Audio signal tone modification processing method and device, storage medium and electronic equipment

By performing windowing and interpolation processing on the audio signal, combined with historical frame data, the problem of large delay in audio tone shift processing in the prior art is solved, and audio tone shift processing with lower delay and higher user experience is achieved.

CN118918907APending Publication Date: 2024-11-08XG TECHNOLOGIES PTE LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410963704.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing audio tone-changing processing scheme is based on the overlapping addition method of time domain processing, resulting in a large delay and reducing user experience.

Method used

By using the first window function to window the audio signal, combined with interpolation processing and historical frame data, the delay of the audio signal after the tone change processing is reduced, and the compensation of the window function envelope is realized, thereby reducing the signal Overlapping part.

Benefits of technology

It effectively reduces the delay of the audio signal, improves user experience, reduces the frame overlap rate, meets the perfect reconstruction conditions at a smaller frame overlap rate, and reduces fluctuations and distortions during signal superposition.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118918907A_ABST
    Figure CN118918907A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses an audio signal tone modification processing method and device, a storage medium and electronic equipment, and the method comprises the steps: carrying out the processing of a first audio signal according to a first window function, and obtaining a second audio signal, the first audio signal being a current frame audio signal; respectively executing first interpolation processing on the second audio signal and the first window function according to a preset tone modification multiple to correspondingly obtain a third audio signal and a second window function; and obtaining a fourth audio signal according to the third audio signal, the second window function and historical frame data. According to the embodiment of the invention, the overlapping part in the third audio signal is reduced, so that the time delay is reduced on the premise that the obtained fourth audio signal conforms to the preset tone modification multiple.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the technical field of audio processing, and in particular to a method, device, storage medium and electronic device for processing pitch shift of an audio signal. Background Art

[0002] As we all know, the basic elements of sound are pitch, intensity and timbre. Audio pitch shifting refers to changing the pitch of a sound, raising or lowering the pitch, while keeping the playback time of the audio signal unchanged. Audio pitch shifting can change the characteristics of a sound and is a popular technology in audio processing. It has a wide range of applications in many areas, such as music synthesis, dubbing, online video chat, and secure phone calls.

[0003] Audio pitch shifting methods, also known as pitch shifting without changing speed, time domain compression and expansion, etc., are commonly used methods in the audio field. Based on operations in the time domain or frequency domain, the algorithm controls the fundamental frequency of the input signal, thereby achieving control of the human voice pitch, achieving corrective or interesting sound effects, or playing an important role in speech synthesis. In the audio pitch shifting method, delay is an important indicator. The lower the delay of the output signal compared to the input signal, the better, otherwise it will cause a decrease in the sense of hearing. Summary of the invention

[0004] Embodiments of the present disclosure provide a method, device, storage medium and electronic device for processing pitch shift of an audio signal.

[0005] According to one aspect of an embodiment of the present disclosure, a method for processing pitch shift of an audio signal is provided, comprising:

[0006] Processing a first audio signal according to a first window function to obtain a second audio signal, wherein the first audio signal is a current frame audio signal;

[0007] According to a preset modulation multiple, performing a first interpolation process on the second audio signal and the first window function respectively, and correspondingly obtaining a third audio signal and a second window function;

[0008] A fourth audio signal is obtained according to the third audio signal, the second window function and the historical frame data.

[0009] According to another aspect of an embodiment of the present disclosure, there is provided a pitch shift processing device for an audio signal, comprising:

[0010] a windowing processing module, configured to process the first audio signal according to a first window function to obtain a second audio signal, wherein the first audio signal is a current frame audio signal;

[0011] An interpolation processing module, configured to perform a first interpolation processing on the second audio signal and the first window function according to a preset modulation multiple, to obtain a third audio signal and a second window function respectively;

[0012] A signal synthesis module is used to obtain a fourth audio signal according to the third audio signal, the second window function and the historical frame data.

[0013] According to another aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the pitch shift processing method of the audio signal described in any of the above embodiments.

[0014] According to another aspect of the embodiments of the present disclosure, an electronic device is provided, the electronic device comprising:

[0015] processor;

[0016] a memory for storing instructions executable by the processor;

[0017] The processor is used to read the executable instructions from the memory and execute the instructions to implement the pitch shift processing method of the audio signal described in any one of the above embodiments.

[0018] Based on the above-mentioned embodiments of the present disclosure, a method, device, storage medium and electronic device for processing the pitch shift of an audio signal are provided. A windowing process is performed on a second audio signal to be synthesized by a first window function, so as to facilitate subsequent signal synthesis processing based on the window function. In addition, a first interpolation process is performed on the second audio signal and the first window function to compensate for the information difference caused by the pitch shift by a preset pitch shift multiple. A third audio signal after the first interpolation process and the second window function are synthesized with historical frame data. Due to the existence of envelope compensation of the window function, the overlapping part in the third audio signal is reduced, so that the delay of the obtained fourth audio signal is reduced while meeting the preset pitch shift multiple.

