TWI899642B - Circuit, method for audio signal processing with excursion estimation compensation, and non-transitory storage medium - Google Patents

Abstract

Circuit, method for audio signal processing with excursion estimation compensation, and non-transitory storage medium are provided. The circuit comprises a delay circuit, compensation filter, excursion estimator, peak detector, gain determination circuit, and gain adjustment circuit. The delay circuit is for delaying a digital audio signal to output a delayed digital audio signal. The compensation filter is for generating a compensated digital audio signal according to the digital audio signal for excursion estimation compensation for a speaker type. The excursion estimator is for determining an estimated excursion signal for the speaker type according to the compensated digital audio signal. The gain determination circuit is for generating a gain setting signal according to the estimated excursion signal and a threshold value. The gain adjustment circuit is for generating an adjusted digital audio signal according to the gain setting signal and delayed digital audio signal.

Description

Circuit, method and non-transitory storage medium for audio signal processing with impulse estimation compensation

The present disclosure relates to a circuit and method for processing audio signals with speaker impulse protection, and more particularly to a circuit, method, and non-transitory storage medium for processing audio signals with impulse estimation compensation.

Speakers operate by varying the flow of electric current, which causes variations in the magnetic field generated by an electromagnet. These variations cause the speaker's cone to move. The cone's movement, in turn, creates pressure variations in the air, which in turn creates sound waves.

Excursion describes the distance a speaker travels from its normal resting position. Overexcursion occurs when the excursion limit of a speaker diaphragm is exceeded, which can lead to nonlinear damping, sound distortion, and ultimately speaker damage or a significantly shortened lifespan. Furthermore, when the diaphragm experiences overexcursion, the sound produced by the speaker becomes distorted due to nonlinearity, which can damage the speaker and voice coil.

The present disclosure aims to provide a circuit and method for audio signal processing with speaker impulse protection, and in particular, a circuit and method for audio signal processing with impulse estimation compensation.

To achieve at least the foregoing objectives, the present disclosure provides a circuit for processing audio signals with stroke estimation compensation. The circuit includes a delay circuit, a compensation filter, a stroke estimator, a gain determination circuit, and a gain adjustment circuit. The delay circuit is configured to delay a digital audio signal to output a delayed digital audio signal. The compensation filter is configured to generate a compensated digital audio signal based on the digital audio signal for use in stroke estimation compensation for a speaker type. The stroke estimator is configured to determine an estimated stroke signal for the speaker type based on the compensated digital audio signal. The gain determination circuit is used to generate a gain setting signal based on the estimated impulse signal and the threshold value. The gain adjustment circuit is used to generate an adjusted digital audio signal based on the gain setting signal and the delayed digital audio signal.

To achieve at least the foregoing objectives, the present disclosure provides a method for audio signal processing with stroke estimation compensation. The method comprises the following steps: a digital audio signal is delayed to output a delayed digital audio signal; a compensation filter generates a compensated digital audio signal based on the digital audio signal for use in stroke estimation compensation for a speaker type; an estimated stroke signal for the speaker type is determined based on the compensated digital audio signal; and a gain setting signal is generated based on the estimated stroke signal and a threshold value. Generates an adjusted digital audio signal based on the gain setting signal and the delayed digital audio signal.

In some embodiments of the circuit or method, the compensation filter has a filter response for compensating for the range estimation of the speaker type and is configured to generate a compensated digital audio signal based on the digital audio signal and the filter response.

In some embodiments of the circuit or method, a filter response of the compensation filter is associated with a specific frequency band for range estimation compensation for the speaker type, and the compensation filter is capable of amplifying a component of the digital audio signal for the specific frequency band to produce a compensated digital audio signal.

In some embodiments of the circuit or method, the filter response of the compensation filter is associated with a particular frequency band in which a speaker of the speaker type can operate to obtain maximum excursion.

In some embodiments of the circuit or method, the compensation filter can amplify a component of the digital audio signal based on the digital audio signal and the filter response for the particular frequency band to generate a compensated digital audio signal.

In some embodiments of the circuit or method, the compensation filter is based on an infinite impulse response (IIR) filter or a finite impulse response (FIR) filter.

In some embodiments of the circuit or method, the compensation filter is based on a shelf filter for performing impulse estimation compensation for speaker types having a shelf response.

In some embodiments of the circuit or method, the compensation filter is based on a peak filter for performing range estimation compensation for speaker types with peak responses.

