Signal processing method

Abstract

To provide a signal processing method that provides an adaptive tonality control, a program, and a digital signal processing apparatus.SOLUTION: A method includes: calculating a shape dependent filter magnitude response (30); calculating a level dependent filter magnitude response (20); calculating a result response based on the shape and level dependent filter magnitude responses (40); and calculating an output signal by convolving an input signal with the result response (50). The calculation of the shape dependent filter magnitude response comprises: calculating a magnitude spectrum of a signal (110); filtering the resulting magnitude spectrum by using a first smoothing filter (120); applying a predetermined weighting curve to the spectrum (130); detrending the weighted spectrum by fitting a curve to the weighted spectrum and subtracting the fitted curve from the weighted spectrum (140); and inverting the resulting spectrum to form the shape dependent filter magnitude response (150).SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] <Field of the Invention> The present invention relates to a signal processing method for processing an audio signal, in particular for adjusting the tone of a music signal. [Background technology]

[0002] Description of the Prior Art Adjusting the tone of a musical signal in a high-quality production is a demanding task that requires a specialized sound engineer. Tone adjustment is usually required to make the signal sound pleasant. The balance of overtone series partials is one of the key aspects that defines the timbre and tone of a sound. Several psychoacoustic metrics have been shown to correlate with specific relationships between the levels of the partials. Many musical instruments and singing vocals produce an overtone series. The shape of the overtone series, i.e. their relative levels, can be adjusted to obtain a more balanced and pleasant sound.

[0003] In addition to tone adjustment, dynamic range is also typically controlled. Both of these adjustments require manual effort and control.

[0004] In prior art configurations, processing requires manual effort and expertise, typically from a sound engineer given an input signal and an understanding of how that signal should sound to a listener, and the sound engineer needs to know how to adjust this particular signal to reach a desired result. A simpler way to process sound that requires less effort from a professional sound designer is needed. Summary of the Invention [Problem to be solved by the invention]

[0005] The invention is directed to a method according to the characterizing portion of the independent claims. The invention also relates to a digital signal processing device according to the characterizing portion of the independent device claim. The invention also relates to a computer program product according to the characterizing portion of the independent computer program product claim. [Means for solving the problem]

[0006] The present invention solves the problems of the prior art by providing a signal processing method for a digital signal processing device that provides adaptive tonality control, where a user can specify a desired weighting curve and the processing device calculates how to filter the input signal. In one embodiment of the present invention, this calculation is performed as follows:

[0007] In the method, a shape dependent filter magnitude response is calculated, a level dependent filter magnitude response is calculated, a result response is calculated based on the shape and level dependent filter magnitude responses, and an output signal is calculated by convolving the input signal with the result response. In particular, the calculation of the shape dependent filter magnitude response includes at least - calculating the amplitude spectrum of the signal; - filtering the resulting amplitude spectrum using a first smoothing filter; - applying a predefined weighting curve to the spectrum; - detrend the weighted spectrum by fitting a curve to the weighted spectrum and subtracting the fitted curve from the weighted spectrum; - inverting the resulting spectrum to form a shape-dependent filter magnitude response.

[0008] In a further embodiment of the invention, the method further comprises, after the step of low-pass filtering, a step of resampling the spectrum to reduce computational complexity.

[0009] In a further embodiment of the invention, the method further comprises the step of resampling the filter magnitude response after the inverting step.

[0010] In a further embodiment of the invention, the method further comprises the step of applying a phase transformation function to the filter magnitude response after the inverting step.

[0011] In a further embodiment of the present invention, the calculation of the level dependent filter magnitude response comprises at least - filtering the calculated amplitude spectrum of the signal using a second smoothing filter; - applying a second predetermined weighting curve to the spectrum; - forming a level-dependent filtered magnitude spectrum by applying a level limiting function to the smoothed spectrum.

[0012] In one embodiment of the present invention, the resultant response is calculated based on the level-dependent filter magnitude response and the shape-dependent filter magnitude response in a manner that changes dynamically such that the less one of the responses influences the resultant response at a point, the more the other one influences the resultant response at that point, and vice versa.

[0013] In a further embodiment of the invention, the result response comprises: the less the level-dependent filter amplitude response influences the result response at a point, the more the shape-dependent filter amplitude response influences the result response at that point; The more the level-dependent filter magnitude response influences the result at a point, the less the shape-dependent filter magnitude response is calculated to influence the result response at that point.

