Method and system for naturally sounding audio equalization

Abstract

A method for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment is provided. The method comprises applying one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal. To obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the method comprises: Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer.

Description

[0001]Method and System for Naturally Sounding Audio Equalization Description: of the Invention The present invention relates to a method and a system for audio signal processing, in particular to a method and a system for audio equalization, and, more particularly, to a method and a system for naturally sounding audio equalization. Equalization is the process of adjusting the frequency response of audio signals. In the equalization process, unpleasant frequencies in the audio signal are attenuated, while other frequencies in the audio signal that highlight desired characteristics of it are accentuated. This influences the perceived sound quality of an audio signal from a listener’s perspective. Equalization is often performed with regards to playback systems that consumers listen to audio content on. Every playback system, based on its physical and electronic properties, modifies an audio signal before it reaches the listener’s ears. It is therefore desirable to equalize the source audio signal in a way that keeps its frequency response within desired bounds when played back on certain playback systems. Equalization has historically been performed by human engineers. Based on their experience, human engineers can decide which frequencies enhance the perceived quality of an audio signal and highlight desired parts of its content to the listener, and which frequencies compromise its perceived quality. Engineers typically use equalizers (EQs), and parametric equalizers with frequency, gain and Q factor parameters in particular, to perform equalization. A parametric equalizer (EQ) may, e.g., implement a filter function, e.g., a second-order filter function. Usually, a parametric equalizer exhibits a center frequency, a gain (or level) and a Q factor. Q may, e.g., be considered to determine the sharpness of the bandwidth, and determines how much those frequencies are boosted or cut relative to frequencies much above or below the center frequency. The Q factor may, e.g., be defined as the ratio of the center frequency to the bandwidth. E.g., with a given center frequency, the bandwidth is inversely proportional to Q. In other words, by increasing the Q factor, the bandwidth of the equalizer is narrowed. MCH2301-D-2024098091.DOCX A common difficulty in designing automatic equalization systems is the determination of equalization settings that attenuate or accentuate frequencies as described earlier while preserving fundamental characteristics of the original audio signal and without making the resulting audio signal sound artificial to the listener. The presented approach introduces a system that finds natural-sounding equalization adjustments for an audio signal given a loosely-defined target frequency distribution. Some automatic EQ algorithms require user input for optimal settings. State of the art approaches often work with frequency adjustments in the FFT-domain, or do not consider phase information of the signal, which can sound unnatural. In prior art technologies, EQ adjustments are often conducted in the frequency domain by altering FFT frequency bins or similar, and user input is required to determine the ‘aggressiveness’ of the EQ adjustment. It would however be appreciated, if improved concepts for automatic audio equalization would be provided. A method for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment is provided. The method comprises applying one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal. To obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the method comprises: - Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. - Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. MCH2301-D-2024098091.DOCX - Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. - Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer. Moreover, a method for automatic audio equalization of an audio input signal to obtain an equalized audio signal according to an embodiment is provided. The method comprises: - Determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer. And: - Applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal. Determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response. Furthermore, a system for automatic audio equalization of an audio input signal to obtain an equalized audio signal according to an embodiment is provided. The system is configured for: Determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer, and applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal. Determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response. Moreover, a computer program for implementing the above described method, when the computer program is executed on a computer or signal processor, according to an embodiment is provided. MCH2301-D-2024098091.DOCX Furthermore, a non-transitory computer-readable medium comprising a computer program for implementing the above described method, when the computer program is executed on a computer or signal processor, according to an embodiment is provided. Moreover, a digital storage medium comprising a computer program for implementing the above described method, when the computer program is executed on a computer or signal processor, according to an embodiment is provided. Furthermore, a data signal comprising computer-readable instructions for implementing the above described method, when the computer-readable instructions are executed on a computer or signal processor, according to an embodiment is provided. Moreover, a non-transitory computer-readable medium comprising a data signal comprising computer-readable instructions for implementing the above described method, when the computer-readable instructions are executed on a computer or signal processor, according to an embodiment is provided. Furthermore, a digital storage medium comprising a data signal comprising computer- readable instructions for implementing the above described method, when the computer- readable instructions are executed on a computer or signal processor, according to an embodiment is provided. Moreover, an equalized audio signal being obtained by executing the above described method, according to an embodiment is provided. Furthermore, a non-transitory computer-readable medium comprising an equalized audio signal being obtained by executing the above described method, according to an embodiment is provided. Moreover, a digital storage medium comprising an equalized audio signal being obtained by executing the above described method, according to an embodiment is provided. Furthermore, a system for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment is provided. The system is configured for applying one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal. To obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the system is configured for: Determining an equalizer setting for each of the plurality of MCH2301-D-2024098091.DOCX parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer. Moreover, a system for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment is provided. The system comprises a configurator and a processor. The processor is configured for applying one of a plurality of parametric audio equalizers on the audio input signal to obtain an audio output signal. To obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the configurator is configured for: Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer. According to some embodiments, the above-described systems may, e.g., be configured to conduct one of the above-described methods described below. Before embodiments of the present invention are described in detail using the accompanying figures, it is to be pointed out that the same or functionally equal elements are given the same reference numbers in the figures and that a repeated description for elements provided with the same reference numbers is omitted. Hence, descriptions provided for elements having the same reference numbers are mutually exchangeable. Brief Description of the Drawings: MCH2301-D-2024098091.DOCX Fig.1 illustrates a method for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment. Fig.2 illustrates a particular method for automatic audio equalization of an audio input signal according to an embodiment. Fig.3 illustrates an audio input signal that is received and a desired target frequency response. Fig.4 illustrates a logarithmic-domain frequency response of the audio input signal, which is obtained by transforming the audio input signal from a time domain to a frequency domain. Fig.5 illustrates a logarithmic-domain positive frequency response difference signal, which is obtained by subtracting the target frequency response from the input signal frequency response of the audio input signal and by depicting the positive values of the frequency response difference signal which results from the subtraction. Fig.6 illustrates a general equalized model of an embodiment and a determination of an allowed gain range of a parametric audio equalizer according to an embodiment. Fig.6a illustrates four logarithmic-domain frequency responses of four equalizer settings according to an embodiment. Fig.6b depicts a smoothed frequency response for selected Q factors according to an embodiment. Fig.6c illustrates the settings of the plurality of linear-domain convolution kernels according to an embodiment. Fig.6d illustrates the logarithmic-domain smoothed frequency responses of the embodiment of Fig.6b in more detailed representations. Fig.6e illustrates a representation of a discrete lookup table resulting from the method of Fig.6, in which the adjustment scope / the allowed gain range, for each Q factor and frequency setting is stored. MCH2301-D-2024098091.DOCX Fig.6f illustrates a logarithmic-domain frequency response of an equalizer setting according to an embodiment. Fig.6g illustrates all four logarithmic-domain frequency responses of all four equalizer settings of Fig. 6a according to an embodiment in a better resolution. Fig.7 illustrates an equalizer result evaluation according to an embodiment. Fig.8 illustrates a method according to particular embodiment, which employs the steps described with respect to the illustrations of Fig.2 to Fig.7. Fig.9 illustrates a system for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment. Detailed Description of the Invention: Fig. 1 illustrates a method for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment. The method comprises applying (120) one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal. To obtain (130) said one the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the method comprises: - Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. - Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. - Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. MCH2301-D-2024098091.DOCX - Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer. Said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal after the equalizer setting for each of the plurality of parametric audio equalizers has been determined may, e.g., be the parametric audio equalizer having a best quality among the plurality of audio equalizers. When a plurality of iterations are conducted for determining the equalizer setting for each of the plurality of parametric audio equalizers, the parametric audio equalizer having a best quality among the plurality of audio equalizers after the last iteration may, e.g., be said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal. According to an embodiment, determining