In some embodiments of the circuit, the gain determination circuit includes a peak detector and a gain calculation circuit. The peak detector is configured to output a peak-indicating excursion signal based on an estimated excursion signal. The gain calculation circuit is configured to generate a gain setting signal based on the peak-indicating excursion signal and a threshold value.

In some embodiments of the method, generating a gain setting signal based on the estimated stroke signal and a threshold value includes: outputting a peak-indicating stroke signal based on the estimated stroke signal; and generating the gain setting signal based on the peak-indicating stroke signal.

100, 100A: Circuit

110: Delay circuit

120: Compensation filter

130: Stroke Estimator

140, 140A: Gain determination circuit

141: Peak Detector

145: Gain calculation circuit

150: Gain adjustment circuit

160: Digital-to-Analog Converter

170: Amplifier

180: Speaker

501, 503: Estimated displacement

502, 504: Actual displacement

800: Filter

DN: Actual negative displacement

DM: Process Model

DP: Actual positive displacement

G(t): Gain setting signal

OUT(t): Adjusted digital audio signal

S10, S20, S30: Steps

S40, S50: Steps

X(t): digital audio signal

Xc(t): Compensated digital audio signal

Xd(t): delayed digital audio signal

Xmax: stroke threshold

Y(t): Estimated impulse signal

Yp(t): Peak indication impulse signal

α: coefficient

ω: parameter

FIG1 is a schematic diagram showing the architecture of a circuit for audio signal processing with stroke estimation compensation according to an embodiment of the present disclosure.

FIG2A is a diagram showing an example of the impulse response of the first type of speaker.

FIG2B is a diagram illustrating an example of a filter response that the compensation filter of the circuit shown in FIG1 may be configured to implement for use in compensating for impulse estimation of the first type of loudspeaker illustrated in FIG2A , according to one embodiment of the present invention.

FIG3A is a diagram showing an example of the impulse response of the second type of speaker.

FIG3B is a diagram illustrating another example of a filter response that the compensation filter of the circuit shown in FIG1 can be configured to implement, according to one embodiment of the present invention, for use in compensating for the stroke estimation of the second type of loudspeaker illustrated in FIG3A .

FIG4 is a diagram showing an example of the actual negative displacement (DN), the stroke model (DM), and the actual positive displacement (DP) of a loudspeaker of the first type.

FIG5A is a diagram showing an example of an audio waveform.

FIG5B is a graph showing an example of the estimated displacement based on the stroke response and the actual displacement measured in response to the audio waveform shown in FIG5A without stroke estimation compensation.

FIG5C is a diagram showing an example of a second-frequency audio waveform.

FIG5D is a graph showing an example of the estimated displacement based on the stroke response and the actual displacement measured in response to the audio waveform shown in FIG5C without stroke estimation compensation.

FIG6 is a schematic diagram showing an embodiment of a circuit for audio signal processing with stroke estimation compensation based on the architecture shown in FIG1.

Figure 7 is a diagram showing an example of peak detection and gain calculation.

FIG8 is a diagram showing an example of dynamic range compression for peak detection.

FIG9 is a flow chart illustrating a method for audio signal processing with range estimation compensation according to an embodiment of the present disclosure.

To facilitate understanding of the purpose, features, and effects of the present disclosure, the following provides examples and accompanying drawings to illustrate the present disclosure in detail.

FIG1 illustrates the architecture of a circuit for audio signal processing with range estimation compensation according to an embodiment of the present disclosure. As shown in FIG1 , to reproduce audio from a digital audio signal (denoted by X(t)) using a speaker 180, a circuit 100 for audio signal processing with range estimation compensation can be employed to output a conditioned digital audio signal (denoted by OUT(t)) based on the digital audio signal (denoted by X(t)). Circuit 100 can perform audio signal processing with overrange protection and range estimation compensation for speaker 180. Circuit 100 can be configured or incorporated into an electronic device with computing or digital audio signal or data processing capabilities to facilitate audio reproduction, such as a mobile phone, tablet computer, laptop computer, desktop computer, multimedia player, digital audio player, or portable wireless or wired speaker. In practical applications, additional circuitry or audio signal processing devices, such as a digital-to-analog converter 160 for converting the conditioned digital audio signal OUT(t) into an analog signal, and an amplifier 170 for driving a speaker 180 based on the analog signal, can be electrically coupled between circuit 100 and speaker 180, as shown in FIG. 1 , for audio reproduction.