[0019] The technical solution of the present disclosure is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other purposes, features and advantages of the present disclosure will become more apparent by describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. In the drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1is a flowchart of a method for processing pitch shifting of an audio signal provided by an exemplary embodiment of the present disclosure;

[0022] Figure 2 is a schematic diagram of a flow chart of determining a fourth audio signal in a pitch shift processing method for an audio signal provided by an exemplary embodiment of the present disclosure;

[0023] Figure 3 is a waveform diagram of envelope compensation for a sine wave in an exemplary embodiment of the present disclosure;

[0024] Figure 4 is a flowchart of a method for processing pitch shifting of an audio signal provided by another exemplary embodiment of the present disclosure;

[0025] Figure 5a is a schematic diagram of a flow chart of determining a first audio signal in a pitch shift processing method of an audio signal provided by another exemplary embodiment of the present disclosure;

[0026] Figure 5b is another flowchart of determining a first audio signal in a pitch shift processing method for an audio signal provided by another exemplary embodiment of the present disclosure;

[0027] Figure 6 is a schematic flow chart of determining a first audio signal in a pitch shift processing method for an audio signal provided by another exemplary embodiment of the present disclosure;

[0028] Figure 7 is a comparative schematic diagram of the autocorrelation function values ​​before and after weighting provided by the present disclosure;

[0029] Figure 8 is a structural schematic diagram of a pitch shift processing device for an audio signal provided by an exemplary embodiment of the present disclosure;

[0030] Figure 9a is a structural schematic diagram of a signal synthesis module in a pitch shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure;

[0031] Figure 9b is a structural schematic diagram of an interpolation processing module in a pitch shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure;

[0032] Fig.9c is a structural schematic diagram of a pitch shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure;

[0033] Fig.10 is a structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0034] Below, the exemplary embodiments according to the present disclosure will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure, and it should be understood that the present disclosure is not limited to the exemplary embodiments described here.

[0035] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless specifically stated otherwise.

[0036] Those skilled in the art can understand that the terms "first" and "second" in the embodiments of the present disclosure are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor indicate the necessary logical order between them.

[0037] It should also be understood that in the embodiments of the present disclosure, “plurality” may refer to two or more than two, and “at least one” may refer to one, two, or more than two.

[0038] It should also be understood that any component, data or structure mentioned in the embodiments of the present disclosure can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0039] In addition, the term "and / or" in the present disclosure is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present disclosure generally indicates that the associated objects before and after are in an "or" relationship.

[0040] It should also be understood that the description of the various embodiments in the present disclosure focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0041] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0042] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0043] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0044] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0045] The disclosed embodiments can be applied to electronic devices such as terminal devices, computer systems, servers, etc., which can operate with many other general or special computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, servers, etc. include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0046] Electronic devices such as terminal devices, computer systems, servers, etc. can be described in the general context of computer system executable instructions (such as program modules) executed by computer systems. Generally, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules can be located on local or remote computing system storage media including storage devices.

[0047] Application Overview

[0048] In the process of implementing the present disclosure, the inventors found that the existing audio pitch shifting processing scheme is usually based on the overlap-add method of time domain processing, such as: synchronous waveform superposition method, waveform similarity overlap-add method. However, when implemented, the above two types of methods have a large delay, for example, the delay exceeds 20ms, which reduces the user experience in practical applications.

[0049] In the phase vocoder method and the overlap-add method, the overlap-add method has a lower frame length requirement, so the delay is lower. Its algorithm is based on the audio frame, and its inherent delay is frame overlap = frame length - frame shift (synthesis step length). In addition, considering the search target waveform module in the algorithm, the offset is calculated based on the autocorrelation function, which will bring additional O m Delay, so the total delay of the algorithm can be expressed as: m +KH s =O m +α*L; where L is the frame length, H s is the frame shift, α is the frame overlap rate, and the frame overlap rate can be set to a fixed value.

[0050] Exemplary Methods

[0051] Figure 1FIG. 1 is a flow chart of a method for processing a pitch change of an audio signal provided by an exemplary embodiment of the present disclosure. This embodiment can be applied to electronic devices, such as Figure 1 As shown, the following steps are included:

[0052] Step 102: Process the first audio signal according to a first window function to obtain a second audio signal.

[0053] The first audio signal is a current frame audio signal. Optionally, processing the first audio signal may be windowing the first audio signal according to a first window function. Windowing is to multiply the first audio signal by the first window function, and the windowing is to make the fade-in and fade-out and smooth transition between time domain signals so that the signal is more continuous globally.

[0054] Optionally, the first window function may include but is not limited to: a Hanning window function, a Hamming window function, etc.

[0055] In some optional examples, the windowing process of the first audio signal may be implemented based on the following formula (1):

[0056] ccc(n)=hanning_L(n)*x 1 (n), n = 1, 2, ... L formula (1)

[0057] Where n represents the coordinates of the window function; hanning_L represents the Hanning window function at point L; x 1 is the first audio signal, and ccc(n) represents the amplitude corresponding to each window function coordinate in the second audio signal.

[0058] Optionally, the first audio signal is a signal segment that is most similar to a target signal (for example, the target signal is an audio signal output after audio pitch shifting of a previous frame of audio signal) in waveform, which is found in the audio signal received based on the current frame. The first audio signal can be determined by the similarity with the target waveform, for example, by identifying and matching an audio signal with a high similarity to the target waveform from the received current frame of audio.