the equalizer setting of each of the plurality of parametric audio equalizers may, e.g., comprise determining an allowed maximum value or an allowed minimum value or an allowed range of at least one parameter of the two or more parameters of the equalizer setting. The parametric audio equalizers may, e.g., be evaluated in parallel or quasi-parallel. In an embodiment, determining the allowed maximum value or the allowed minimum value or the allowed range of said at least one parameter of the two or more parameters of said equalizer setting may, e.g., be conducted depending on a difference between the frequency response of the audio input signal and the target frequency response. According to an embodiment, determining the allowed maximum value or the allowed minimum value or the allowed range of said at least one parameter of the two or more parameters of said equalizer setting may, e.g., be conducted depending on one or more other parameters of said equalizer setting. In an embodiment, the equalizer setting of each parametric audio equalizer comprises a frequency, a Q factor and a gain. MCH2301-D-2024098091.DOCX The method may, e.g., comprise determining the allowed maximum value or the allowed minimum value or the allowed range of the gain of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the Q-factor of the parametric setting. Or, the method may, e.g., comprise determining the allowed maximum value or the allowed minimum value or the allowed range of the frequency of the equalizer setting of each of the plurality of parametric audio equalizers depending on the gain and the Q- factor of the parametric setting. Or, the method may, e.g., comprise determining the allowed maximum value or the allowed minimum value or the allowed range of the Q-factor of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the gain of the parametric setting. According to an embodiment, the method may, e.g., comprise determining the allowed maximum value or the allowed minimum value or the allowed range of the gain of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the Q-factor of the parametric setting by: - Determining a difference signal depending on the frequency response of the audio input signal and the target frequency response. - Determining a convolution kernel depending on the frequency and the Q-factor of the equalizer setting. - Conducting a convolution of the difference signal or of a portion of the difference signal and the convolution kernel to obtain a convoluted signal. And: - Determining the allowed maximum value or the allowed minimum value of the gain of the equalizer setting depending on the convoluted signal. In an embodiment, the allowed maximum value or the allowed minimum value may, e.g.,, depend on a value of the convoluted signal at the frequency of the equalizer setting. According to an embodiment, determining the convolution kernel may, e.g., be conducted by MCH2301-D-2024098091.DOCX - Generating a frequency response of the equalizer setting in a logarithmic domain depending on the frequency and the Q-factor of the equalizer setting. - Transforming or resampling the frequency response of the equalizer setting from the logarithmic domain to a linear domain so that the frequency response exhibits a linear x-scale to obtain the convolution kernel, and normalizing the convolution kernel, so that the sum of all values of the convolution kernel is 1. In an embodiment, the method may, e.g., comprise determining the difference signal so that the difference signal is represented in a logarithmic domain and indicates at least all positive values of a frequency response of the audio input signal and the target frequency response, and transforming or resampling the difference signal from the logarithmic domain to a linear domain so that the difference signal exhibits a linear x-scale. According to an embodiment, determining quality information for each parametric audio equalizer of the plurality of parametric audio equalizers may, e.g., be conducted by applying said parametric audio equalizer on the audio input signal to obtain an equalized audio signal, and by determining the quality information using the equalized audio signal. In an embodiment, determining quality information for said parametric audio equalizer may, e.g., be conducted by: - transforming the equalized audio signal to the frequency domain to obtain a frequency-domain equalized audio signal: - subtracting the target frequency response from absolute values of the frequency- domain equalized audio signal to obtain an adjusted frequency-domain equalized audio signal, and - calculating a standard deviation of the adjusted frequency-domain equalized audio signal to obtain the quality information. According to an embodiment, transforming the equalized audio signal to the frequency domain is conducted using a Discrete Cosine Transform or a Discrete Sine Transform or a Discrete Fourier Transform. In an embodiment, the following steps of determining, selecting and updating: MCH2301-D-2024098091.DOCX - determining quality information for each of the plurality of parametric audio equalizers, - selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer, - updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer, may, e.g., be repeatedly executed one or more times to obtain resulting equalizer settings of the plurality of parametric audio equalizers, and wherein said one of the plurality of parametric audio equalizers may, e.g., be applied with its resulting equalizer settings on the audio input signal to obtain the audio output signal. According to an embodiment, updating the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer, may, e.g., be conducted further depending on a stored equalizer setting of one of the plurality of parametric audio equalizers of a previous iteration preceding a current iteration of executing the steps of determining, selecting and updating, which exhibits a best quality value among the quality values of the equalizer setting of the plurality of parametric audio equalizers of all iterations of executing the steps of determining, selecting and updating. In an embodiment, updating the equalizer setting of each of the plurality of parametric audio equalizers may, e.g., be conducted depending on the equalizer setting of the selected audio equalizer and further depending on noise or on a random process. According to an embodiment, the steps of determining, selecting and updating may, e.g., be repeatedly executed until a predefined number of iterations is reached. In an embodiment, the steps of determining, selecting and updating may, e.g., be repeatedly executed until the equalizer settings of the plurality of parametric audio equalizers exhibit a predefined quality. According to an embodiment, the steps of determining, selecting and updating may, e.g., be repeatedly executed until at least one of two or more of the following occurs: MCH2301-D-2024098091.DOCX - the equalizer settings of the plurality of parametric audio equalizers exhibit a predefined quality, - a predefined number of iterations is reached, - the equalizer settings of the plurality of parametric audio equalizers differ less than a defined threshold, - the equalizer settings for the plurality of parametric audio equalizers have not changed at all or not above or below a defined threshold over a set number of consecutive iterations. In an embodiment, one of the above-described methods may, e.g., be executed two or more times. When repeating said one of the above-described methods, the audio output signal originating from an immediately preceding execution of said one of the above- described methods may, e.g., be employed as the audio input signal of a current execution. When repeating said one of the above-described methods, a maximum allowed value for the frequency of the equalizer setting of at least one of the plurality of parametric equalizers may, e.g., be defined, wherein the method comprises increasing the maximum allowed frequency value for a current execution of said one of the above-described methods compared to an immediately preceding execution of said one of the above- described methods. This allows to start with determining an optimized equalizer only for low frequencies, to update the audio input signal and to then continue to search for an optimized equalizer for higher frequencies in later executions of the method. Often, this approach may often allow to realize even further improvements. Example values for a maximum allowed frequency may, e.g., be 100 Hz in the first iteration, 1000 Hz in the second iteration and 10000 Hz in the third iteration. Other values are also possible, for example, 150 Hz in the first iteration, 800 Hz in the second iteration and 8000 Hz in the third iteration. Or, e.g., 130 Hz in the first iteration, 700 Hz in the second iteration and 12000 Hz in the third iteration. Or, e.g., 20 Hz in the first iteration, 600 Hz in the second iteration and 9000 Hz in the third iteration. Other values are also possible. MCH2301-D-2024098091.DOCX In an embodiment, a minimum allowed frequency value may, e.g., also be increased, in subsequent repetitions. This concept may often be particularly beneficial, when the above-described methods are employed that determine a range for the gain or a maximum absolute gain value. Fig. 9 illustrates a system for automatic audio equalization of an audio input signal to obtain an audio output signal according to an embodiment. The system comprises a configurator 920 and a processor 930. The processor 930 is configured to apply one of a plurality of parametric audio equalizers on the audio input signal to obtain an audio output signal. To obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the configurator 920 is configured for: - Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters. - Determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response. - Selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer. - Updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer. According to some embodiments, the system may, e.g., be configured to conduct one of the above-described methods. In the following, further particular embodiments are described. MCH2301-D-2024098091.DOCX As already outlined, parametric equalization is conducted using parametric equalizers wherein each of the parametric equalizers is defined by equalization parameters, e.g., by a frequency, a gain and a Q factor. In the following, methods and systems for automatic audio equalization are provided, which, on receipt of an audio input signal, automatically determine suitable parametric equalizers, such as the (center) frequency, the gain and the Q factor. Fig.2 illustrates a particular method 200 for automatic audio equalization of an audio input signal according to an embodiment. First, first audio equalization parameters of a parametric audio equalizer are determined in step 210. The first audio equalization parameters may, e.g., comprise a frequency and a Q factor for an adjustment of an equalizer. Choosing the frequency and the Q factor may, e.g., be conducted randomly or quasi-randomly. In step 220, an important step of the inventive concept is conducted: The allowed gain adjustment / a maximum absolute allowed gain range for the chosen setting (the first audio equalization parameters, e.g., frequency and Q factor) is determined. For example, allowed gain adjustment for the chosen setting may, e.g., be looked up (for example, from a lookup table). In some embodiments, the convolution could be performed (e.g., inside the optimization loop) in each iteration, instead of via a lookup table. An actual gain is chosen depending on the allowed gain adjustment / depending on the maximum absolute allowed gain range. Details on how to determine the maximum absolute gain adjustment range are provided later on. According to embodiments, step 220 may, e.g., be