Circuit 100 includes a delay circuit 110, a compensation filter 120, an impulse estimator 130, a gain determination circuit 140, and a gain adjustment circuit 150. Alternatively, circuit 100 can be viewed as comprising a signal path and an impulse estimation path. The signal path includes the delay circuit 110 and the gain adjustment circuit 150 from the input to the output of circuit 100. The impulse estimation path includes the compensation filter 120, the impulse estimator 130, and the gain determination circuit 140 between the input and the gain adjustment circuit 150.

In the signal path, the delay circuit 110 is used to delay the digital audio signal X(t) to output a delayed digital audio signal Xd(t). Since audio signal processing in the estimated stroke path requires processing time, the delay circuit 110 can be configured to delay the digital audio signal X(t) by a delay time consistent with the processing time. This allows the gain adjustment circuit 150 to generate an adjusted digital audio signal OUT(t) based on the delayed digital audio signal Xd(t) and the gain setting signal G(t) from the estimated stroke path.

In the estimated range path, the compensation filter 120 is used to generate a compensated digital audio signal (denoted by Xc(t)) based on the digital audio signal X(t) for estimating the range of the speaker 180. The compensation filter 120 can generate the compensated digital audio signal Xc(t) based on the type of speaker 180 (also referred to as the speaker type) to compensate for the asymmetric range of the speaker 180. The range estimator 130 is configured to determine an estimated range signal Y(t) of the speaker 180 based on the compensated digital audio signal Xc(t). The range estimator 130 may be implemented to use the range response of the speaker 180 to determine the estimated range signal Y(t), where the estimated range signal Y(t) represents the estimated speaker range that varies according to the compensated digital audio signal. The gain determination circuit 140 is configured to generate a gain setting signal G(t) based on the estimated range signal Y(t) and a threshold. The gain determination circuit 140 may be implemented to determine whether the estimated range signal Y(t) meets a determination criterion based on the threshold. For example, if the estimated excursion signal Y(t) satisfies a determination criterion, such as whether the estimated excursion signal Y(t) is greater than a threshold, such as a predetermined excursion threshold or a maximum excursion limit allowed by the speaker 180, the gain determination circuit 140 generates a gain setting signal G(t) having a reduced gain value to prevent the speaker 180 from over-excursion.

The gain adjustment circuit 150 is configured to generate an adjusted digital audio signal OUT(t) based on the delayed digital audio signal Xd(t) and the gain setting signal G(t). For example, the adjusted digital audio signal OUT(t) can be generated based on the gain setting signal G(t) and the delayed digital audio signal Xd(t), as expressed by the equation OUT(t) = G(t)·Xd(t). The gain adjustment circuit 150 can be implemented using an amplifier, a multiplier, a mixer, or any other suitable circuit.

As will be explained below by way of example, circuit 100 can be implemented to perform audio signal processing to provide over-range protection with range estimation compensation for specific speaker types (e.g., speakers having specific types of characteristics such as range response).

For example, speaker 180 may be a first type of speaker. FIG2A illustrates an example of the excursion response of a first type of speaker. As shown in FIG2A , the example excursion response of the first type of speaker, represented by the solid curve, represents a low shelving response, wherein the horizontal dashed line represents an exemplary excursion threshold (or maximum displacement) (denoted by Xmax) of the first type of speaker. The first type of speaker may be various low-frequency speakers, such as a woofer or a subwoofer.

In another example, speaker 180 may be a second type of speaker. FIG3A illustrates an example of the excursion response of the second type of speaker. As shown in FIG3A , the example of the excursion response of the second type of speaker, represented by the solid curve, represents a peak response, wherein the horizontal dashed line represents an exemplary excursion threshold (or maximum displacement) (denoted by Xmax) of the second type of speaker. The second type of speaker may be various mid-range speakers.

In some embodiments, compensation filter 120 has a filter response for compensating for impulse estimation of a specific type of speaker 180 and is configured to generate a compensated digital audio signal Xc(t) based on the digital audio signal X(t) and the filter response. In some embodiments, compensation filter 120 is based on an infinite impulse response (IIR) filter or a finite impulse response (FIR) filter.

For example, when the speaker 180 belongs to the first speaker type and has a range response of a shelf response as shown in FIG2A , the compensation filter 120 can be implemented based on the filter response of a low shelf filter, as shown in FIG2B .