[0059] Step 104: Perform a first interpolation process on the second audio signal and the first window function according to a preset modulation multiple, and obtain a third audio signal and a second window function accordingly.

[0060] Optionally, the second audio signal and the first window function can be interpolated based on the same interpolation processing method, and the information missing or redundancy problem caused by the lengthening or shortening of the signal duration during the pitch change processing can be compensated by the interpolation processing. The interpolation processing can adopt a commonly used interpolation processing method such as a linear interpolation processing method or a nonlinear interpolation processing method, and the information missing in the second audio signal or the information redundancy in the second audio signal can be compensated by the interpolation processing, and in order to ensure that the window function is consistent with the audio signal, the window function is also subjected to the corresponding first interpolation processing.

[0061] Step 106: Obtain a fourth audio signal according to the third audio signal, the second window function and the historical frame data.

[0062] Optionally, the fourth audio signal is an audio signal obtained by performing modulation processing on the current frame audio signal by a preset modulation multiple.

[0063] In this embodiment, the historical frame data may include a frame of audio signal that is located before the current frame of audio signal when the audio signal is framed; or, it may also include a window function for windowing the frame of audio signal; or, the historical frame data includes an audio signal after windowing and interpolation processing of the previous frame of audio signal, and a window function after interpolation processing of the previous window function for windowing the previous frame of audio signal. Optionally, the third audio signal is added to the audio signal in the historical frame data to obtain an audio signal with a certain overlapping portion; and the second window function is added to the window function in the historical frame data; and the result of adding the audio signals is processed based on the result of the addition of the window functions, so that the overlapping portion in the fourth audio signal obtained can be reduced.

[0064] The above-mentioned embodiment of the present disclosure provides a method for processing the pitch shift of an audio signal. The method performs windowing processing on a second audio signal to be synthesized through a first window function, so as to facilitate subsequent signal synthesis processing based on the window function. In addition, a first interpolation processing is performed on the second audio signal and the first window function to compensate for the information difference caused by the pitch shift by a preset pitch shift multiple. The third audio signal after the first interpolation processing and the second window function are synthesized with the historical frame data. Due to the existence of the window function envelope compensation, the overlapping part in the third audio signal is reduced, so that the delay of the obtained fourth audio signal is reduced while meeting the preset pitch shift multiple.

[0065] In some optional embodiments, step 104 may include:

[0066] According to the preset modulation multiple, scaling processing is performed on the second audio signal in the time domain to obtain a sixth audio signal.

[0067] Optionally, the first audio signal is an audio signal that has been processed (stretched or compressed) by a preset modulation multiple. For example, the preset modulation multiple is 0.5. At this time, the first audio signal is stretched relative to before the modulation, and the time interval between every two amplitude points in the audio changes. Direct playback will cause distortion. The corresponding second audio signal also has this problem. Therefore, the second audio signal is interpolated to compensate for the distortion of the audio signal after the modulation.

[0068] A first interpolation process is performed on the sixth audio signal to obtain a third audio signal.

[0069] Optionally, the first interpolation processing may include linear interpolation processing or nonlinear interpolation processing. Optionally, the linear interpolation processing may include but is not limited to: first-order linear interpolation, second-order linear interpolation, bicubic interpolation, etc.; the nonlinear interpolation processing may include but is not limited to: edge diffusion or edge banding, Lagrange interpolation, etc. In this embodiment, the preset modulation multiple is a user input or a preset modulation multiple received when the audio signal is modulated. The preset modulation multiple can be any multiple, for example, 0.5, 0.8, 2, 3, etc.

[0070] In some optional examples, taking a linear interpolation method as an example, the interpolation processing of the sixth audio signal may be implemented based on the following formula (2):

[0071]

[0072] Among them, x 2 is the sixth audio signal; x 3 is the third audio signal; β is the interpolation factor, which represents the interpolated signal (x 3 ) relative to the pre-interpolation signal (x 2 ), in this embodiment, it can be determined based on the modulation multiple, for example, the interpolation factor is equal to the modulation multiple; n is the coordinate of the second audio signal; floor is the floor rounding function; ceil is the ceiling rounding function; L′ represents the length of the third audio signal after interpolation processing.

[0073] In some optional embodiments, step 104 may further include:

[0074] According to the preset modulation multiple, scaling processing is performed on the first window function in the time domain to obtain a fourth window function.

[0075] A first interpolation process is performed on the fourth window function to obtain a second window function.

[0076] In this embodiment, the processing of the first window function can refer to the processing of the second audio signal. First, the waveform corresponding to the first window function is scaled in the time domain (the scaling factor corresponds to the preset modulation factor), and the obtained fourth window function is linearly interpolated to obtain the second window function. The specific interpolation process can refer to the above formula (2) and only needs to replace the sixth audio signal with the fourth window function and the third audio signal with the second window function. The corresponding L' represents the length of the second window function after the interpolation process. It should be understood that the above formula is only used to understand the specific implementation of an interpolation method, and does not limit the first interpolation processing method for the sixth audio signal in the present disclosure. The existing and later disclosed technical solutions that can be used for interpolation processing can be used as the first interpolation processing.