conducted in different ways: A first option would be to determine the maximum absolute allowed gain / the allowed absolute gain adjustment, for example, a range of an allowed gain adjustment or a maximum absolute allowed gain, and to determine the gain for the chosen setting from the allowed range or so that the absolute value of the allowed gain is not greater than the absolute value of the maximum allowed gain. Such a gain may, e.g., then be determined, e.g., randomly or quasi-randomly, from the gain values within the allowed range. MCH2301-D-2024098091.DOCX A second option would to determine the maximum absolute allowed gain and to always use the additive inverse of the allowed gain as the gain for the chosen setting. A third option would be to at first determine an (initial) gain value for the chosen setting, e.g., randomly or quasi-randomly, to determine an allowed absolute range or the maximum or minimum allowed gain, and to adjust the initial gain value to a border value of the allowed range, if the initial gain value is outside the allowed range; or to set the initial absolute gain value to the maximum allowed absolute gain, if the initial absolute gain value is greater than the maximum allowed absolute gain value. Concepts for determining the maximum absolute gain value according to particular embodiments will be provided in the following. In embodiments, where an allowed range of gains is determined, different options exist for determining the minimum absolute gain value. A first option would be to set the minimum absolute gain value always to a predefined value, for example to 0. A second option would be to determine the minimum absolute gain value depending on the maximum absolute gain value. For example, the minimum absolute gain value may, e.g., set to 0.5, or to 0.25, or to 0.2, or to 0.1,or to 0.01, or to the additive inverse of those numbers of the maximum allowed absolute gain. Instead of determining an allowed maximum absolute value, in other embodiments, for example, an allowed minimum absolute value may, e.g., be determined. For example, instead of computing a maximum gmaxof the absolute value of a gain value g, one could likewise determine an allowed minimum value gneg,minof a negative gain gneg, wherein gneg= – g. In step 230 the finally determined setting (the chosen frequency and Q factor, and the determined gain) are applied to the audio input signal to obtain the equalized audio signal (the EQed audio signal) 240. An equalizer result evaluation may, e.g., be conducted in step 250. An example for such an equalizer result evaluation 250 is provided below with reference to Fig.7. In step 260 it may, e.g., be decided, whether or not the equalized audio signal exhibits sufficient quality depending on equalizer result evaluation 250. For example, the equalizer result evaluation 250 may, e.g., provide a measure, e.g., a quality value (e.g., a MCH2301-D-2024098091.DOCX quality score), for the quality of the equalized audio signal, and in step 260, the quality value may, e.g., be compared against a threshold value. Alternatively or additionally: a quality score might also come from a user; and / or a quality score might come from a comparison to an external audio signal; and / or a quality score might be linked to the environment of the system, e.g. speaker feedback. If the quality of the equalized audio signal is sufficient, the method 200 for automatic audio equalization may, e.g., be aborted. E.g., the equalized audio signal may, e.g.,then be output. For example, the equalized audio signal may, e.g., then be output via an interface, or may, e.g., be transmitted, for example, via the internet, to another computer or terminal or other device, which, for example, has provided the audio input signal. If the quality of the equalized audio signal is not sufficient, the method 200 may, e.g., be continued in step 210. There, e.g., the frequency and the Q factor may be chosen again. For example, a different frequency and a different Q factor may, e.g., be chosen. Or, for example, a same frequency and a different Q factor may, e.g., be chosen. Or, for example, a different frequency and a same Q factor may, e.g., be chosen. Or, for example, a same frequency and a same Q factor may, e.g., be chosen, and the gain is then chosen in step 220 randomly or quasi randomly from within the allowed range or below the maximum allowed absolute gain value; or, e.g., its absolute value is set to the maximum allowed absolute gain value if that was not the case in the first iteration. In step 230, according to an embodiment, the determined equalization setting (e.g., the chosen frequency and Q factor and the determined gain) may, e.g., be applied on the equalized audio signal that has been determined in the previous iteration. Or, in another embodiment, the determined setting may, e.g., be applied again on the (initial) audio input signal. In other embodiments, step 260 does not depend on a quality of the equalized audio signal. In such embodiments, for example, a predefined number of iterations (e.g., a defined limit of iterations) may, e.g., be conducted, for example, 20 iterations or 50 iterations or 100 iterations. When the maximum number of iterations is reached, the method 200 for automatic audio equalization is terminated and the (resulting) equalized audio signal is output. In such embodiments, the EQ result evaluation step 250 is not necessary. MCH2301-D-2024098091.DOCX In other embodiments, EQ result evaluation takes place, but when a maximum number of iterations is reached, the method 200 for automatic audio equalization is terminated, irrespective of the EQ result evaluation of the equalized audio signal 240 In further embodiments, no iterations are conducted, and a single equalizer setting processing of the steps 210, 220 and 230 is conducted to obtain the equalized audio signal 240. Then, steps 250 and 260 are not necessary. Or, in yet further embodiments, no iterations are conducted, but steps 210 and 220 are conducted for a plurality of different equalizer settings, e.g., determined in parallel or quasi-parallel, and the plurality of different equalizer settings are all applied on the audio input signal in step 230 to obtain the equalized audio signal 240 Regarding step 210, in all of the above embodiments, the equalizer parameters, e.g., frequency and Q value, may, e.g., not be chosen randomly or quasi-randomly, but may, e.g., alternatively be chosen according to a predefined rule. For example, a predefined list of settings may, e.g., be provided for the different number of iterations. Or, a predefined algorithm may, e.g., be provided how to change the frequency and / or the Q value in subsequent iteration steps. For example, the frequency may, e.g., be increased by a constant value, or may, e.g., be increased exponentially by a value ^^^^^^^^, where y is increased by a constant value. Analogous approaches may, e.g., be applied for adjusting the Q factor in subsequent iteration cycles. Or, for example, the frequency may, e.g., exhibit the same value in all iteration cycles. Or, for example, the Q factor, e.g., exhibit the same value in all iteration cycles. Or, for example, a user of the system may influence or limit the generation of EQ parameters. According to some embodiments, each iteration may, e.g., be performed on different parts of the input audio signal, e.g. chronologically or randomly For example, in live / realtime audio processing, e.g., each iteration of the loop works on a different buffer of the input audio signal Further details of particular embodiments are now provided with reference to Figures 3 to 8. Fig.3 depicts an illustration 300 with an audio input signal 310 that is received. Moreover, Fig.3 illustrates a desired target frequency response 320. MCH2301-D-2024098091.DOCX The target frequency response 320 may, e.g., be predefined. For example, the same target frequency response 320 may, e.g., be used for all audio input signals 310. Or, for example, the target frequency response 320 may, e.g., depend on the audio input signal 310. For example, the target frequency response 320 may, e.g., depend on the content of the audio input signal. E.g., a different target frequency response 320 may, e.g., be employed depending on whether the audio input signal is a music signal or whether the audio input signal is a speech signal. Or, e.g., a different target frequency response 320 may, e.g., be employed depending on a genre of the content of the audio input signal. Or, e.g., a different target frequency response 320 may, e.g., be employed depending on elements that can be detected in the audio signal. Or, e.g., a different target frequency response 320 may, e.g., be employed depending on a presence of vocals, no vocals, of particular instruments, e.g., guitars etc. Or, for example, in other embodiments a desired target frequency response may, e.g., be provided as input to the system in addition to the audio input signal. Or, for example, in an embodiment, a desired target frequency response may, e.g., be adjusted dynamically by a user after each loop iteration. Or, for example, in an embodiment, if the loop works on different parts of the audio signal, the target frequency response may also change depending on the section / the volume / other properties of the audio signal. Or, for example, in an embodiment, a desired target frequency response may, e.g., be determined based on a target device the output audio signal of the system is received by or played back on. Or, for example, in an embodiment, a target frequency response based on the environment of the system, e.g. speaker feedback, microphone input, or similar. Fig. 4 depicts an illustration 400 of a frequency response 420 of the audio input signal. The frequency response 420 of the audio input signal (which may be referred to as an input signal frequency response) is obtained by, e.g., transforming in step 410 the audio input signal, e.g., from a time domain to a frequency domain. For example, the audio input signal may, e.g., be transformed by conducting a Fast Fourier Transform from the time domain to the frequency domain. The person skilled in the art is aware that various other transformations and other approaches exists to transform an audio signal from the time domain to the frequency domain, which may be applied in other embodiments, for example, a Discrete Cosine Transform (DCT), or a Discrete Sine Transform (DST), or a Short-Time Fourier Transform (STFT), etc. MCH2301-D-2024098091.DOCX Fig. 5 depicts an illustration 500 where the target frequency response 320 is subtracted from the input signal frequency response 420 of the audio input signal 310 in step 510 to obtain a frequency response difference signal. In step 520, the frequency response difference signal is further processed, e.g., using a max(0, x) function, so that only the positive values, e.g., where the frequency response of the input signal exceeds the target frequency response, are kept. The resulting signal may, e.g., be referred to as a positive frequency response difference signal 530, and is illustrated in Fig.5. In other embodiments, instead of setting negative values of the frequency response difference signal to zero, a processed frequency response difference signal, which comprises the absolute values of the frequency response difference signal, may, e.g., be employed. In further embodiments, both the positive and the negative values may, e.g., be kept. In Fig. 5, the positive frequency response difference signal 530 is represented in a logarithmic domain (the frequency values on the abscissa axis are not arranged linearly). In embodiments, the positive frequency response difference signal 530 may, e.g., be resampled to a linear x-scale, see step 540. Fig.6 illustrates a general equalized model of an embodiment and a determination of an allowed gain range of a parametric