal in frequencies outside the first frequency interval. For example, the high pass processor 303 may be a high pass filter which removes all frequencies below 120 Hz. In some embodiments, the high pass signal may be generated as a residual signal following subtraction of the first signal from the input signal. For example, for a digital implementation, the signal samples being generated by the band pass processor 301 may be fed to the high pass processor 303 which may proceed to subtract these samples from a suitably delayed version of the input signal thereby generating the residual signal.

The band pass processor 301 is furthermore coupled to a frequency compressor 305 which is operable to perform the frequency compression to generate a frequency compressed signal from the first signal.

The frequency compressor 305 is coupled to a combiner 307 which is furthermore coupled to the high pass processor 303 and which is arranged to combine the frequency compressed signal and the high frequency signal from the high pass processor 303. The combiner 307 may for example perform a simple addition of the frequency compressed signal and the high frequency signal.

Thus, the output of the combiner 307 is a signal which corresponds to the input signal wherein a first frequency interval has been compressed. Thus, in the generated signal, the first frequency interval is represented by the frequency compressed signal whereas the remaining frequencies are represented by the high frequency signal. Thus, for frequencies above 120 Hz the input signal components are maintained whereas the energy of the frequencies below 120Hz is compressed into a smaller second frequency interval.

It will be appreciated that in some embodiments, the combined signal may in the first frequency interval comprise signal components from both an uncompressed and a compressed signal. For example, an attenuated version of the first signal may be added to a scaled version of the frequency compressed signal. In such an example, the weights of the first signal and the frequency compressed signal may be adjusted in response to the sound level indication.

In the specific example, the combiner 307 is coupled to a switch 309 which is furthermore fed the input signal. In the example, the switch 309 is arranged to switch between the input signal and the combined signal depending on the sound level indication. Specifically, for sound levels below a predetermined threshold, the input signal may be selected and for sound levels above the predetermined threshold, the combined signal may be selected.

The output of the switch 309 may be fed directly to the loudspeaker. It will be appreciated that the drive processor 103 may typically comprise audio amplifier functionality, volume adjustment functionality, digital to analog conversion functionality etc as will be known to the skilled person. For clarity and brevity such functionality is not illustrated in the figures or further described herein.

It will be appreciated that any suitable method of frequency compression may be used.

For example, in a digital implementation, the first signal may be converted to the frequency domain using an N-point Discrete Fourier Transform (DFT) and specifically an N-point Fast Fourier Transform. The resulting frequency bin values may then be concentrated into a smaller number of bins and the remaining bin values set to zero. For

example, N/2 consecutive bin values may be generated by averaging bin values of pairs of neighboring bins of the FFT. The resulting bin values are then allocated to the bins around the resonance frequency and the bin value of the non-assigned bins is set to zero. An inverse FFT can then be applied to generate a time domain version of the frequency compressed signal. This approach may accordingly correspond to compression of the bandwidth of the first signal by a factor of two with the compressed spectrum being located around the resonance frequency. It will be appreciated that the bandwidth of the frequency compressed signal may be varied by changing the number of bin values that are allocated values from the original transformed spectrum. For example, a frequency compression by a factor of four can be achieved by assigning bin values to only N/4 bins. As an extreme example, a bin value may be assigned to only a single bin corresponding to the entire frequency range being compressed into a single bin.

As another example, an N-point FFT may be used to transform the received first signal into the frequency domain. A number of additional bins may be added to generate an increased number of bin values with each bin value being set to zero. For example, an extra N zero value bins may be added resulting in a frequency spectrum of 2N bins. A 2N inverse FFT may be performed in these 2N bins resulting in a frequency compression by a factor of two (a sampling frequency multiplication by a factor of two will also result and accordingly a time domain decimation may be performed on the resulting signal). In some embodiments, the proportion of frequency bins that are assigned values from the bin values resulting from the FFT of the input signal is adjusted in response to the sound level indication. For example, for an increasing sound level the proportion of non-zero bins is reduced thereby resulting in an increased frequency compression to an increasingly narrow frequency band around the resonance frequency. FIG. 4 illustrates a specific example of a frequency compressor 305 of FIG. 3.

In the example, the frequency compressor 305 comprises an amplitude detector 401 which is fed the first signal and which generates an amplitude signal reflecting the amplitude of the input signal.

The amplitude detector 401 may for example consist in a single low pass filter with a cut-off frequency of less than 5 Hz. As another example, the amplitude detector 401 may comprise a peak detector or envelope detector with a suitable time constant.

The frequency compressor 305 furthermore comprises a frequency generator 403 which generates a carrier signal having a frequency falling in the second frequency interval. In the specific example, the carrier frequency is a fixed frequency that is set to be

identical or very close to the resonance frequency (such as a Helmholtz frequency) of the sound transducer.