[0077] like Figure 2 As shown in the above Figure 1 Based on the embodiment shown, the historical frame data includes a first historical frame audio signal and a first historical frame window function corresponding to the first historical frame audio signal; step 106 may include the following steps:

[0078] Step 1061: Obtain a fifth audio signal according to the third audio signal and the first historical frame audio signal.

[0079] The first historical frame audio signal is an audio signal outputted in a frame before the current frame signal. Optionally, the first historical frame audio signal is also subjected to windowing processing and interpolation processing by a window function like the third audio signal.

[0080] Optionally, since the cost of windowing is that the two ends of a frame signal are weakened, there needs to be a partial overlap when two consecutive frames of audio signals are connected. Therefore, in this embodiment, the third audio signal is overlapped and added with the first historical frame audio signal. Optionally, during the overlap-addition process, the third audio signal and the first historical frame audio signal are merged based on the overlapping part, and the overlapping part usually accounts for a part of the audio signal, for example, 20%, 30%, etc. At this time, there are corresponding 80% or 70% non-overlapping parts in the third audio signal and the first historical frame audio signal, respectively. Therefore, the length of the fifth audio signal is not the sum of the lengths of the third audio signal and the first historical frame audio signal, but the sum of the two minus the length of the overlapping part.

[0081] Step 1062: Obtain a third window function according to the second window function and the second historical frame window function.

[0082] The second historical frame window function is a window function obtained by performing a second interpolation process on the first historical frame window function.

[0083] Optionally, the second interpolation process may be the same as or different from the first interpolation process; usually, in order to ensure the consistency of continuous frame signals, the second interpolation process and the first interpolation process adopt the same interpolation method.

[0084] In this embodiment, the processing of the second window function and the second historical frame window function is similar to that of the audio signal, and can also be overlap-addition.

[0085] Step 1063: Obtain a fourth audio signal according to the fifth audio signal and the third window function.

[0086] In this embodiment, compensation processing is performed on the fifth audio signal based on the third window function to overcome the fluctuation and distortion of the signal caused by the audio signal in the overlap-add process. Optionally, the compensation processing may be envelope compensation processing. Envelope compensation processing is performed on the fifth audio signal based on the third window function to obtain a fourth audio signal.

[0087] In this embodiment, the fluctuation and distortion generated when the signal is superimposed are solved by envelope compensation based on the window function, the requirement of the perfect reconstruction condition for the frame overlap rate is changed, and the requirement of the frame overlap rate of 50% is lifted. The frame overlap rate can be reduced to 20% or lower, that is, the requirement of frame overlap is reduced, thereby reducing the delay, for example, reducing the delay by more than 15ms. The perfect reconstruction condition in this embodiment is the perfect reconstruction condition of windowing. In signal analysis and synthesis, the perfect reconstruction condition of windowing is the COLA (Constant Overlap-Add) criterion. The COLA criterion requires that the sum of the squares of the window of the part of the overlapped buffer that the windowed signal meets must be 1, so that the analysis and synthesis are free of damage and distortion to the signal. This means that the sum of the squares of the analysis window and the synthesis window should be equal to ensure perfect reconstruction. To meet the perfect reconstruction condition, the frame overlap rate is usually required to be 50%. In this embodiment, envelope compensation based on the window function is used to achieve the perfect reconstruction condition at a smaller frame overlap rate (for example, 20%).

[0088] In some optional embodiments, the fifth audio signal and the third window function are both time domain signals.

[0089] Step 1063 may include:

[0090] Based on the first amplitude at each time point in the fifth audio signal and the second amplitude corresponding to each time point in the third window function, the third amplitude corresponding to each time point of the fourth audio signal is determined.

[0091] A fourth audio signal is obtained based on the third amplitude corresponding to each time point.

[0092] In this embodiment, the envelope compensation process may be to divide the amplitude of the fifth audio signal at each time point by the amplitude corresponding to the third window function at the time point to determine the third amplitude after envelope compensation at the time point.

[0093] Optionally, the third amplitude may be determined based on the following formula (3):

[0094]

[0095] Among them, y m+1 represents the fourth audio signal corresponding to the m+1th frame; y′ m+1 represents the fifth audio signal corresponding to the m+1th frame; w m+1 is the third window function corresponding to the m+1th time point; i represents each time point (for example, Figure 3 In some optional examples, a sine wave is used as an example to show the effect of envelope compensation. Figure 3 As shown, before compensation (the waveform represented by the solid line in the figure), the envelope of the sine wave fluctuates and is distorted due to the influence of frame overlap. After compensation, the envelope of the sine wave (the waveform represented by the dotted line in the figure) returns to a constant state.

[0096] Figure 4 FIG. 1 is a flow chart of a method for processing pitch change of an audio signal provided by another exemplary embodiment of the present disclosure. Figure 4 As shown, the following steps are included:

[0097] Step 402: Obtain a first audio signal according to the seventh audio signal and the first autocorrelation process.