[0014] In yet another embodiment of the present invention, the result response comprises: the less the shape-dependent filter amplitude response influences the result response at a point, the more the level-dependent filter amplitude response influences the result response at that point; The more the shape-dependent filter magnitude response influences the response at a point, the less the level-dependent filter magnitude response is calculated to influence the resulting response at that point.

[0015] The above summary relates to only a few of the many embodiments of the invention disclosed herein and is not intended to limit the scope of the invention as set forth in the claims. These and other features of the invention are described in more detail below in the Detailed Description of the Invention and in conjunction with the following drawings. [Brief description of the drawings]

[0016] Various embodiments of the invention are described in detail below, by way of example only, with reference to the accompanying drawings, in which: [Figure 1] FIG. 1 illustrates a method according to one embodiment of the present invention. [Diagram 2] FIG. 2 shows an apparatus according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Detailed Description of Specific Embodiments The following embodiments are illustrative. Although the specification may refer to "a," "one," or "several" embodiments, this does not necessarily mean that such references each refer to the same embodiment or that a feature applies only to a single embodiment. Features of different embodiments can be combined to provide further embodiments.

[0018] In the following, the features of the present invention will be described using a simple example of a sound processing method in which various embodiments of the present invention can be implemented. Only elements relevant for explaining the embodiments will be described in detail. Details commonly known to those skilled in the art may not be specifically described herein.

[0019] Figure 1 shows a signal processing method according to an embodiment of the present invention. The method follows a basic structure in which a shape-dependent filter magnitude response is calculated (30), a level-dependent filter magnitude response is calculated (20), a result response is calculated based on the shape and level-dependent filter magnitude responses (40), and an output signal is calculated by convolving an input signal with the result response (50). The input signal is processed in blocks containing a predetermined number of signal samples. This basic structure shown in Figure 1 is repeated to form a successive signal analysis and processing process.

[0020] Generally, in signal processing methods using this basic structure, the aim of shape-dependent processing is to reduce the deviation of the shape of the spectrum of the signal from a relative target spectrum, regardless of the level of the signal. Correspondingly, level-dependent processing generally aims to adjust the level of the signal at various frequencies without reference to the overall shape of the signal spectrum. These two types of processing are then combined to form a resultant filter response for processing the actual signal.

[0021] According to this embodiment of the invention shown in FIG. 1, the calculation 30 of the shape-dependent filter magnitude response includes at least - calculating (110) the amplitude spectrum of the signal; - filtering (120) the resulting amplitude spectrum using a first smoothing filter; - applying (130) a first predefined weighting curve to the spectrum; - detrending (140) the weighted spectrum by fitting a curve to the weighted spectrum and subtracting the fitted curve from the weighted spectrum; - inverting (150) the resulting spectrum to form a shape-dependent filter magnitude response.

[0022] In step 130, applying a weighting curve to the spectrum can be performed using multiplication, for example by multiplying the spectrum with a predefined weighting curve. In step 120, the first smoothing filter can be, for example, a low pass filter.

[0023] In certain further embodiments of the present invention, the filtered spectrum is resampled after the filtering step 120 in order to reduce computational complexity. This resampling can be performed in various ways in different embodiments of the present invention. The resampling rate may be different in different implementations of the present invention and may depend on the specific needs of a particular implementation.

[0024] Furthermore, the resampling rate does not have to be constant across the spectrum. In an embodiment, the resampling rate can be different in different parts of the spectrum. For example, in a further embodiment, samples are added to the spectrum at low frequencies by interpolation, while samples are removed at high frequencies by decimation.

[0025] The step 140 of detrending the weighted spectrum can be performed in many different ways in different embodiments of the invention. For example, in one embodiment of the invention, the weighted spectrum is detrended by overpass filtering, which removes low frequency background from the spectrum while preserving details.

[0026] In a further embodiment, detrending is performed by fitting a curve to the weighted spectrum and subtracting the fitted curve from the weighted spectrum. The fitting of a curve to the weighted spectrum can be performed in different ways in different embodiments of the invention. For example, the fitting can be performed by weighted or non-weighted linear regression. The fitting can also be implemented using non-linear regression, for example using weighted or non-weighted polynomial regression, or by using neural networks or deep learning. The fitting of a curve can be performed by any fitting method known to a person skilled in the art in different embodiments of the invention.