In another example, when the speaker 180 belongs to the second speaker type and has a peak response as shown in FIG3A , the compensation filter 120 can be implemented based on the filter response of the peak filter, as shown in FIG3B .

In some embodiments, the filter response of the compensation filter 120 is associated with a specific frequency band for range estimation compensation for a specific type of speaker 180, and the compensation filter 120 can amplify the component of the digital audio signal X(t) for the specific frequency band to generate the compensated digital audio signal Xc(t).

Furthermore, the filter response of the compensation filter 120 is associated with a specific frequency band in which the speaker 180 can operate to obtain maximum range.

For example, when speaker 180 belongs to the first speaker type and has a shelf response as shown in FIG2A , and compensation filter 120 is implemented based on the filter response of a low-shelf filter as shown in FIG2B , compensation filter 120 is associated with a specific frequency band below a frequency value (e.g., 30, 40, or 50 Hz) and is used to compensate for the excursion of speaker 180 . Components of digital audio signal X(t) in the specific frequency band are amplified to generate compensated digital audio signal Xc(t). As shown in FIG2A , speaker 180 can operate to have maximum excursion in the specific frequency band.

In another example, when speaker 180 belongs to the second speaker type and has a peak response as shown in FIG3A , and compensation filter 120 is implemented based on the filter response of a peak filter as shown in FIG3B , compensation filter 120 is associated with a specific frequency band from a first frequency value to a second frequency value (e.g., from 400 Hz to 1400 Hz) for use in compensating for the excursion of speaker 180 . Components of digital audio signal X(t) are amplified for the specific frequency band to generate compensated digital audio signal Xc(t). As shown in FIG3A , speaker 180 can operate to have maximum excursion in the specific frequency band.

Therefore, the compensation filter 120 can amplify the component of the digital audio signal X(t) for the specific frequency band according to the digital audio signal X(t) and the filter response to generate a compensated digital audio signal Xc(t).

The following example illustrates how the circuit 100 architecture can address the overexcursion problem caused by asymmetric excursion nonlinearity by using excursion estimation compensation. FIG4 illustrates the actual negative displacement (DN), the excursion model (DM), and the actual positive displacement (DP) of a first type of loudspeaker having an excursion response, such as a shelving response. In FIG4 , the solid line DM represents the excursion response of the loudspeaker.

Theoretically, as in a learned approach, the impulse response can be obtained symmetrically based on impulse modeling based on loudspeaker parameters (including force factor (Bl), mechanical damping factor (Rms), mechanical mass (Mms), mechanical compliance (Cms), and DC resistance (RE)), without considering asymmetric nonlinearities such as the variation of force factor with respect to displacement. The impulse response obtained in this manner is typically in the form of a transfer function of loudspeaker impulse or displacement (in mm/V) with respect to frequency. Therefore, the impulse estimator 130 can be implemented based on the impulse response obtained based on impulse modeling.

However, in actual applications, it can be observed that the loudspeaker 180 produces an asymmetrical positive excursion and an actual negative excursion in the frequency band where the maximum excursion may occur (e.g., frequencies below, for example, 60 Hz). As shown in FIG4 , the dashed line DN and the dot-dashed line DP respectively represent the actual negative excursion and the actual positive excursion of the loudspeaker 180 obtained through measurement, assuming that the maximum excursion of the loudspeaker 180 in this example is approximately 0.72 millimeters (mm), as indicated by the horizontal dashed line. As shown in Figure 4, at a frequency of 20 Hz, although the estimated stroke (shown by the solid line DM) obtained based on the stroke response is lower than the maximum withstand stroke (e.g., 0.72 mm), the actual negative stroke (shown by the dashed line DN) exceeds the maximum withstand stroke (e.g., 0.72 mm). Therefore, if the stroke estimation path lacks the stroke estimation compensation provided by the compensation filter 120, the stroke estimator 130 implemented based on the stroke response obtained through stroke modeling may produce erroneous estimates. In a worse case scenario, even though the gain determination circuit 140 is configured to generate a gain setting signal having a reduced gain value, overexcursion may still occur, particularly during a specific frequency band where the speaker 180 can operate to generate maximum excursion, resulting in damage to the speaker 180 due to an error in the estimated excursion.