[0098] Optionally, the seventh audio signal is a signal for which a waveform similar to that of the seventh audio signal is desired to be found. For example, the seventh audio signal is an audio signal after the audio signal of the previous frame of audio signal is output after the audio pitch shifting process is performed and the audio signal is moved by a synthesis step length (the length of the audio signal for each audio pitch shifting process); optionally, an audio signal obtained by moving the synthesis step length forward based on the first historical frame of audio signal is used as the seventh audio signal; in some optional examples, the synthesis step length is equal to the product of the analysis step length (which is a preset value) and a preset pitch shift multiple. The first autocorrelation process can be based on the autocorrelation function to determine an audio signal segment with a high degree of similarity to the seventh audio signal within a certain range after the seventh audio signal as the first audio signal.

[0099] Step 404: Obtain a fourth audio signal based on the first audio signal and the historical frame data.

[0100] In this embodiment, after determining the first audio signal, the first audio signal can be fused with the historical frame data using any signal fusion method to obtain a fourth audio signal. In some optional examples, based on a preset overlap rate, the historical audio signal in the historical frame data is overlapped with the first audio signal at a preset overlap rate; for example, the preset overlap rate is 50%, and 50% of the historical audio signal and the first audio signal are overlapped to obtain a fourth audio signal having a length of 3 / 2 of the first audio signal. In some other optional examples, the historical audio signal in the historical frame data is directly connected to the first audio signal to obtain a fourth audio signal having a length equal to the sum of the lengths of the historical audio signal and the first audio signal. In some further optional examples, the first audio signal can be processed with reference to the audio signal processing method provided in any of the above embodiments to obtain the fourth audio signal outputted by the current frame. For example, with the above Figure 1 Step 102, step 104 and step 106 in the provided method for processing pitch shift of an audio signal determine a fourth audio signal.

[0101] like Figure 5a As shown in the above Figure 4 Based on the illustrated embodiment, the first autocorrelation processing is weighted autocorrelation processing; step 402 may include the following steps:

[0102] Step 4021: Determine a corresponding eighth audio signal based on the seventh audio signal.

[0103] Optionally, based on searching the seventh audio signal forward within the range of the current frame audio signal, determine the range of audio segments that have a high similarity to the seventh audio signal, for example, the last time point of the current frame audio signal is the end point, and an audio segment of length L+T is obtained in the reverse direction of time from the end point as the eighth audio signal, where L is the length of one frame of audio signal and T is the search length. It can be understood that the eighth audio signal is an audio signal after the seventh audio signal, and by determining the eighth audio signal, the audio matching range is narrowed, thereby improving the efficiency of determining the first audio signal.

[0104] Step 4022: Perform weighted autocorrelation processing based on the eighth audio signal and the seventh audio signal to obtain a first audio signal.

[0105] In this embodiment, an audio segment (for example, a ninth audio signal in the following embodiments) having a high similarity to the seventh audio signal and a small time difference with the seventh audio signal is determined from the eighth audio signal based on weighted autocorrelation as the first audio signal, wherein the weight corresponds to the time difference between the audio signal and the seventh audio signal, and the larger the time difference, the smaller the corresponding weight value, and the smaller the time difference, the larger the corresponding weight value. That is, this embodiment expects to obtain an audio segment having a small time difference and a large similarity with the seventh audio signal, thereby reducing the delay on the basis of achieving better waveform superposition.

[0106] Optionally, step 4022 may include:

[0107] A plurality of ninth audio signals of preset lengths are determined based on the eighth audio signal.

[0108] Optionally, the eighth audio signal is divided into a ninth audio signal having the same time length as the seventh audio signal, and a division method may be to take each time point in the eighth audio signal as a starting point, and obtain an audio segment of a corresponding time length backward from the starting point as the ninth audio signal.

[0109] A plurality of first similarity values ​​between the plurality of ninth audio signals and the seventh audio signal are determined.

[0110] Optionally, each first similarity value reflects the weighted similarity between each ninth audio signal and the seventh audio signal. The larger the value, the more similar the waveforms of the two audios are. The first similarity value is obtained based on a specific calculation and is usually expressed as a decimal less than 1, for example, 0.9, 0.8, etc.

[0111] In this embodiment, the first similarity value may be a weighted similarity value. Optionally, the second similarity value is determined based on an autocorrelation function value between each ninth audio signal and the seventh audio signal.

[0112] Optionally, each second similarity value represents the autocorrelation similarity between each ninth audio signal and the seventh audio signal. The larger the value, the more similar the waveforms of the two audios are. The second similarity value is obtained based on solving the autocorrelation function and is usually expressed as a decimal less than 1, for example, 0.7, 0.8, etc.