[0027] In a further embodiment of the invention where the filtered spectrum is resampled after the filtering step 120, the inverted result is resampled after the inverting step in order to bring the result back to the same sample rate that the filtered spectrum had before the first resampling after the filtering step 120, in other words to undo the effect of the first resampling on the sampling scheme of the spectrum.

[0028] In a further embodiment of the invention, the method further comprises, after the inverting step, applying a phase transformation function to the filter magnitude response. Such an embodiment may use a phase transformation function that reduces filter pre-ringing (or artifacts appearing in the filtered signal) while retaining an approximate magnitude response. This step may be performed, for example, by applying a Hilbert transform to the filter magnitude response.

[0029] Although the Hilbert transform is a good choice for generating a minimum phase filter response, those skilled in the art are aware of several variants of such transforms that generate transforms that are close to a minimum phase filter response, i.e., that are near minimum if not absolutely minimum, and such variants are less expensive to compute. Thus, the present invention is not limited to the use of only the Hilbert transform, and various embodiments of the present invention may use other phase transformation functions known to those skilled in the art.

[0030] In a further embodiment of the present invention, the calculation 20 of the level dependent filter magnitude response comprises at least - a step 210 of filtering said calculated amplitude spectrum of the signal using a second smoothing filter; - a step 220 of applying a second predetermined weighting curve to said spectrum; - forming 230 a level-dependent filtered magnitude spectrum by applying a level limiting function to the smoothed spectrum.

[0031] In one embodiment of the invention, the second predefined weighting curve is at least partially different from the first predefined weighting curve, however, in a further embodiment of the invention, the first and second predefined weighting curves are the same.

[0032] In various embodiments of the present invention, the level limiting function can be implemented in a variety of ways. The level limiting function can be implemented, for example, using frequency dependent thresholds, knee functions, and time constants and program dependent functions. The level limiting function can be implemented, for example, as an arbitrary frequency dependent mapping function g that outputs a scaling factor for the amplitude (or magnitude) of each frequency as follows: y[f]=g(x[f],f) * x[f]

[0033] The instantaneous scaling coefficients may be further smoothed over time using any recursive smoothing function to avoid any abrupt changes in the frequency-dependent scaling coefficients. Those skilled in the art will know many different ways to construct level limiting functions, and therefore such functions will not be described in further detail herein.

[0034] In a further embodiment of the invention, the result response is calculated such that the less the level-dependent filter magnitude response influences the result response at a point, the more the shape-dependent filter magnitude response influences the result response at that point, and the more the level-dependent filter magnitude response influences the result response at a point, the less the shape-dependent filter magnitude response influences the result response at that point. Such a calculation can be implemented in many different ways. Examples of such calculations are given in the following paragraphs.

[0035] In further embodiments of the invention, the method uses parameters that allow a user of the device performing the method to adjust how much and at what level any of these responses dominate the resulting response, in such embodiments the method effectively provides a dynamic selection between a level-dependent filter magnitude response and a shape-dependent filter magnitude response.

[0036] In one embodiment of the present invention, the result response is calculated using the following formula: |H|=p * A^(B^q) * B (1) Where: |H| is the outcome response, p is a predetermined constant gain compensation coefficient greater than zero; q is a predetermined dynamic transition steepness coefficient between 0 and 1; A is the shape dependent filter magnitude response; B is the level dependent filter magnitude response.

[0037] In this embodiment, the parameter q is a user adjustable parameter, meaning that it affects the relative strengths of A and B with respect to the final filtering.

[0038] Equation (1) can be expressed in a variety of ways, including: |H|=p * exp(log(A) * exp(q * log(B))+log(B)) (2) =p * exp(log(A))^exp(q * log(B) * exp(log(B)) (3) =p * A^(B^q) * B

[0039] Which of these expressions, or other expressions derived therefrom, are used in a particular implementation of an embodiment of the present invention will depend on the practical needs of such implementation and which computational methods are deemed most suitable for that implementation.

[0040] In a further embodiment of the invention, the blending of the shape-dependent filter magnitude response A and the level-dependent filter magnitude response B is performed using linear blending, for example using one of the following equations: |H|=p(B * A+(1-B))B (4) =p(B^2 * A+BB^2) (5) =p(B^2(A-1)+B) (6)

[0041] These different ways of combining A and B to calculate the result response are merely examples, and the present invention is not limited to using only these examples. Those skilled in the art may generate variations of these examples, whereby further variations and examples are not further discussed herein.