audio equalizer according to an embodiment. As already outlined, an EQ adjustment (or a determination of a parametric audio equalizer / a determination of its audio equalization parameters) is commonly defined by three parameters: the (center) frequency, the gain, and the Q factor. The center frequency is the peak of the adjustment, e.g., the frequency at which the adjustment has the most effect on the audio signal (see peak 6115 in Fig.6f). The gain is the amount by which the amplitude of the center frequency is increased (gain > 0) or decreased (gain < 0). The Q factor determines the width of the EQ adjustment, e.g., the amount by which the frequencies surrounding the center frequency are affected. Parametric EQs usually employ Q factors between 0.2 (wide adjustment slope, last image above), or lower, to 8.0 (narrow adjustment slope, first image above), or higher. See: https: / / www.w3.org / TR / audio-eq-cookbook / . MCH2301-D-2024098091.DOCX Fig.6f depicts a logarithmic-domain magnitude / frequency response of an equalizer setting according to an embodiment, which has a center frequency of about 140 Hz. The peak of the logarithmic-domain frequency response 611 of the equalizer setting is depicted by reference signal 6115. The frequency response 611 of the equalizer setting of Fig.6f has a Q factor of 0.2. The illustration 610 of Fig. 6a also illustrates this logarithmic-domain frequency response 611 of the equalizer setting of Fig.6f, and, as examples, three further logarithmic-domain frequency responses of the three further equalizer settings having other Q factors. For example, logarithmic-domain frequency responses of equalizer settings with Q factors from a narrowest possible EQ Q factor, e.g. Q=8.0, to a widest possible EQ Q factor, e.g. Q=0.2, may, e.g., be employed. Fig. 6g illustrates all four logarithmic-domain frequency responses of all four equalizer settings of Fig.6a according to an embodiment in a better resolution. The employed logarithmic-domain frequency responses of equalizer settings may, e.g., be transformed or resampled to a linearly spaced x-scale in step 612 to obtain a plurality of linear-domain convolution kernels 614. For example, the logarithmic-domain frequency responses of the equalizer settings may, e.g., be resampled to a linear x-scale to obtain the convolution kernels. To obtain the plurality of linear-domain convolution kernels 614, e.g., normalization, e.g., to a kernel entry sum of 1, may be conducted in step 613. According to some embodiments, the convolution kernels may, e.g., be obtained as follows: To calculate a corresponding convolution kernel ^̂^^^ for a Q factor q: 1. Create an EQ setting hc= ^f, g, q^, where: - f is set to the center of the frequency range of the system on a logarithmically scaled x axis, e.g. f = (20000)0.5− 1 ≈ 140Hz if the system operates on a frequency scale from 0 Hz to 20000 Hz. This may, e.g., be done so that in step 3, there is the same number of linear x and corresponding y values to both sides of f. MCH2301-D-2024098091.DOCX - g is set to a constant value, e.g., g = 5 dB. g might be set more intelligently once it is known what typical range of absolute gain values the system ends up using in EQ settings. If it typically only uses absolute gain values between 0 dB and 1 dB, for instance, g could be set to g = 0.5 dB to better model the average frequency response of those EQ settings. 2. Calculate the frequency response Hcof hcon a logarithmically scaled x axis for the frequency range of the system. 3. Resample logarithmically scaled Hcon a new linear x axis. The sampled values of Hcproduce the unnormalized convolution kernel c for q. Alternatively: Approximate Hcwith a different function on a linear x axis, e.g. via the Generalized Normal Distribution. Then its values are the unnormalized convolution kernel c. 4. Normalize c to ^̂^^^ so that the sum of its entries is 1: Fig. 6c depicts the settings of the plurality of linear-domain convolution kernels 614 according to an embodiment. Each convolution kernel 614 is then convoluted with the difference of input and target frequency response, e.g., with the positive frequency response difference signal after transformation of said difference signal (in step 540) of Fig. 5. As both the difference signal and each of the convolution kernels 614 exhibit a same scale, here, a linear x- scale, convolution is possible. The convolution (which may, e.g., be referred to as “blurring”, see 622 of Fig.6b) of each convolution kernel 614 with the difference signal yields a smoothed frequency response 623 for each Q factor. Fig. 6b depicts an illustration 620 of the smoothed frequency response 623 for some Q factors. MCH2301-D-2024098091.DOCX Fig. 6d illustrates the smoothed frequency responses in more detailed representations 6231, 6232, 6233. The y-value (g1, g2, ...) of each smoothed frequency response represents the maximum allowed absolute EQ gain value i.e. the adjustment scope, using that Q setting at that specific frequency. In an embodiment depicted by Fig. 6e, the adjustment scope for each Q factor and frequency setting may, e.g., again be stored in a discrete lookup table. Fig. 7 depicts an illustration 700 for an equalizer (EQ) result evaluation according to an embodiment. In step 720, the equalized audio signal 240 is transformed into the frequency domain, e.g., by employing a Discrete Cosine Transform (DCT), and absolute values of the Discrete Cosine Transform (DCT) are determined to preserve phase information. In other embodiments, the Discrete Sine Transform may, e.g., be employed. To adjust the values of the DCT, in step 730, the target frequency response 320 is subtracted from the absolute values of the DCT-transformed equalized audio signal to obtain an adjusted DCT-transformed equalized audio signal. Finally, in step 740, the standard deviation of the adjusted DCT-transformed equalized audio signal is calculated and returned as a quality measure for the EQ adjustment. The lower the standard deviation, the better the applied EQ setting. For example, in an embodiment, the standard deviation may, e.g., then be compared with a threshold value, for example, in step 260 of Fig. 2. If the standard deviation is smaller than the threshold value, the quality of the equalized audio signal may, e.g., be considered sufficient, and the equalized audio signal may, e.g., be output. Otherwise, the iterations of Fig.2 may, for example, be continued. The standard deviation is a robust measure to identify how far the actual frequency response of the EQed audio signal deviates from the target frequency response. Throughout its optimization cycles, the system will find out if a wide EQ adjustment is more effective at bringing the audio signal closer to the target frequency response than a narrow adjustment. At the same time, the standard deviation of the DCT is a measure for the contribution of individual frequencies or groups of frequencies to the overall signal. The smaller the standard deviation, the more distinct fundamental frequencies the audio signal is comprised of, leading to a higher perceived clarity of the audio. Compared to MCH2301-D-2024098091.DOCX other time to frequency transformation approaches, the using the DCT also incorporates a certain level of phase information of the equalized signal in the quality score. According to some embodiments, equalizer result evaluation may, e.g., be employed for optimizing equalizer settings. According to a preferred embodiment, optimizing the audio equalization parameters of a plurality of parametric audio equalizers may, e.g., be conducted employing equalizer result evaluation by conducting one or more of the following steps: At first, in step 210, for each of the plurality of parametric audio equalizers, first audio equalization parameters, e.g., frequency and Q factor, may, e.g., randomly, be determined. In step 220, e.g., for each of the plurality of parametric audio equalizers, e.g., the absolute gain value, may be determined, by taking the maximum gain value that depends on the frequency and the Q factor of the respective parametric audio equalizer, into account. Then, in step 230, for each of the plurality of parametric audio equalizers, the respective audio equalizer is applied on the original audio signal to obtain an equalized audio signal for each of the plurality of parametric audio equalizers, and a quality value for each obtained equalized audio signal is obtained by conducting equalizer result evaluation 250. The method described with respect to Fig.7 may, e.g., be employed. For example, if the plurality of parametric audio equalizers comprises 20 parametric audio equalizers, then 20 equalized audio signals are obtained in step 240, by employing each of the 20 parametric audio equalizers on the original audio signal to obtain 20 equalized audio signals. Then, equalizer result evaluation is conducted in step 250 for each of the 20 equalized audio signals (each of the 20 equalized audio signals is associated with one of the 20 parametric audio equalizers that has been used to generate it). Thus, 20 quality values are obtained from the equalizer result evaluation 250, namely one quality value for each of the 20 parametric audio equalizers. The equalizer result evaluation 250 may, e.g., be conducted to identify a best quality value and by this a best parametric audio equalizer among the plurality of parametric audio equalizers. The best parametric audio equalizer may, e.g., be referred to as a selected audio equalizer. For example, in the embodiment of Fig. 7, the best parametric audio MCH2301-D-2024098091.DOCX equalizer among the plurality of parametric audio equalizers may, e.g., be the parametric audio equalizer having a smallest standard deviation. In such an embodiment, step 260 of Fig. 2 may, e.g., be implemented such that a predetermined number of loops / interations is conducted, e.g., for example, 30 loops / iterations. Other concepts are possible, as explained above. In the next iteration, it step 210, the audio equalization parameters (e.g., frequency, Q- factor and gain) of the best parametric audio equalizer may, e.g., be kept unmodified. In another embodiment some random deviation from the previous parameters may be introduced. Regarding the other parametric audio equalizers, the audio equalization parameters (e.g., frequency, Q-factor and gain) may, e.g., be determined depending on their previous values, but the respective audio equalization parameters (e.g., frequency, Q-factor and gain) of the best parametric audio equalizer may, e.g., influence their updated audio equalization parameters (e.g., frequency, Q-factor and gain). For example, an updated parameter value ^^^^^^^^,^^^^^^^^^^^^^^^^^^^^^^^^may, e.g., be determined as follows: ^^^^^^^^,^^^^^^^^^^^^^^^^^^^^^^^^ = (1 − ^^^^1) ∙ ^^^^^^^^ + ^^^^1 ∙ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^: a parameter of an i-th parametric audio equalizer ^^^^^^^^,^^^^^^^^^^^^^^^^^^^^^^^^: an update of the ^^^^^^^^parameter of the i-th parametric audio equalizer for the next iteration ^^^^^^^^^^^^^^^^^^^^: a corresponding parameter of the best parametric audio equalizer^^^^1: a real weighting factor with 0 < ^^^^1 < 1 .In a particular embodiment, when a plurality of loops / a plurality of iterations have already been conducted, a setting of a best audio equalization parameter of all iterations (best_global) may have been stored and may, together with the equalizer setting of the best parametric audio equalizer of the current iteration (best_local) also influence the parameters of the next iteration: ^^^^^^^^: a parameter of an i-th parametric audio equalizer MCH2301-D-2024098091.DOCX ^^^^^^^^,^^^^^^^^^^^^^^^^^^^^^^^^: an update of the ^^^^^^^^parameter of the i-th parametric audio equalizer for the next iteration ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^: a corresponding parameter of the best parametric audio equalizer of the current iteration ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^^^^^: a corresponding parameter of the best parametric audio equalizer of all iterations^^^^1: a first real weighting factor with 0 < ^^^^1 < 1 .