[0113] In order to determine the autocorrelation similarity between the two audio signals, it can be determined based on the autocorrelation function calculation. In order to better identify the similarity between the two signals (which can be understood as the similarity of the time domain waveforms corresponding to the signals), the two signals can be sampled first (one sampling point is obtained at each preset number of time points, and the preset number can be determined according to the specific application scenario), and multiple first sampling points and multiple second sampling points are obtained respectively; for example, the second similarity value corresponding to each ninth audio signal can be determined based on the following formula (4):

[0114]

[0115] Wherein, Cor[n] represents the second similarity; n is the autocorrelation function coordinate. In this embodiment, different values ​​of n correspond to different ninth audio signals; T represents the number of ninth audio signals included in the eighth audio signal (corresponding to the search length); L represents the number of sampling points included in the seventh audio signal; x t represents the eighth audio signal; x rrepresents the seventh audio signal; i represents each sampling point in the seventh audio signal. In this formula, the autocorrelation similarity between each ninth audio signal included in the eighth audio signal and the seventh audio signal is directly calculated. It is also possible not to sample the audio signal, and the autocorrelation similarity between each ninth audio signal included in the eighth audio signal and the seventh audio signal is calculated by the above formula. In this case, i in the formula represents each time point in the seventh audio signal, and L represents the length of the seventh audio signal.

[0116] Optionally, the first similarity corresponding to each ninth audio signal is determined based on the product of the first weight value and the second similarity. Optionally, a corresponding first weight value may be preset for each ninth audio signal.

[0117] Based on the plurality of first similarity values, a first audio signal is determined.

[0118] Optionally, a ninth audio signal corresponding to a first similarity value having the largest value is obtained from a plurality of first similarity values ​​as the first audio signal. The first similarity obtained in this embodiment is combined with a first weight value corresponding to each ninth audio signal, so that signal matching is performed not only by similarity but also by combining the first weight value. When the first weight value is related to time, the delay in synthesizing the audio signal can be reduced.

[0119] Optionally, obtaining a plurality of first similarity values ​​based on the plurality of second similarities and a first weight value corresponding to each ninth audio signal includes:

[0120] Based on an interval between each ninth audio signal and the seventh audio signal, a first weight value corresponding to each ninth audio signal is determined.

[0121] Optionally, the weight value corresponding to each sampling point may be determined based on the following formula (5):

[0122]

[0123] Wherein, T represents the time length of the eighth audio signal; n is the autocorrelation function coordinate. Based on the above formula, it can be known that the autocorrelation function value with n close to 0 obtains a greater gain, and the autocorrelation function value with n close to T obtains a smaller gain; the smaller n is, the better, and the larger the weighted similarity value is, that is, this embodiment expects to obtain a ninth audio signal that is closer to the seventh audio signal in time through the weight value, so as to achieve a balance between time and similarity.

[0124] Based on the plurality of second similarities and the corresponding first weight values, a plurality of first similarity values ​​are determined.

[0125] In this embodiment, each second similarity is multiplied by the corresponding first weight value to determine the first similarity value by the product.

[0126] like Figure 5b As shown in the above Figure 4 Based on the illustrated embodiment, the first autocorrelation processing is autocorrelation processing; step 402 may include the following steps:

[0127] Step 4021: Determine a corresponding eighth audio signal based on the seventh audio signal.

[0128] The determination process of the eighth audio signal in this embodiment can refer to Figure 5a The embodiments shown will not be described in detail here.

[0129] Step 4023: Perform autocorrelation processing based on the eighth audio signal and the seventh audio signal to obtain a first audio signal.

[0130] In this embodiment, an audio segment (for example, a ninth audio signal) having a higher similarity to the seventh audio signal is determined from the eighth audio signal based on autocorrelation as the first audio signal.

[0131] Optionally, step 4023 may include:

[0132] A plurality of ninth audio signals of preset lengths are determined based on the eighth audio signal.

[0133] Determine a plurality of second similarity values ​​between the plurality of ninth audio signals and the seventh audio signal. Optionally, the second similarity value is determined based on an autocorrelation function value between each ninth audio signal and the seventh audio signal. For example, after sampling, the autocorrelation similarity between each ninth audio signal in the eighth audio signal and the seventh audio signal is calculated based on the above formula (4); and the second similarity value corresponding to each ninth audio signal is obtained.

[0134] Based on the plurality of second similarity values, a first audio signal is determined.

[0135] Optionally, a ninth audio signal corresponding to a second similarity value having the largest value is obtained from the plurality of second similarity values ​​as the first audio signal. This embodiment implements signal matching based on similarity, and can quickly determine the first audio signal.

[0136] Figure 6 FIG. 1 is a flow chart of determining a first audio signal in a method for processing a pitch change of an audio signal provided by another exemplary embodiment of the present disclosure. Figure 6 As shown, the input audio signal x (including multiple frames of audio signal), that is Figure 6The gray bar in the upper middle part indicates that the input audio signal can be any sound signal, such as a human voice signal, a music audio signal, etc. The output audio signal y (which is the audio signal after the pitch change processing, corresponding to the fourth audio signal in the above embodiment) is shown in FIG. 1 . Assume that the audio signal of the (m+1)th frame is currently being processed, and the audio signal of the mth frame of audio signal after the pitch change processing (corresponding to the first historical frame of audio signal in the above embodiment) is as follows: Figure 6 As shown in the red dotted box. In this embodiment, the synthesis step length is H s (In this embodiment, the synthesis step length = analysis step length H a (preset value)×preset pitch shift multiple), based on the m-th frame audio signal and the synthesis step length, the synthesis step length is moved in the positive direction of time to determine the corresponding target audio signal (corresponding to the seventh audio signal in the above embodiment) as Figure 6 The corresponding audio signal search range (corresponding to the eighth audio signal in the above embodiment) is determined according to the target audio signal. The audio signal search range is (m+1)H a is the end point, where H a is the analysis step length (its value is a preset empirical value). From the end point to the reverse direction of time, an audio segment of length L+T is obtained as the audio signal search range, where L is the length of one frame of audio signal and T is the search length. The obtained audio signal search range is as follows: Figure 6 Shown in the blue dotted box.