Based on the example of FIG. 4 , the inventors of the present disclosure considered the application scenario regarding asymmetric impulses as described above and therefore introduced impulse estimation compensation by using compensation filter 120 in the impulse estimation path, as shown in FIG. Subsequently, as shown in FIG. 2B , compensation filter 120 can be implemented based on the filter response of a low-shelf filter. In this way, referring to FIG. 2B and FIG. 4 , the filter response can be configured to have a specific frequency band associated with the shelf response of the first type of loudspeaker 180 (e.g., frequencies below a frequency value such as 60 Hz). Accordingly, the compensation filter 120 can amplify components of the digital audio signal X(t) for a specific frequency band based on the digital audio signal X(t) and the filter response to generate a compensated digital audio signal Xc(t). Consequently, with the support of the compensation filter 120, the impulse estimator 130 can determine the estimated impulse signal Y(t) of the speaker 180 based on the compensated digital audio signal Xc(t), without underestimating the impulse even in a specific frequency band, such as a frequency of 20 Hz.

Furthermore, in implementation, for example, the range estimator 130 may utilize a time-domain impulse response (e.g., represented by H(t)) derived from the frequency-domain impulse response of the speaker 180 (e.g., represented by H(s)) to determine an estimated range signal Y(t) of the speaker 180 based on the compensated digital audio signal Xc(t). The impulse response may include a series of impulse response coefficients. The estimated range signal Y(t) may be determined by convolution of the compensated digital audio signal Xc(t) and the impulse response of the speaker 180, for example, as shown in the equation: Y(t)=Xc(t)*H(t).

The following examples illustrate the effect of asymmetric impulse on time-domain impulse estimation. FIG5A illustrates an audio waveform (e.g., a sinusoidal signal) at a first frequency (e.g., 20 Hz). FIG5B illustrates an example of an estimated displacement of a loudspeaker based on its impulse response (e.g., as shown by curve 501) and measured actual displacement of the loudspeaker (e.g., as shown by dashed line 502) in response to the audio waveform shown in FIG5A without deriving an impulse estimate compensation from the input signal of FIG5A. In both FIG5A and FIG5B , an audio signal with an amplitude of 1 V at a frequency of 20 Hz is associated with (or corresponds to) an estimated impulse (e.g., vibration of the loudspeaker) with an amplitude of 0.675 mm. For example, the speaker's excursion threshold (Xmax) is 0.675 mm. In Figure 5B , the actual excursion exhibits an offset, which can be determined by comparing the actual amplitude of the second half-cycle of displacement to the ideal amplitude. The offset is calculated as: offset = |-0.75| - |-0.675| = 0.075 mm. The offset in decibels (dB) is calculated as: offset in dB = 20 * log (|-0.75| / |-0.675|) = +0.915 dB. Figure 5C shows an example of an audio waveform at a second frequency (e.g., 40 Hz). FIG5D shows an example of an estimated displacement of a loudspeaker based on its displacement response (e.g., as shown by curve 503) and an actual displacement of the loudspeaker measured (e.g., as shown by dashed line 504) in response to the audio waveform shown in FIG5C, without any displacement estimation compensation derived from the input signal of FIG5C.

In some embodiments of the circuit 100, the gain determination circuit 140 may be implemented to determine whether the estimated excursion signal Y(t) is greater than or equal to a threshold, such as a predetermined excursion threshold or a maximum excursion limit of the speaker 180, and generate a gain setting signal G(t) having a reduced gain value based on the determination result to prevent over-excursion of the speaker 180. If the estimated excursion signal Y(t) (e.g., its magnitude or absolute value) is greater than the threshold, the gain determination circuit 140 outputs the gain setting signal G(t) having a reduced gain value (e.g., less than 1 or 0 dB) to reduce the delay of the digital audio signal Xd(t). If the estimated impulse signal Y(t) (e.g., its magnitude or its absolute value) is less than or equal to the threshold, the gain determination circuit 140 outputs a gain setting signal G(t) having a gain value (e.g., a value of 1 or 0 dB) indicating that the delayed digital audio signal Xd(t) will not be modified.

For example, using a shelf filter as the compensation filter 120, when the frequency of a component of the digital audio signal X(t) (e.g., represented by Fin) is less than or equal to a specific frequency value (e.g., 60 Hz), the gain determination circuit 140 outputs a gain value to prevent the excursion (or coil displacement) of the speaker 180 from exceeding a threshold.