The frequency compressor 305 furthermore comprises a modulator 405 which is coupled to the amplitude detector 401 and the frequency generator 403 and which is operable to modulate the amplitude signal from the amplitude detector 401 onto the carrier. The modulator 405 may specifically be implemented as a multiplier.

Thus, the output of the modulator 405 is a single tone having an amplitude corresponding to the amplitude of the first signal. Thus, the energy of the first signal in the first frequency interval is compressed into a narrow frequency range around the carrier frequency. Specifically, the frequency bandwidth of the resulting signal is equivalent to the frequency bandwidth of the amplitude signal, which typically may be 5 Hz or less.

In the specific example, the amplitude detector 401 is arranged to generate the amplitude signal in response to a correlation of the first signal and the carrier signal.

For a digital implementation, the amplitude detector 401 can multiply the input signal and the carrier signal followed by a low pass filtering of the result. The low pass filtered signal accordingly corresponds to the correlation of the carrier signal and the input signal.

For example, the following signals may be generated:

X(t) = x(t) ^■ ήn{2πft), Y(t) = x(t) ^■ cos{2πft) .

where x(t) is the input signal and f is the carrier frequency. The signals may be individually low pass filtered to generate the signals X'(t) and Y'(t). Accordingly, the amplitude signal can be generated as:

In many embodiments, the generation of a correlated amplitude may allow improved performance. For example, it may reduce the impact of noise on the generated signal. For example, if the first signal consists only in non-correlated white noise, standard amplitude detection would result in a non-zero (constant) amplitude estimate resulting in the noise being converted into a single constant tone output which may be very noticeable to the listener. However, for the described correlation, the non-correlated white noise will have no

correlation with the carrier and will accordingly result in a (close to) zero amplitude estimate. Thus, the conversion of noise into a single frequency tone can be avoided or reduced.

FIG. 5 illustrates another example of a frequency compressor 305. In the example of FIG. 5, the frequency compressor 305 also comprises an amplitude detector 401, a frequency generator 403 and a modulator 405. However, in contrast to the example of FIG. 4, the frequency compressor 305 of FIG. 5 is arranged to adjust the carrier frequency in response to the first signal. Specifically, the carrier frequency is adjusted in response to a frequency detected in the first signal.

In the example, the frequency compressor 305 comprises a frequency tracker 501 which is fed the first signal and which is arranged to detect a frequency of the first signal.

In the example, the frequency tracker 501 can track a frequency of the input signal by using the following recursion:

r _k = r _k _ _x + x _k _{γ [χ _k + χ _k _ ₂ - 2x _k _ _x r _k _ _x ]

where x represents samples of the first, γ is a design parameter determining the convergence speed and r* is defined as

with/ the sampling frequency, /the tracked frequency and k being a time index.

Thus, the output of the frequency tracker 501 is a frequency which specifically may correspond to a dominant or average frequency of the input signal. In the system, the carrier frequency is set dependent on the tracked frequency.

Specifically, the carrier frequency is determined as a function of the first frequency where the function depends on the sound level indication.

In the specific example, the frequency tracker 501 is coupled to a frequency adjuster 503 which implements the function. The frequency adjuster 503 is coupled to the frequency generator 403 and controls the output carrier frequency.

The function can specifically define a frequency reduction of the difference between the carrier frequency and the resonance frequency relative to the difference between

the tracked frequency and the resonance frequency. Thus, the function may define the first difference as:

where f _R is the resonance frequency and fp is the tracked frequency. The carrier frequency can then be determined from:

where f _c is the carrier frequency, f{x, s)\ ≤ x and f(x) is dependent on the sound level indication s.

Thus, in the example, the carrier frequency as a frequency determined as an offset to the resonance frequency of the sound transducer. Furthermore, this offset is determined as a continuous function of the second frequency difference and the sound level indication. By using a continuous function relating the offsets, a gradual adaptation of the frequency compression can be achieved. This may avoid or reduce audio artifacts such as clicks caused by sudden transitions or changes in the frequency compression.

The adaptation of the carrier frequency to be dependent on the tracked frequency may provide improved performance in many embodiments. For example, a music signal often comprises one dominant low frequency tone and the system may be adapted to track this frequency and adjust the carrier frequency accordingly. For example, for lower signal levels, the carrier frequency may be set substantially equal to the dominant frequency thereby resulting in a psycho acoustic perception which is very close to that corresponding to an undistorted input signal.

As the sound level increases, it may become increasingly important that the carrier frequency approaches the resonance frequency and thus the reduction of the frequency offset to the resonance frequency by the frequency adjuster 503 may be increased. At sound levels above a given level, the carrier frequency may be set to the resonance frequency regardless of the tracked frequency.