[0137] Based on the above formula (4), the second similarity between each signal segment with the same length as the target audio signal in the audio signal search range and the target audio signal can be calculated and determined.

[0138] Combining the above formula (4) and formula (5), the following formula (6) can be obtained, which is used to reflect the first similarity of each signal segment in the audio signal search range:

[0139]

[0140] Among them, Cor w [n] represents the first similarity of each signal segment; other parameters are understood by referring to the above formulas (4) and (5), that is, the first similarity is determined by combining the second similarity with the weight. By weighting the autocorrelation function, the autocorrelation function value closer to 0 obtains a greater gain (the larger the weight value), and the autocorrelation function value closer to T obtains a smaller gain (the smaller the weight value).

[0141] Based on the above formula (6), the coordinates corresponding to the maximum value of the weighted autocorrelation function are calculated using the following formula (7):

[0142] O m+1 =max n Corw [n]Formula (7)

[0143] Among them, O m+1 Indicates the first audio signal and the rightmost boundary of the audio signal search range ((m+1)H a ) offset, such as Figure 6 The offset between the right edge of the red solid line frame and the right edge of the blue dotted line frame; based on O m+1 Get an audio segment of the same length as the target audio signal in the reverse direction of time (such as Figure 6 The red solid line frame in the middle) is used as the first audio signal.

[0144] The autocorrelation function values ​​before and after weighting (the second similarity value before weighting is shown as the blue curve in the figure, and the first similarity value after weighting is shown as the orange dotted curve in the figure) are as follows Figure 7 As shown in (the horizontal axis is the offset, and the vertical axis is the value of measuring the similarity of the two signal waveforms), in the voiced segment of human voice, the autocorrelation function value before weighting often has multiple local maxima (such as Figure 7 ), and after weighting them, the algorithm tends to find similar waveforms with smaller offsets. Figure 7 In the example, the unweighted autocorrelation function is used to determine the first audio signal based on the third peak of the blue curve, and the weighted autocorrelation function can be used to determine the first audio signal based on the first peak of the orange dashed curve; the first peak and the third peak are separated by nearly 200 sampling points (250-50), so the current frame delay is reduced by nearly 200 sampling points, and at a sampling rate of 16k, the delay is reduced by more than 10ms. For the entire signal, it can be guaranteed that the delay of each frame is reduced or unchanged, and the specific reduction value depends on the shape of the waveform of each frame.

[0145] Any audio signal pitch change processing method provided by the embodiments of the present disclosure can be executed by any appropriate device with data processing capabilities, including but not limited to: terminal devices and servers, etc. Alternatively, any audio signal pitch change processing method provided by the embodiments of the present disclosure can be executed by a processor, such as the processor executing any audio signal pitch change processing method mentioned in the embodiments of the present disclosure by calling corresponding instructions stored in a memory. No further details will be given below.

[0146] Exemplary Devices

[0147] Figure 8 FIG. 1 is a schematic diagram of a structure of a pitch-shift processing device for an audio signal provided by an exemplary embodiment of the present disclosure. Figure 8 As shown, the device provided in this embodiment includes:

[0148] The windowing processing module 81 is used to process the first audio signal according to a first window function to obtain a second audio signal.

[0149] The first audio signal is a current frame audio signal.

[0150] The interpolation processing module 82 is used to perform a first interpolation process on the second audio signal and the first window function according to a preset modulation multiple, so as to obtain a third audio signal and a second window function accordingly.

[0151] The signal synthesis module 83 is used to obtain a fourth audio signal according to the third audio signal, the second window function and the historical frame data.

[0152] Optionally, after obtaining the fourth audio signal after the tone shifting processing, the fourth audio signal can be played through an audio playing device (for example, a sound box, a speaker, etc.) to achieve output of the tone shifting audio signal.

[0153] The above-mentioned embodiment of the present disclosure provides a pitch shift processing device for an audio signal, which performs windowing processing on a second audio signal to be synthesized through a first window function, so as to facilitate subsequent signal synthesis processing based on the window function; in addition, a first interpolation processing is performed on the second audio signal and the first window function, so as to compensate for the information difference caused by the pitch shifting by a preset pitch shift multiple, and a third audio signal after the first interpolation processing is synthesized with the second window function and the historical frame data. Due to the existence of the window function envelope compensation, the overlapping part in the third audio signal is reduced, so that the delay of the obtained fourth audio signal is reduced while meeting the preset pitch shift multiple.

[0154] Figure 9a FIG. 1 is a schematic diagram of the structure of a signal synthesis module in a pitch-shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure. Figure 9a As shown, the signal synthesis module 83 includes:

[0155] The audio processing unit 831 is configured to obtain a fifth audio signal according to the third audio signal and the first historical frame audio signal.