^^^^2: a second real weighting factor with 0 < ^^^^2 < 1 , wherein ^^^^1 + ^^^^2 < 1In embodiments, additional (e.g., random) noise may intentionally influence the updated parameter values of the equalizer settings of the plurality of parametric audio equalizers. In a particular embodiment, the equalizer setting of the best parametric audio equalizer of all iterations (best_global) may, e.g., also result in an update of the equalizer setting of the best parametric audio equalizer of the current iteration, for example, according to: ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^_ ^^^^^^^^^^^^^^^^^^^^^^^^ = (1 − ^^^^1) ∙ ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^ + ^^^^1 ∙ ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^: a corresponding parameter of the best parametric audio equalizer of the current iteration ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^^^^^: a corresponding parameter of the best parametric audio equalizer of all iterations ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^,^^^^^^^^^^^^^^^^^^^^^^^^: an update of the ^^^^^^^^^^^^^^^^^^^^_^^^^^^^^^^^^^^^^^^^^parameter of the best parametric audio equalizer of the current iteration^^^^1: a real weighting factor with 0 < ^^^^1 < 1 .In step 220 of the next iteration, for each of the plurality of parametric audio equalizers, e.g., the gain value, may be updated, taking the maximum gain value that depends on the frequency and the Q factor of the respective parametric audio equalizer, into account. For example, in a particular embodiment, e.g., the gain value of the parametric audio equalizers may, e.g., be represented as a value between 0 and 1. For example, a gain value of one of the parametric audio equalizers may, e.g., be 0.8. Then, the maximum gain value of that parametric audio equalizer is recalculated in step 220 depending on the updated frequency value and updated Q factor value. While the relative gain value remains the same (here, 0.8), the absolute gain value changes from 0.8 ∙ -1 ∙ gmaxto 0.8 ∙ -1 ∙ gmax_update, when the maximum absolute gain value for that parametric audio equalizer has changed from gmax_updateto gmax. MCH2301-D-2024098091.DOCX In an embodiment, all parameters of the equalizer setting of the plurality of parametric audio equalizers may, e.g., be represented as relative values, e.g., between 0 and 1, and relative with respect to a maximum value. For example, a maximum frequency value may, e.g., always be 20000 Hz and a maximum Q-factor may, e.g., always be 8.0. Other maximum values are possible. The maximum absolute gain value changes dynamically in each iteration, and may, e.g., be individual for each of the parametric audio equalizers. In step 230, each of the plurality of parametric audio equalizers, now with an updated equalizer setting, is again applied on the audio input signal to obtain a new equalized audio signal 240 for each of the plurality of parametric audio equalizers and equalizer result evaluation is again conducted in step 250, to obtain a quality value for each of the plurality of parametric audio equalizers. Again a best parametric audio equalizer for the current iteration is determined, and the method continues with the next iteration in step 210. After the last iteration, (In step 260: abort: “yes”), the plurality of parametric audio equalizers each exhibit an optimized or at least improved equalizer setting. Then, in an embodiment, all of the plurality of parametric audio equalizers may, e.g., be applied on a same single version audio input signal to produce a single (equalized) audio output signal. Instead of conducting a predefined number of iterations, the above method may, e.g., terminate, when a predefined quality is reached. For example, in each iteration, it may be tested, whether applying all of the plurality of parametric audio equalizers on a same single version of the audio input signal results in a signal with sufficient quality. A threshold test may, e.g., be conducted as described above. In another embodiment, EQ result evaluation may, e.g., be performed on the audio input signal without applying any EQ, and if the quality score is below or above a threshold, the whole optimization process is skipped. In an embodiment, the iterations may, e.g., terminate, when at least one of two or more of the following occurs: - a sufficient quality is achieved - a predefined number of iterations has been conducted. - The settings of the plurality of parametric audio equalizers differ less or more than a defined threshold. MCH2301-D-2024098091.DOCX - The settings for the plurality of parametric audio equalizers have not changed at all or not above or below a defined threshold over a set number of consecutive iterations. Fig. 8 illustrates a method according to particular embodiment, which employs the steps described with respect to the illustration 200 of Fig. 2, the illustration 300 of Fig. 3, the illustration 400 of Fig. 4, the illustration 500 of Fig. 5, the illustration 610 of Fig. 6a, the illustration 620 of Fig.6b, the illustration 630 of Fig.6e, and the illustration 700 of Fig.7. As already explained with respect to Fig. 2 and the other figures, it is apparent for the person skilled in the art that in other embodiments some of the method steps are not necessary or may be replaced by other, alternative method steps. It should be noted that instead of determining a maximum gain value in step 220 depending on the frequency and the Q-factor of an equalizer setting of a parametric audio equalizer, it is likewise possible to: - determining a maximum or minimum frequency or a frequency range in step 220 depending on the gain value and the Q-factor of an equalizer setting of a parametric audio equalizer; or - determining a maximum or minimum Q-factor or a Q-factor range in step 220 depending on the gain value and the frequency of an equalizer setting of a parametric audio equalizer by similarly applying the principles of described above. In the following, concepts of particular embodiments are described in more detail. At first, a system description according to embodiments is provided. The system receives the audio signal x as input, modifies that audio signal, and outputs the modified audio signal y. The modification the system performs to the audio signal is an equalization (EQ) adjustment with setting hbest. The goal of the system is to find hbestto attenuate unpleasant frequencies in x. MCH2301-D-2024098091.DOCX A peaking EQ adjustment is commonly defined by three parameters: the center frequency f, the gain g, and the Q factor q. f locates the peak of the adjustment, e.g., the frequency at which the EQ adjustment has the most effect on the audio signal. g is the amount, typically in decibels, by which the amplitude of f is increased or decreased. q determines the bandwidth of the EQ adjustment, i.e. the extent to which the frequencies surrounding f are affected. Parametric EQs typically allow for q values between 0.2 or lower for a wide adjustment slope to 8.0 or higher for a narrow adjustment slope. We define an EQ adjustment setting as h = ^f, g, q^, where f ∈ [fmin, fmax] and q ∈ [qmin, qmax] for the system. fminand fmaxcan e.g. be set to fmin= 0, fmax= 20000 if the system is supposed to calculate settings across the full humanly audible frequency range, or to a narrower interval if only adjustments within that interval are desired. The Q factor bounds qminand qmaxcan be set to limit the range of Q factors the system may use, e.g., qmin= 0.2, qmax= 8.0. The presented system only considers peaking EQ filters, and not other types of EQ filters, e.g., high shelf or low shelf filters, but the system can be modified to work with those filter types as well. The system is comprised of two parts. The first part is an optimization loop that finds an EQ adjustment setting hbestfor x based on a cost function J. It starts with a set of random EQ settings and eventually converges towards an EQ setting hbestwith the lowest cost for x. The second part is an optimization guide that constrains all h during the optimization loop that may be applied to x, making the optimization loop converge within dynamically calculated bounds. In the following, the system input according to embodiments is considered. The only required user input to the system is the time-based discrete source audio signal x defined with N mono samples. In addition, a target magnitude response T is required. T can be static, or cater to different use cases of the system, e.g., music, podcasts, video sounds etc., or to achieve a desired target magnitude response for other purposes, e.g. to reflect the personal preference of an audio engineer or a user. T is considered to be available as a continuous function In particular embodiments, input adjustments may, e.g., be conducted. MCH2301-D-2024098091.DOCX x is normalized to a set perceived integrated loudness level, e.g. according to ITU-R BS.1770, and T is defined with respect to that same loudness level. If x consists of too many samples that cannot be processed by the system for computational or other reasons, or if the system is intended to only analyze parts of the original audio signal, a characteristic sub-sequence x’ of x with length L can be extracted as input to the system. For example, the sub-sequence with the highest absolute amplitude sum in the frequency band [fmin, fmax] can be extracted from x as shown in Equation 1, where eq refers to a band-pass filter or consecutive high-pass and low-pass filters. If the system is used to equalize a stereo signal with a left channel , x can be obtained using Equation 2, where b can be used to incorporate more information from the left or right channel into x, respectively, e.g., b = 0.6 for more contribution of the left channel. If the system is used to equalize a spatial audio or multichannel surround audio signal, that signal is rendered down or converted to a stereo signal first, e.g., according to ITU-R BS.2127, and then Equation 2 is applied to obtain the mono signal x as input to the system. In particular embodiments, an optimization loop may, e.g., be executed. The optimization loop uses Particle Swarm Optimization (PSO) to find an optimal EQ setting hbestfor x. The particle population p for the PSO algorithm consists of P particle MCH2301-D-2024098091.DOCX position vectors, e.g. P = 25, individually referred to as piin each iteration with j = 3 dimensions as shown in Equation 3. In the following, a cost calculation is considered: The cost of each particle position piis calculated by first converting piinto a corresponding EQ setting hi. This conversion is explained starting with Equation 13 below. For each pi, the corresponding EQ setting hiis then used to equalize the system’s input signal x to receive a modified signal vivia an eq function as per Equation 4, e.g. by converting hito filter coefficients (see https: / / www.w3.org / TR / audio-eq-cookbook / ) and applying those to x. yiof length N is then used to calculate the cost Jifor pi, using the DCT-II Di of yias shown in Equation 5. Jiis the standard deviation of the difference between the absolute values of Diand the system’s target magnitude response T. Subtraction of T from Diis possible by evaluating frequencies as per Equation 6, where r is the sample rate of the system, e.g., r = 44100 [Hz]. fkfrom Equation 6 will also be used in equations below. MCH2301-D-2024098091.DOCX In the following, iterations according to embodiments are considered. Before the start of the optimization loop, each particle position dimension pijis initialized as pij∼ Uniform(0, 1). The velocity vi for each particle position piis