By using the compensation filter 120, an example of an estimated stroke with an amplitude equal to the stroke threshold (Xmax) of 0.675 (mm) when Fin < 60 Hz will be compensated (e.g., amplified in this case when Fin < 60 Hz) to a compensated predicted stroke with an amplitude of 0.75 (mm/V).

Gain determination circuit 140 determines a gain value (e.g., represented by Gv) such that the compensated predicted stroke (amplitude: 0.75 (mm/V)) is lower than the stroke threshold (e.g., amplitude of 0.675 (mm)). For example, according to the formula representing that the output stroke amplitude (OUT_E) is equal to the input stroke amplitude (IN_E) multiplied by the gain value (Gv), i.e., OUT_E = IN_E * Gv, the gain value Gv is equal to 0.675/0.75 = 0.9, or 20 * log (0.9 / 1) = -0.915 dB. Therefore, the gain setting signal G(t) can be determined based on the gain value Gv.

Calculations show how this gain value addresses the issue of asymmetric excursion distortion. Since the actual excursion should be within an amplitude range of 0.675 (mm/V), the displacement amplitude corresponding to the delayed digital audio signal Xd(t) multiplied by the gain value Gv (e.g., 0.9) is ideally 0.6075 (mm/V). Consequently, even if the speaker's asymmetric excursion occurs below 50Hz, the actual negative period of the displacement will be -0.675 (mm/V), never exceeding the (negative) excursion threshold (i.e., -Xmax).

In an embodiment of another type of loudspeaker having a range response based on a peak response (e.g., the second type illustrated in FIG. 3A ), a peak filter is used as a compensation filter, and circuit 100 having compensation filter 120 and gain determination circuit 140 can be configured to similarly find a gain value Gv with appropriate judgment criteria, for example, for range estimation compensation of loudspeaker 180 when the frequency is within a specific frequency band from a first frequency to a second frequency (e.g., from 400 Hz to 1400 Hz). In further embodiments, circuit 100 can be configured or implemented to perform audio signal processing with range estimation compensation for other types of speakers (e.g., speakers having a range response based on a combination of the first and second types of speakers, etc.) by combining the above examples of range estimation compensation for the first and second types of speakers as appropriate.

Furthermore, in some embodiments, gain determination circuit 140 in FIG. 1 can be implemented using dynamic range control and peak and hold time control. FIG. 6 illustrates an embodiment of a circuit 100A for audio signal processing with stroke estimation compensation based on the architecture of FIG. In FIG. 6 , gain determination circuit 140A of circuit 100A for audio signal processing with stroke estimation compensation, compared to circuit 100 , includes a peak detector 141 and a gain calculation circuit 145 . Peak detector 141 is configured to output a peak-indicating stroke signal (e.g., represented by Yp(t)) based on the estimated stroke signal Y(t). The gain calculation circuit 145 is used to generate a gain setting signal G(t) based on the peak indicating stroke signal Yp(t) and a threshold (such as a stroke threshold or a maximum stroke limit).

Figure 7 illustrates an example of peak detection and gain calculation. For ease of illustration, the upper portion of Figure 7 shows the digital audio signal X(t) as a waveform. The middle portion of Figure 7 shows an estimated range signal Y(t) (mm/V) (e.g., the absolute range value) associated with the digital audio signal X(t) with range estimation compensation, along with a peak-indicating range signal Yp(t). The horizontal line represents a threshold, such as the speaker's range threshold (Xmax). The peak-indicating range signal Yp(t) can be obtained, for example, from the estimated range signal Y(t) using peak detection or averaging. For example, a peak detection method retains the maximum value of the estimated impulse recorded over time within a time window; another peak detection method obtains the envelope of the estimated impulse by averaging the time window over time. The lower portion of Figure 7 shows the gain setting signal G(t), which indicates the gain value used to adjust the associated digital audio signal X(t) (actually, the delayed digital audio signal Xd(t)). For example, when the peak-indicating impulse signal Yp(t) exceeds the impulse threshold Xmax (e.g., 0.2 mm), the gain setting signal G(t) has a negative value (or a value less than 1) in dB, for example, using the equation G(t) = Xmax/Yp(t). When the peak indication stroke signal Yp(t) is lower than or less than the stroke threshold Xmax (e.g., 0.2mm), the corresponding gain value of the gain setting signal G(t) is 0dB (or equal to 1).