[0042] In a further embodiment of the present invention, the first smoothing filter is a low-pass filter and the second smoothing filter is a low-pass filter, the cut-off point of the second smoothing filter being lower than the cut-off point of the first smoothing filter.

[0043] In such an embodiment, the level-dependent filter magnitude response B is in part the result of low-pass filtering having a lower cutoff point than the shape-dependent filter magnitude response A. This has the effect that for high-level signal segments the method optimizes for time domain selectivity, and for low-level signal segments the method optimizes for frequency domain selectivity. However, this particular choice of filters is only one example, and the filters may be arranged in different ways in different embodiments of the invention.

[0044] The present invention can be implemented in many different ways. The method of the present invention can be implemented in a dedicated signal processing device, as part of a sound processing device such as a digital mixer unit or another device, or using a general-purpose computing device configured to process, for example, digital sound signals. Furthermore, the method of the present invention can be implemented as computer software stored on a computer readable medium or as downloadable software made available to customers, for example, via a computer network.

[0045] The present invention may also be implemented as a computer program product comprising instructions which, when executed by a computing device, cause the computing device to perform the steps of the methods of the present invention.

[0046] 2 shows a digital signal processing device 300. This device can be, for example, a digital mixer unit or part of a dedicated sound processing device, or for example a general purpose computer configured to process sound.

[0047] The apparatus 300 comprises a processor 310 operatively connected to an input 302 and an output 304. The apparatus also comprises a memory element 320 that stores instructions 322, 324 to be executed by the processor. The instructions, when executed by the processor 310, cause the processor to perform the methods of the present invention as described herein and defined in the claims.

[0048] In further embodiments of the present invention, the device 300 may also include other functional elements known to those skilled in the art, such as analog-to-digital (AD) and digital-to-analog (DA) converters to enable the device to input and output analog signals. The device may also include multiple processors. The processors may be dedicated signal processing processors, general purpose programmable microprocessors, or processors implemented, for example, using an FPGA (Field Programmable Gate Array).

[0049] The present invention may also be implemented as a software product as instructions stored on a storage medium such that, when executed by a processor, the instructions cause the processor to perform the inventive method as recited in the claims. The storage medium may be a volatile memory device or a semiconductor-based device such as, for example, a non-volatile flash memory device. Some further examples of such machine-readable storage media include RAM, ROM, read-only compact disc (CD-ROM), recordable compact disc (CD-R), rewritable compact disc (CD-RW), read-only digital versatile disc (e.g., DVD-ROM, dual-layer DVD-ROM), various recordable / rewritable DVDs (e.g., DVD-RAM, DVD-RW, DVD+RW, etc.), flash memory (e.g., SD card, mini SD card, micro SD card, etc.), magnetic and / or solid-state hard drives, read-only and recordable Blu-ray discs, ultra-density optical discs, and any other optical or magnetic media. The present invention may also be implemented as a downloadable software product that can be installed in an audio processing device.

[0050] Most of the methods shown in Fig. 1 concern an analysis of an input signal, followed in step 50 by actual processing of the input signal blocks. Whether the processing of the signal blocks in step 50 concerns the same signal blocks for which an analysis was performed in a previous step varies in different embodiments of the invention. Furthermore, in some embodiments of the invention, the size of the input signal blocks used for analysis and the size of the input signal blocks used for processing may differ from each other.

[0051] In one embodiment of the present invention, the input signal block of N samples processed in step 50 is the same signal block on which the previous analysis step was based. Such an embodiment has the advantage of being computationally efficient and having little adaptation delay. However, such an embodiment has some latency, since the input block needs to be analyzed before it can be processed.

[0052] However, in further embodiments, the processing step may use the analysis results of a previous analysis round, for example an input signal block may be processed using a result response calculated from a preceding input signal block.

[0053] In further embodiments of the invention, the analysis steps 20, 30, 40 are performed using input signal blocks of N samples, and the resulting response calculated using such blocks is used to convolve an input signal block of M samples. For example, in embodiments where M is smaller than N, e.g., N is 1024 and M is 256, the processing latency is relatively small, but the processing load and adaptation delay from the analysis step is higher than in embodiments where the analysis is based on signal blocks of M samples.