As a specific example, a linear scaling of the offset to the resonance frequency may be defined by the function. The scale factor may furthermore depend directly on the sound level indication. For example, the carrier frequency may be determined as:

where g(s) is a scale factor dependent on the sound level s. The value of the scale factor is between 0 and 1 dependent on the sound level. For example, g(s) may be determined as:

It will be appreciated that in different embodiments, different algorithms or functions for determining the sound level indication may be used.

For example, in some embodiments, the current volume setting may be detected and used directly as the sound level indication. Thus, a simple detection of the current volume setting can be used to control the frequency compression and thus the low- frequency operation of the device. Accordingly, in such an embodiment, no evaluation or processing of any of the signals is required to determine the sound level indication.

In other embodiments, the sound level indication may directly be determined in response to one or more of the signals being processed. For example, the sound level processor 107 may perform an amplitude detection on the input signal and set the sound level indication to correspond to the current amplitude of the input signal. As the desired volume level in many embodiments may be reflected by the amplitude of the input signal (e.g. if a gain/volume setting stage precedes the driving device), the amplitude detector may provide a good indication of the desired sound level.

As another example, an amplitude detection may be performed on the drive signal itself and the sound level indication may be said to reflect this amplitude level. This may in some embodiments allow improved performance as the sound level indication may provide a closer indication of the actual generated sound level rather than the desired sound level indicated by the input signal.

In some embodiments, the sound level indication may be generated in response to the first signal, i.e. in response to the band pass filtered signal. For example, an amplitude detection may be performed on the output of the band pass filter 301 and the resulting amplitude estimate may be used as the sound level indication.

In the examples, the amplitude detection may for example be an envelope or peak amplitude detection with a suitable time constant.

As another example, the device may be arranged to detect an operational characteristic of the sound transducer itself. The operational characteristic may for example be detected by a mechanical movement detector mounted on the sound transducer itself and/or may e.g. be detected by an external microphone placed close to the sound transducer. The sound level indication may be generated in response to the signal, for example as a signal reflecting the current mechanical movements or sound output of the sound transducer 105. In many embodiments the sound level indication varies only slowly relative to the audio frequencies being generated by the sound transducer. This may allow improved performance and may in particular ensure that the adaptation of the frequency compression does not become audible. Specifically the frequency of variations in the sound level indication may typically be kept below 1 -5 Hz. FIG. 6 illustrates a method of generating a drive signal for a sound transducer.

The method initiates in step 601 wherein an input signal is received. Step 601 is followed by step 603 wherein a sound level indication is determined for the sound transducer. The sound level indication is directly or indirectly dependent on the input signal. Step 603 is followed by step 605 wherein the drive signal is generated from the input signal. The generation of the drive signal comprises performing a frequency compression of a first frequency interval of the input signal to a second frequency interval corresponding to a resonance frequency of the sound transducer in response to the sound level indication. It will be appreciated that the above description for clarity has described embodiments of the invention with reference to different functional units and processors. However, it will be apparent that any suitable distribution of functionality between different functional units or processors may be used without detracting from the invention. For example, functionality illustrated to be performed by separate processors or controllers may be performed by the same processor or controllers. Hence, references to specific functional units are only to be seen as references to suitable means for providing the described functionality rather than indicative of a strict logical or physical structure or organization.

The invention can be implemented in any suitable form including hardware, software, firmware or any combination of these. The invention may optionally be

implemented at least partly as computer software running on one or more data processors and/or digital signal processors. The elements and components of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way. Indeed the functionality may be implemented in a single unit, in a plurality of units or as part of other functional units. As such, the invention may be implemented in a single unit or may be physically and functionally distributed between different units and processors.

Although the present invention has been described in connection with some embodiments, it is not intended to be limited to the specific form set forth herein. Rather, the scope of the present invention is limited only by the accompanying claims. Additionally, although a feature may appear to be described in connection with particular embodiments, one skilled in the art would recognize that various features of the described embodiments may be combined in accordance with the invention. In the claims, the term comprising does not exclude the presence of other elements or steps.

Furthermore, although individually listed, a plurality of means, elements or method steps may be implemented by e.g. a single unit or processor. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that a combination of features is not feasible and/or advantageous. Also the inclusion of a feature in one category of claims does not imply a limitation to this category but rather indicates that the feature is equally applicable to other claim categories as appropriate. Furthermore, the order of features in the claims do not imply any specific order in which the features must be worked and in particular the order of individual steps in a method claim does not imply that the steps must be performed in this order. Rather, the steps may be performed in any suitable order. In addition, singular references do not exclude a plurality. Thus references to "a", "an", "first", "second" etc do not preclude a plurality. Reference signs in the claims are provided merely as a clarifying example shall not be construed as limiting the scope of the claims in any way.