defined as and its dimensions are initialized as vij∼ Uniform(-0.1, 0.1). For each of the P particles, the historical positions pbestwith the lowest associated cost across iterations are stored as Before the start of the optimization loop, pbestis initialized as pbest= p. The single particle position pglob_bestwith the lowest cost out of all particles p is calculated and initialized as shown in Equation 7. The optimization loop starts by first updating particle velocities as shown in Equation 8, where w, c1and c2are chosen constants, e.g. w = 0.5, c1= 0.6, c2= 0.2, with and and The velocities can be constrained to constant bounds vmin and vmax as shown in Equation 9, e.g. vmin = −0.1, vmax = 0.1. The new particle dimensions p′ijare then calculated as shown in Equation 10, adding randomized local search noise . (10) The particle dimensions are then constrained to [0, 1] as shown in Equation 11. MCH2301-D-2024098091.DOCX For each particle, its new historical position pi_bestwith the lowest cost is then updated as shown in Equation 12. Finally, the global best particle position pglob_best is calculated again using Equation 7 and the loop repeats. The loop ends after a constant number of iterations have been completed, or if J(pglob_best) no longer decreases by more than a defined threshold. The EQ setting hi and the equalized signal yi that are evaluated during the final cost calculation for pglob_best are the optimal EQ setting hbest and output signal y of the system, respectively. If x is a signal that was derived from a stereo or other audio signal as described above, then the EQ setting hbest is applied to each channel of the original audio signal. In the following, an optimization guide according to embodiments is described. The optimization guide converts each particle position pi from the optimization loop above to an EQ setting hi via Equation 13. This makes the optimization loop converge towards an EQ setting within dynamically calculated gain bounds. The gain bound for a setting is determined by the gmaxfunction specified in Equation 14, with precalculations dependent on the system input signal x. MCH2301-D-2024098091.DOCX Z is the excess magnitude response of x, consisting of the positive values of the difference between the magnitude response X of x and the system’s target magnitude response T at corresponding frequencies fk(k). Via a resampling function , Z is evaluated on a logarithmic scale using a linearly spaced sequence a of length L > 2, e.g. L = 1000, as per Equation 15. Further explaining Equation 14, ^̂^^^^^^^is a linear convolution kernel that models the magnitude response of a peaking EQ adjustment with q on the same resampled x-axis corresponding to sequence a from Equation 15. ^̂^^^^^^^is convoluted with the resampled version of Z, and the convolution result is sampled back to the original x -axis of Z via ^^^^^^^−^1. MCH2301-D-2024098091.DOCX gmaxfor f and q is then given as the additive inverse of the closest available value of ^^^^′ at approximately frequency f. How ^̂^^^^^^^is calculated is explained in the following. First, a helper EQ setting is defined as center of the frequency range of the system on a logarithmically scaled x-axis, e.g. if system operates on a frequency scale from 0 Hz to 20000 Hz. This is done so that in the later resampling step, there is the same number of linear x and corresponding y values to both sides of fc. gcis a constant value, e.g. gc= 5db. gcmight be set more intelligently once it is known what typical range of gain values the system ends up using in EQ settings. If it typically only uses gain values between 0 db and 1 db, for instance, gccould be set to gc= 0.5db to better model the average magnitude response of those EQ settings. Then, the M-order numerator b and the N-order denominator a for hqare used to calculate its magnitude response Hqvia Equation 16 (see https: / / www.w3.org / TR / audio-eq-cookbook / ). The magnitude responseHq in relation to frequencies f can be obtained via Equation 17,where r is the sampling rate of the system. In order to obtain the symmetrical unnormalized convolution kernel cq, Hqis evaluated at fk(k) frequencies with k ∈ 0, 1, ..., N − 1 and then resampled using fras shown in Equation 18. MCH2301-D-2024098091.DOCX cqis finally normalized to ^̂^^^^^^^so its entries sum up to 1, as shown in Equation 19. Calculating ^^^^′ for each particle position in each optimization loop iteration can be computationally intensive. It is possible to precalculate ^^^^′ for a range of different q values before the optimization loop starts. For a q that has no precalculated gmaxvalue, the gmaxvalue can be inter- or extrapolated from the precalculated gmaxvalues based on the nearest q values used in precalculations. The described system assumes the input signal x to be static throughout the optimization process. For live-audio or other realtime applications, buffered sequences of x can also be passed into the system, such that different iterations of the optimization loop work on different sub-sequences of x. In that case, the system may never converge towards one constant EQ setting hbest, but instead continuously produce new temporarily optimal EQ settings. Alternatively, instead of dynamically calculating the g value in each loop, f might also be found dynamically, e.g. via the following steps: 1. Calculate g and q for hias follows: (20) where gmin and gmax are set for the system. 2. Via favailable, obtain a sequence F of all the frequency values at which the convoluted difference magnitude response for q equals −1・ g. Then, the value for f in hiis obtained MCH2301-D-2024098091.DOCX by multiplying pi0with the length M of F, rounding that result to the nearest integer n, and selecting the n-th element from F. If no frequencies can be found for a desired g value, i.e. the length of F is 0, g may e.g. be set to 0 to prevent any adjustments. Or, the search for frequencies via favailablemight be repeated with a new g value derived from g, e.g.0.5・g, until at least one frequency is found. In the following, further embodiments are provided. A method for automatic audio equalization of an audio input signal to obtain an equalized audio signal according to an embodiment is provided. The method comprises: - Determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer. And: - Applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal. Determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response. Furthermore, a system for automatic audio equalization of an audio input signal to obtain an equalized audio signal according to an embodiment is provided. The system is configured for: Determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer, and applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal. Determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the MCH2301-D-2024098091.DOCX parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response. Although some aspects have been described in the context of a system or in the context of an apparatus, it is clear that these aspects also represent a description of the corresponding method, where a block or device corresponds to a method step or a feature of a method step. Analogously, aspects described in the context of a method step also represent a description of a corresponding block or item or feature of a corresponding system or of a corresponding apparatus. Some or all of the method steps may be executed by (or using) a hardware apparatus, like for example, a microprocessor, a programmable computer or an electronic circuit. In some embodiments, one or more of the most important method steps may be executed by such an apparatus. Depending on certain implementation requirements, embodiments of the invention can be implemented in hardware or in software or at least partially in hardware or at least partially in software. The implementation can be performed using a digital storage medium, for example a floppy disk, a DVD, a Blu-Ray, a CD, a ROM, a PROM, an EPROM, an EEPROM or a FLASH memory, having electronically readable control signals stored thereon, which cooperate (or are capable of cooperating) with a programmable computer system such that the respective method is performed. Therefore, the digital storage medium may be computer readable. Some embodiments according to the invention comprise a data carrier having electronically readable control signals, which are capable of cooperating with a programmable computer system, such that one of the methods described herein is performed. Generally, embodiments of the present invention can be implemented as a computer program product with a program code, the program code being operative for performing one of the methods when the computer program product runs on a computer. The program code may for example be stored on a machine readable carrier. Other embodiments comprise the computer program for performing one of the methods described herein, stored on a machine readable carrier. In other words, an embodiment of the inventive method is, therefore, a computer program having a program code for performing one of the methods described herein, when the computer program runs on a computer. MCH2301-D-2024098091.DOCX A further embodiment of the inventive methods is, therefore, a data carrier (or a digital storage medium, or a computer-readable medium) comprising, recorded thereon, the computer program for performing one of the methods described herein. The data carrier, the digital storage medium or the recorded medium are typically tangible and / or non-transitory. A further embodiment of the inventive method is, therefore, a data stream or a sequence of signals representing the computer program for performing one of the methods described herein. The data stream or the sequence of signals may for example be configured to be transferred via a data communication connection, for example via the Internet. A further embodiment comprises a processing means, for example a computer, or a programmable logic device, configured to or adapted to perform one of the methods described herein. A further embodiment comprises a computer having installed thereon the computer program for performing one of the methods described herein. A further embodiment according to the invention comprises an apparatus or a system configured to transfer (for example, electronically or optically) a computer program for performing one of the methods described herein to a receiver. The receiver may, for example, be a computer, a mobile device, a memory device or the like. The apparatus or system may, for example, comprise a file server for transferring the computer program to the receiver. In some embodiments, a programmable logic device (for example a field programmable gate array) may be used to perform some or all of the functionalities of the methods described herein. In some embodiments, a field programmable gate array may cooperate with a microprocessor in order to perform one of the methods described herein. Generally, the methods are preferably performed by any hardware apparatus. The apparatus described herein may be implemented using a hardware apparatus, or using a computer, or using a combination of a hardware apparatus and a computer. The methods described herein may be performed using a hardware apparatus, or using a computer, or using a combination of a hardware apparatus and a computer. MCH2301-D-2024098091.DOCX The above described embodiments are merely illustrative for the principles of the present invention. It is understood that modifications and variations of the arrangements and the details described herein will be apparent to others skilled in the art. It is the intent, therefore, to be limited only by the scope of the impending patent claims and not by the specific details presented by way of description and explanation of the embodiments herein. MCH2301-D-2024098091.DOCX