[0054] In a further embodiment of the invention, the processing and analysis is based on input signal blocks of N samples formed every N / M samples from the input signal sample stream, such that the actual processing step 50 of the input signal blocks to form output signal blocks is based on the analysis of the previous input signal block. In such an embodiment, the main penalty of such an arrangement is a small adaptivity delay of N / M samples, since successive signal sample blocks partially overlap with some samples in the input sample stream.

[0055] Specific Advantages of the Invention The present invention has many advantages, many of which have already been described herein. Furthermore, the present invention helps to automate parts of the audio signal conditioning process, which saves time for the audio engineer. Furthermore, the present invention allows amateur audio engineers to perform audio processing that would not be possible without the method of the present invention. The present invention provides an automated method for handling both tone control and dynamic range control in a single process, which saves work and time for the audio engineer and allows both controls to be performed in a more optimal manner.

[0056] Further specific embodiments of the present invention Several embodiments of the present invention are described below.

[0057] According to a first aspect of the invention, there is provided a method for processing an audio signal in a digital signal processing device, in which a shape-dependent filter magnitude response is calculated, a level-dependent filter magnitude response is calculated, a resultant response is calculated based on the shape and the level-dependent filter magnitude response, and an output signal is calculated by convolving the input signal with the resultant response. According to a first embodiment of this first aspect of the invention, the calculation of the shape-dependent filter magnitude response comprises at least the steps of calculating (110) an amplitude spectrum of the signal, filtering (120) the resulting amplitude spectrum using a first smoothing filter, applying (130) a first predefined weighting curve to the spectrum, detrending (140) the weighted spectrum, and inverting (150) the resulting spectrum to form a shape-dependent filter magnitude response.

[0058] According to a second embodiment of this first aspect of the invention, the method further comprises, after said filtering step, a step of performing a first resampling of the amplitude spectrum to reduce computational complexity.

[0059] According to a third embodiment of this first aspect of the invention, the method further comprises, after said inverting step, a step of resampling said filter magnitude response to return to said first resampling.

[0060] According to a fourth embodiment of this first aspect of the invention, the method further comprises, after said inverting step, applying a phase transformation function to said filter magnitude response.

[0061] According to a fifth embodiment of this first aspect of the invention, the step of calculating (20) a level-dependent filter magnitude response comprises at least the steps of filtering (210) said calculated magnitude spectrum of the signal using a second smoothing filter, applying (220) a second predetermined weighting curve to the spectrum, and forming (230) a level-dependent filter magnitude spectrum by applying a level limiting function to the smoothed spectrum.

[0062] According to a sixth embodiment of this first aspect of the invention, the result response is calculated such that the less the level-dependent filter magnitude response influences the result response at a point, the more the shape-dependent filter magnitude response influences the result response at that point, and such that the more the level-dependent filter magnitude response influences the result response at a point, the less the shape-dependent filter magnitude response influences the result response at that point.

[0063] According to a seventh embodiment of this first aspect of the invention the result response comprises: |H|=p * exp(log(A) * exp(q * log(B))+log(B)) =p * exp(log(A))^exp(q * log(B) * exp(log(B)) =p * A^(B^q) * B It is calculated by Where: |H| is the outcome response, p is a predetermined constant gain compensation factor; q is a predetermined dynamic transition steepness factor; A is the shape dependent filter magnitude response; B is the level dependent filter magnitude response.

[0064] According to an eighth embodiment of this first aspect of the present invention, the first smoothing filter is a low-pass filter and the second smoothing filter is a low-pass filter, and a cut-off point of the second smoothing filter is lower than the cut-off point of the first smoothing filter.

[0065] According to a second aspect of the invention there is provided a computer program product comprising instructions which, when executed by a computing device, cause the computing device to perform the steps of the method of the first aspect of the invention.

[0066] According to a third aspect of the invention, there is provided an apparatus for digital signal processing. The apparatus comprises one or more processors (310) and one or more computer readable storage media (320). According to a first embodiment of this third aspect of the invention, the storage media stores instructions (322, 324) which, when executed by the one or more processors, cause the one or more processors to perform the steps of the method of the first aspect of the invention.

[0067] <Specific Further Observations> In view of the above description, it will be apparent to those skilled in the art that various modifications may be made within the scope of the present invention. While preferred embodiments of the present invention have been described in detail, it will be apparent that numerous modifications and variations thereof are possible, all of which are within the true spirit and scope of the invention.