FIG8 illustrates an example of dynamic range compression using a peak detection method. In FIG8 , filter 800 represents an alpha filter structure for dynamic range compression, where the coefficient alpha α can be set to a value of "a," and the parameter omega ω can be set to 1 minus the value of "a." Based on the known application of alpha filter structures for smoothing signals or varying data, peak detector 141 can be implemented using filter 800 for peak hold time control, thereby smoothly outputting a peak-indicating impulse signal Yp(t) based on the estimated impulse signal Y(t) (e.g., as shown in the middle portion of FIG7 ).

In addition, FIG9 shows a flow chart of a method for audio signal processing with impulse estimation compensation according to an embodiment of the present disclosure. The method includes the following steps S10 to S50.

In step S10, the digital audio signal (e.g., X(t)) is delayed to output a delayed digital audio signal (e.g., Xd(t)).

In step S20, a compensation filter (e.g., 120 in FIG. 1 or related examples) generates a compensated digital audio signal (e.g., Xc(t)) based on a digital audio signal (e.g., X(t)) for use in compensating for the range estimation of a type of speaker (e.g., the first or second speaker type in the above examples).

In step S30, an estimated impulse signal (e.g., Y(t)) of the speaker type is determined based on the compensated digital audio signal.

In step S40, a gain setting signal (e.g., G(t)) is generated based on the estimated stroke signal and a threshold (e.g., stroke threshold (Xmax)).

In step S50, an adjusted digital audio signal (e.g., OUT(t)) is generated based on the gain setting signal (e.g., G(t)) and the delayed digital audio signal (e.g., Xd(t)).

In some embodiments of the method, step S40 may include: outputting a peak indicating stroke signal based on the estimated stroke signal; and generating a gain setting signal based on the peak indicating stroke signal.

In some embodiments of the method of FIG. 9 , any one of the embodiments of the circuit 100 (or 100A) described above may also be applied, or two or more of the embodiments may be combined where appropriate.

In some embodiments of the present disclosure, the architecture of circuit 100 or 100A or the pulse estimation path (or both the pulse estimation path and the signal path) can be implemented by a programmable circuit, such as a digital signal processor (DSP), a field programming gate array (FPGA), or an application-specific integrated circuit (ASIC). In some embodiments of the present disclosure, the pulse estimation path (or both the pulse estimation path and the signal path) can be implemented by a dedicated circuit. Circuit 100 or 100A can also be implemented as an integrated circuit or within an integrated circuit, such as an audio signal processor or related audio chip.

The present disclosure further provides a computer program product comprising a plurality of instructions that enable a computing device to execute the method for audio signal processing with impulse estimation compensation as shown in FIG9 according to at least one embodiment described above. In one embodiment, a computer program product includes a storage medium, such as a non-transitory storage medium, storing computer-readable instructions, such as program code, wherein the instructions are executed on at least one computing device (e.g., an electronic device having computing or digital audio signal or data processing capabilities to facilitate audio reproduction, such as a mobile phone, tablet computer, laptop computer, desktop computer, multimedia player, digital audio player, or portable wireless or wired speaker), thereby causing the at least one computing device to execute a method according to at least one embodiment. The method is illustrated in FIG. 9 and is executed according to any one or any combination of the above embodiments. For example, the program code includes one or more programs or program modules for executing the method according to steps S10 to S50 in FIG. 9 in any appropriate order. When a computing device executes the program code, the computing device performs the method according to the embodiment shown in FIG. Examples of storage media include, but are not limited to, optical information storage media, magnetic information storage media, or memory (e.g., a memory card, firmware, ROM, or RAM). For example, the computing device includes communication circuitry, processing circuitry, and storage media. The processing circuitry is electrically coupled to the communication circuitry and storage media. In another example, the computing device may include a programmable circuit, such as a digital signal processor (DSP), a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc., which may be configured to execute the method based on FIG. 9 according to a program code.

Although the present disclosure has been described through specific embodiments, those skilled in the art may make various modifications and variations thereto without departing from the scope and spirit of the present disclosure as set forth in the patent application.