Claims

Claims:

1. A method for automatic audio equalization of an audio input signal to obtain an audio output signal, wherein the method comprises applying one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal, wherein, to obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the method comprises: Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters, determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response, selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer, updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer.

2. The method according to claim 1, wherein said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal after the equalizer setting for each of the plurality of parametric audio equalizers has been determined is the parametric audio equalizer having a best quality indicated by the quality information among the plurality of audio equalizers.

3. The method according to claim 1 or 2, MCH2301-D-2024098091.DOCXwherein determining the equalizer setting of each of the plurality of parametric audio equalizers comprises determining an allowed maximum value or an allowed minimum value or an allowed range of at least one parameter of the two or more parameters of the equalizer setting.

4. The method according to claim 3, wherein determining the allowed maximum value or the allowed minimum value or the allowed range of said at least one parameter of the two or more parameters of said equalizer setting is conducted depending on a difference between the frequency response of the audio input signal and the target frequency response.

5. The method according to claim 3 or 4, wherein determining the allowed maximum value or the allowed minimum value or the allowed range of said at least one parameter of the two or more parameters of said equalizer setting is conducted depending on one or more other parameters of said equalizer setting.

6. The method according to claim 5, wherein the equalizer setting of each parametric audio equalizer comprises a frequency, a Q factor and a gain, wherein the method comprises determining the allowed maximum value or the allowed minimum value or the allowed range of the gain of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the Q-factor of the parametric setting, or wherein the method comprises determining the allowed maximum value or the allowed minimum value or the allowed range of the frequency of the equalizer setting of each of the plurality of parametric audio equalizers depending on the gain and the Q-factor of the parametric setting, or wherein the method comprises determining the allowed maximum value or the allowed minimum value or the allowed range of the Q-factor of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the gain of the parametric setting. MCH2301-D-2024098091.DOCX7. The method according to claim 6, wherein the method comprises determining the allowed maximum value or the allowed minimum value or the allowed range of the gain of the equalizer setting of each of the plurality of parametric audio equalizers depending on the frequency and the Q-factor of the parametric setting by: determining a difference signal depending on the frequency response of the audio input signal and the target frequency response, determining a convoluted signal which depends on the difference signal and which depends on the frequency and the Q-factor of the equalizer setting, and determining the allowed maximum value or the allowed minimum value of the gain of the equalizer setting depending on the convoluted signal.

8. The method according to claim 7, wherein the allowed maximum value or the allowed minimum value or the allowed range depends on a value of the convoluted signal at the frequency of the equalizer setting.

9. The method of claim 7 or 8, wherein the method comprises determining the difference signal so that the difference signal is represented in a logarithmic domain and indicates at least all positive values of a frequency response of the audio input signal and the target frequency response, and transforming or resampling the difference signal from the logarithmic domain to a linear domain so that the difference signal exhibits a linear x-scale.

10. The method according to one of claims 7 to 9, wherein the method comprises determining the convoluted signal using the difference signal and using at least one entry selected from a lookup table depending on the frequency and the Q-factor of the equalizer setting, MCH2301-D-2024098091.DOCXwherein the lookup table comprises a plurality of entries for a plurality of pre- calculated convolution kernels for different Q-factors.

11. The method according to one of claims 7 to 9, wherein the method comprises determining a convolution kernel depending on the frequency and the Q-factor of the equalizer setting, and conducting a convolution of the difference signal or of a portion of the difference signal and the convolution kernel to obtain the convoluted signal.

12. The method according to one of claims 7 to 11, wherein the method comprises calculating a convolution kernel according to: generating a frequency response of the equalizer setting in a logarithmic domain depending on the frequency and the Q-factor of the equalizer setting, transforming or resampling the frequency response of the equalizer setting from the logarithmic domain to a linear domain so that the frequency response exhibits a linear x-scale to obtain the convolution kernel, and normalizing the convolution kernel, so that the sum of all values of the convolution kernel is 1.

13. The method of claim 6, wherein the method comprises determining the allowed maximum value or the allowed minimum value or the allowed range of the frequency of the equalizer setting of each of the plurality of parametric audio equalizers depending on the gain and the Q-factor of the parametric setting.

14. The method according to one of the preceding claims, MCH2301-D-2024098091.DOCXwherein determining quality information for each parametric audio equalizer of the plurality of parametric audio equalizers is conducted by applying said parametric audio equalizer on the audio input signal to obtain an equalized audio signal, and by determining the quality information using the equalized audio signal.