[0068] It is to be understood that the disclosed embodiments of the invention are not limited to the particular structures, process steps, or materials disclosed herein, but extend to equivalents thereof that would be recognized by one of ordinary skill in the art. It is also to be understood that the terminology used herein is used for the purpose of describing particular embodiments only, and is not intended to be limiting.

[0069] Throughout this specification, a reference to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present embodiment. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment.

[0070] As used herein, a plurality of items, structural elements, compositional elements, and / or materials may be presented in a common list for convenience. However, these lists should be construed as if each member of the list were individually identified as a separate and unique member. Thus, the individual members of such lists should not be construed as de facto equivalents of other members of the same list, but should be construed solely based on their presentation in a common group, without any indication to the contrary. In addition, various embodiments and examples of the invention may be referred to herein along with alternatives for the various components thereof. It is understood that such embodiments, examples, and alternatives should not be construed as de facto equivalents of each other, but should be considered as separate, autonomous representations of the invention.

[0071] Furthermore, the described functions, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the foregoing description, numerous specific details, such as examples of lengths, widths, shapes, etc., are provided to provide a thorough understanding of embodiments of the present invention. However, one of ordinary skill in the art will recognize that the present invention may be practiced without one or more of the specific details, or with other methods, components, materials, etc. In other examples, well-known structures, materials, or operations are not shown or described in detail to avoid obscuring aspects of the present invention.

[0072] While the foregoing examples are illustrative of the principles of the present invention in one or more specific applications, it will be apparent to those skilled in the art that numerous modifications in the embodiments, uses, and details can be made without the exercise of inventive faculty and without departing from the principles and concepts of the present invention. Accordingly, the present invention is not intended to be limited except as by the claims set forth below.

Claims

1. A method for processing an audio signal in a digital signal processing device, comprising: A shape-dependent filter magnitude response is calculated (30); A level-dependent filter magnitude response is calculated (20); A resultant response is calculated (40) based on the shape and level dependent filter magnitude response; An output signal is calculated (50) by convolving an input signal with the resulting response, The calculation of the shape dependent filter magnitude response comprises at least Calculating (110) the amplitude spectrum of the signal; filtering (120) the resulting amplitude spectrum with a first smoothing filter; applying (130) a first predetermined weighting curve to the spectrum; detrending (140) the weighted spectrum by fitting a curve to the weighted spectrum and subtracting the fitted curve from the weighted spectrum; and inverting (150) the resulting spectrum to form the shape-dependent filter magnitude response.

2. The method further comprises, after the filtering step:

2. The method of claim 1, further comprising the step of performing a first resampling of the amplitude spectrum to reduce computational complexity.

3. The method further comprises, after the inverting step:

3. The method of claim 2 further comprising the step of resampling the filter magnitude response to return to the first resampling.

4. The method further comprises, after the inverting step:

2. The method of claim 1 further comprising the step of applying a phase transformation function to the filter magnitude response.

5. The calculation (20) of the level-dependent filter magnitude response comprises at least filtering (210) the calculated amplitude spectrum of the signal with a second smoothing filter; applying (220) a second predetermined weighting curve to the spectrum; and forming (230) a level-dependent filtered magnitude spectrum by applying a level limiting function to the smoothed spectrum.

6. 2. The method of claim 1, wherein the result response is calculated according to the following formula: |H|=p ＊ exp(log(A) ＊ exp(q ＊ log(B))+log(B)) =p ＊ exp(log(A))^exp(q ＊ log(B)) ＊ exp(log(B)) =p ＊ A^(B^q) ＊ B Where: |H| is the outcome response, p is a predetermined constant gain compensation factor; q is a predetermined dynamic transition steepness factor; A is the shape-dependent filter magnitude response; B is the level dependent filter magnitude response.

7. the first smoothing filter is a low-pass filter, the second smoothing filter is a low-pass filter; 6. The method of claim 5, wherein the cutoff point of the second smoothing filter is lower than the cutoff point of the first smoothing filter.

8. 13. A computer program product comprising instructions that, when said program is executed by a computing device, cause said computing device to perform the steps of the method of claim 1.

9. A digital signal processing apparatus (300) having one or more processors (310) and one or more computer readable storage media (320), The digital signal processing apparatus, characterized in that the storage medium stores instructions (322, 324) that, when executed by the one or more processors, cause the one or more processors to perform the steps of the method of claim 1.

Images