100: Circuit 110: Delay circuit 120: Compensation filter 130: Threshold estimator 140: Gain determination circuit 150: Gain adjustment circuit 160: Digital-to-analog converter 170: Amplifier 180: Speaker G(t): Gain setting signal OUT(t): Adjusted digital audio signal X(t): Digital audio signal Xc(t): Compensated digital audio signal Xd(t): Delayed digital audio signal Y(t): Estimated threshold signal

Claims (19)

A circuit for processing audio signals with impulse estimation compensation includes: A delay circuit for delaying a digital audio signal to output a delayed digital audio signal; A compensation filter for generating a compensated digital audio signal based on the digital audio signal for use in impulse estimation compensation for a speaker type, the speaker type being a first type speaker with an impulse response that is a shelf response or a second type speaker with an impulse response that is a peak response; An impulse estimator for determining an estimated impulse signal of the speaker type based on the compensated digital audio signal; a gain determination circuit for generating a gain setting signal based on the estimated impulse signal and a threshold; and a gain adjustment circuit for generating an adjusted digital audio signal based on the gain setting signal and the delayed digital audio signal. The audio signal processing circuit of claim 1, wherein the compensation filter has a filter response for compensating for the impulse estimation of the speaker type and is configured to generate the compensated digital audio signal based on the digital audio signal and the filter response. The audio signal processing circuit of claim 2, wherein the filter response of the compensation filter is associated with a specific frequency band for impulse estimation compensation of the speaker type, and the compensation filter is capable of amplifying the component of the digital audio signal for the specific frequency band to generate the compensated digital audio signal. The audio signal processing circuit of claim 2, wherein the filter response of the compensation filter is associated with a specific frequency band in which the speaker type can operate to obtain a maximum range. The audio signal processing circuit of claim 4, wherein the compensation filter is capable of amplifying a component of the digital audio signal for the specific frequency band according to the digital audio signal and the filter response to generate the compensated digital audio signal. The audio signal processing circuit of claim 1, wherein the compensation filter is based on an infinite impulse response (IIR) filter or a finite impulse response (FIR) filter. The audio signal processing circuit of claim 1, wherein the compensation filter is based on a shelf filter for performing impulse estimation compensation for a speaker type having a shelf response. The audio signal processing circuit of claim 1, wherein the compensation filter is based on a peak filter for performing impulse estimation compensation for a speaker type having a peak response. The audio signal processing circuit of claim 1, wherein the gain determination circuit comprises: a peak detector for outputting a peak-indicating range signal based on the estimated range signal; and a gain calculation circuit for generating the gain setting signal based on the peak-indicating range signal and the threshold. A method for processing an audio signal with range estimation compensation, the method comprising: delaying a digital audio signal to output a delayed digital audio signal; generating a compensated digital audio signal based on the digital audio signal through a compensation filter for use in range estimation compensation for a speaker type, the speaker type being a first type speaker with a range response of a shelf response or a second type speaker with a range response of a peak response; determining an estimated range signal for the speaker type based on the compensated digital audio signal; Generating a gain setting signal based on the estimated impulse signal and a threshold; and Generating an adjusted digital audio signal based on the gain setting signal and the delayed digital audio signal. The method of claim 10, wherein the compensation filter has a filter response for compensating for impulse estimation of the speaker type and is configured to generate the compensated digital audio signal based on the digital audio signal and the filter response. The method of claim 11, wherein the filter response of the compensation filter is associated with a specific frequency band for range estimation compensation for the speaker type, and the compensation filter is capable of amplifying components of the digital audio signal for the specific frequency band to generate the compensated digital audio signal. The method of claim 11, wherein the filter response of the compensation filter is associated with a specific frequency band in which a speaker of the speaker type can operate to obtain a maximum range. The method of claim 13, wherein the compensation filter is capable of amplifying a component of the digital audio signal for the specific frequency band based on the digital audio signal and the filter response to generate the compensated digital audio signal. The method of claim 10, wherein the compensation filter is based on an infinite impulse response (IIR) filter or a finite impulse response (FIR) filter. The method of claim 10, wherein the compensation filter is based on a shelf filter for performing impulse estimation compensation for a loudspeaker type having a shelf response. The method of claim 10, wherein the compensation filter is based on a peak filter for compensating for range estimation for a speaker type having a peak response. The method of claim 10, wherein generating the gain setting signal based on the estimated stroke signal and the threshold comprises: outputting a peak-indicating stroke signal based on the estimated stroke signal; and generating the gain setting signal based on the peak-indicating stroke signal and the threshold. A non-transitory storage medium stores a plurality of instructions, wherein the instructions enable a computing device to execute the method described in any one of claims 10 to 18.