15. The method according to claim 14, wherein determining the quality information for said parametric audio equalizer is conducted by: transforming the equalized audio signal to the frequency domain to obtain a frequency-domain equalized audio signal, subtracting the target frequency response from absolute values of the frequency-domain equalized audio signal to obtain an adjusted frequency- domain equalized audio signal, and calculating a standard deviation of the adjusted frequency-domain equalized audio signal to obtain the quality information, wherein determining the quality information for said parametric audio equalizer is, e.g., conducted depending on phase information.

16. The method of claim 15, wherein transforming the equalized audio signal to the frequency domain is conducted using a Discrete Cosine Transform or a Discrete Sine Transform or a Discrete Fourier Transform.

17. The method according to claim 15 or 16, further depending on one of claims 7 to 14.

18. The method according to one of the preceding claims, wherein the following steps of determining, selecting and updating: determining quality information for each of the plurality of parametric audio equalizers, MCH2301-D-2024098091.DOCXselecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer, updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer, are repeatedly executed one or more times to obtain resulting equalizer settings of the plurality of parametric audio equalizers, and wherein said one of the plurality of parametric audio equalizers is applied with its resulting equalizer settings on the audio input signal to obtain the audio output signal.

19. The method according to claim 18, wherein said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, after a last execution of the steps of determining, selecting and updating has been conducted, is the parametric audio equalizer having a best quality indicated by the quality information among the plurality of audio equalizers.

20. The method according to claim 18 or 19, wherein updating the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer, is conducted further depending on a stored equalizer setting of one of the plurality of parametric audio equalizers of a previous iteration preceding a current iteration of executing the steps of determining, selecting and updating, which exhibits a best quality value among the quality values of the equalizer setting of the plurality of parametric audio equalizers of all iterations of executing the steps of determining, selecting and updating.

21. The method according to one of claims 18 to 20, MCH2301-D-2024098091.DOCXwherein updating the equalizer setting of each of the plurality of parametric audio equalizers is conducted depending on the equalizer setting of the selected audio equalizer and further depending on noise or on a random process.

22. The method according to one of claims 18 to 21, wherein the steps of determining, selecting and updating are repeatedly executed until a predefined number of iterations is reached.

23. The method according to one of claims 18 to 21, wherein the steps of determining, selecting and updating are repeatedly executed until the equalizer settings of the plurality of parametric audio equalizers exhibit a predefined quality.

24. The method according to one of claims 18 to 21, wherein the steps of determining, selecting and updating are repeatedly executed until at least one of two or more of the following occurs: the equalizer settings of the plurality of parametric audio equalizers exhibit a predefined quality, a predefined number of iterations is reached, the equalizer settings of the plurality of parametric audio equalizers differ less than or more than a defined threshold, the equalizer settings for the plurality of parametric audio equalizers have not changed at all or not above or not below a defined threshold over a set number of consecutive iterations.

25. The method according to claim 17, further depending on one of claims 18 to 24.

26. The method according to one of the preceding claims, wherein the method according to one of the preceding claims is executed two or more times, MCH2301-D-2024098091.DOCXwherein when repeating the method according to one of the preceding claims, the audio output signal originating from an immediately preceding execution of the method according to one of the preceding claims is employed as the audio input signal of a current execution.

27. The method according to claim 26, further depending on claim 6, wherein when repeating the method according to one of claims 1 to 25, a maximum allowed value or a minimum allowed value for the frequency of the equalizer setting of at least one of the plurality of parametric equalizers is defined, wherein the method of claim 27 comprises increasing the maximum allowed value or the minimum allowed value for the frequency for a current execution of the method according to one of claims 1 to 25 compared to an immediately preceding execution of the method according to one of claims 1 to 25.

28. The method according to claim 27, further depending on claim 7.

29. A method for automatic audio equalization of an audio input signal to obtain an equalized audio signal, wherein the method comprises: determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer, and applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal, wherein determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response.

30. The method according to claim 29, MCH2301-D-2024098091.DOCXwherein determining the allowed maximum gain value or the allowed minimum gain value or the allowed gain range comprises convoluting a frequency response difference signal with a convolution kernel signal to obtain a convolution signal, wherein the frequency response difference signal depends on a difference between a frequency response of the audio input signal and the target frequency response.

31. The method according to claim 30, wherein the convolution kernel signal depends on the first audio equalization parameters.

32. The method according to claim 31, wherein the first audio equalization parameters comprise a frequency and a Q factor.

33. The method according to one of claims 30 to 32, wherein determining the convolution kernel is conducted by generating a frequency response of an equalizer setting in a logarithmic domain depending on the Q-factor of an original equalizer setting, transforming or resampling the frequency response of the equalizer setting from the logarithmic domain to a linear domain so that the frequency response exhibits a linear x-scale to obtain the convolution kernel, and normalizing the convolution kernel, so that the sum of all values of the convolution kernel is 1.

34. The method of claim 33, wherein the method comprises determining the difference signal so that the difference signal is represented in a logarithmic domain and indicates at least all positive values of a frequency response of the audio input signal and the target frequency response, and MCH2301-D-2024098091.DOCXtransforming or resampling the difference signal from the logarithmic domain to a linear domain so that the difference signal exhibits a linear x-scale.

35. A computer program for implementing the method according to one of claims 1 to 34, when the computer program is executed on a computer or signal processor.

36. A non-transitory computer-readable medium comprising a computer program for implementing the method according to one of claims 1 to 34, when the computer program is executed on a computer or signal processor.

37. A digital storage medium comprising a computer program for implementing the method according to one of claims 1 to 34, when the computer program is executed on a computer or signal processor.

38. A data signal comprising computer-readable instructions for implementing the method according to one of claims 1 to 34, when the computer-readable instructions are executed on a computer or signal processor.

39. A non-transitory computer-readable medium comprising a data signal comprising computer-readable instructions for implementing the method according to one of claims 1 to 34, when the computer-readable instructions are executed on a computer or signal processor.

40. A digital storage medium comprising a data signal comprising computer-readable instructions for implementing the method according to one of claims 1 to 34, when the computer-readable instructions are executed on a computer or signal processor.

41. An equalized audio signal being obtained by executing the method according to one of claims 1 to 34.

42. A non-transitory computer-readable medium comprising an equalized audio signal being obtained by executing the method according to one of claims 1 to 34.

43. A digital storage medium comprising an equalized audio signal being obtained by executing the method according to one of claims 1 to 34. MCH2301-D-2024098091.DOCX44. A system for automatic audio equalization of an audio input signal to obtain an audio output signal, wherein the system is configured for applying one of a plurality of parametric audio equalizers on the audio input signal to obtain the audio output signal, wherein, to obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the system is configured for: Determining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters, determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response, selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer, updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer.

45. A system for automatic audio equalization of an audio input signal to obtain an audio output signal, wherein the system comprises a configurator and a processor, wherein the processor is configured for applying one of a plurality of parametric audio equalizers on the audio input signal to obtain an audio output signal, wherein, to obtain said one of the plurality of parametric audio equalizers that is to be applied on the audio input signal to obtain the audio output signal, the configurator is configured for: MCH2301-D-2024098091.DOCXDetermining an equalizer setting for each of the plurality of parametric audio equalizers, wherein the equalizer setting comprises two or more parameters, determining quality information for each of the plurality of parametric audio equalizers depending a frequency response of the audio input signal and depending on a target frequency response, selecting one of the plurality of parametric audio equalizers depending on the quality information for each of the plurality of parametric audio equalizers as a selected audio equalizer, updating, depending on the equalizer setting of the selected audio equalizer, the equalizer setting of each of the plurality of parametric audio equalizers, being different from the selected audio equalizer.

46. A system according to claim 44 or 45, wherein the system is configured to conduct the method according to one of claims 1 to 28.

47. A system for automatic audio equalization of an audio input signal to obtain an equalized audio signal, wherein the system is configured for: determining an actual gain of the parametric audio equalizer depending on first audio equalization parameters of an equalizer setting of a parametric audio equalizer, and applying the parametric audio equalizer on the audio input signal to obtain the equalized audio signal, wherein determining the actual gain comprises determining an allowed maximum gain value or an allowed minimum gain value or an allowed gain range of the parametric audio equalizer depending on the first audio equalization parameters and depending on the audio input signal and depending on a target frequency response.

48. A system according to claim 47, MCH2301-D-2024098091.DOCXwherein the system is configured to conduct the method according to one of claims 29 to 34. MCH2301-D-2024098091.DOCX

Images