[0156] The historical frame data includes a first historical frame audio signal and a first historical frame window function corresponding to the first historical frame audio signal. The first historical frame audio signal is an audio signal outputted in a previous frame of the current frame signal.

[0157] The function processing unit 832 is used to obtain a third window function according to the second window function and the second historical frame window function; the second historical frame window function is a window function obtained by performing a second interpolation process on the first historical frame window function.

[0158] The synthesis processing unit 833 is configured to obtain a fourth audio signal according to the fifth audio signal and the third window function.

[0159] Optionally, the synthesis processing unit 833 is specifically configured to perform envelope compensation processing on the fifth audio signal based on the third window function to obtain a fourth audio signal.

[0160] Wherein, the fifth audio signal and the third window function are both time domain signals.

[0161] The synthesis processing unit 833 is specifically used to determine the third amplitude corresponding to the fourth audio signal at each time point based on the first amplitude at each time point in the fifth audio signal and the second amplitude corresponding to each time point in the third window function; and obtain the fourth audio signal based on the third amplitude corresponding to each time point.

[0162] Figure 9b FIG. 1 is a schematic diagram of the structure of an interpolation processing module in a pitch-shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure. Figure 9b As shown, the interpolation processing module 82 includes:

[0163] The signal scaling unit 821 is configured to perform scaling processing on the second audio signal in the time domain according to a preset modulation multiple to obtain a sixth audio signal.

[0164] The signal interpolation unit 822 is configured to perform a first interpolation process on the sixth audio signal to obtain a third audio signal.

[0165] The function scaling unit 823 is used to perform the scaling process on the first window function in the time domain according to a preset modulation multiple to obtain a fourth window function.

[0166] The function interpolation unit 824 is configured to perform a first interpolation process on the fourth window function to obtain a second window function.

[0167] Fig.9c FIG. 2 is a schematic diagram of a structure of a pitch-shift processing device for an audio signal provided by another exemplary embodiment of the present disclosure. Fig.9c As shown, the device provided in this embodiment also includes:

[0168] The autocorrelation processing module 91 is used to obtain the first audio signal according to the seventh audio signal and the first autocorrelation processing.

[0169] In this embodiment, the seventh audio signal is determined by offsetting the pitch-shifted audio signal output in the previous frame.

[0170] In some optional embodiments, the first autocorrelation processing in the autocorrelation processing module 91 is a weighted autocorrelation processing; the autocorrelation processing module 91 includes:

[0171] a signal searching unit, configured to determine a corresponding eighth audio signal based on the seventh audio signal;

[0172] The weighted autocorrelation unit is used to perform weighted autocorrelation processing based on the eighth audio signal and the seventh audio signal to obtain the first audio signal.

[0173] The first audio signal obtained through weighted autocorrelation processing achieves a balance between the similarity and the offset with the seventh audio signal, and an audio signal with greater similarity and smaller offset is obtained as the first audio signal.

[0174] Optionally, a weighted autocorrelation unit comprises:

[0175] an audio segmentation unit, configured to determine a plurality of ninth audio signals of preset lengths based on the eighth audio signal;

[0176] a similarity determination unit, configured to determine a plurality of first similarity values ​​between a plurality of ninth audio signals and a seventh audio signal;

[0177] The signal screening unit is used to determine the first audio signal based on multiple first similarity values.

[0178] Optionally, the similarity determination unit includes:

[0179] An autocorrelation similarity unit, configured to determine a second similarity value based on an autocorrelation function value between each ninth audio signal and the seventh audio signal;

[0180] The weighted processing unit is used to obtain multiple first similarity values ​​based on the multiple second similarity values ​​and the first weight value corresponding to each ninth audio signal. The weighted processing unit is specifically used to determine the first weight value corresponding to each ninth audio signal based on the interval between each ninth audio signal and the seventh audio signal; and determine the multiple first similarity values ​​based on the multiple second similarities and the corresponding first weight values.

[0181] In some other optional embodiments, the first autocorrelation processing in the autocorrelation processing module 91 is autocorrelation processing; the autocorrelation processing module 91 includes:

[0182] a signal searching unit, configured to determine a corresponding eighth audio signal based on the seventh audio signal;

[0183] The autocorrelation unit is configured to perform autocorrelation processing based on the eighth audio signal and the seventh audio signal to obtain a first audio signal.

[0184] The autocorrelation unit is specifically configured to determine a plurality of ninth audio signals of preset lengths based on the eighth audio signal, determine a plurality of second similarity values ​​between the plurality of ninth audio signals and the seventh audio signal, and determine the first audio signal based on the plurality of second similarity values.

[0185] The beneficial technical effects corresponding to the exemplary embodiment of the present device can be found in the corresponding beneficial technical effects of the above exemplary method section, which will not be repeated here.

[0186] Exemplary Electronic Devices

[0187] Fig.10 A structural diagram of an electronic device provided in an embodiment of the present disclosure includes at least one processor 11 and a memory 12.

[0188] The processor 11 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

[0189] The memory 12 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may run one or more computer program instructions to implement the pitch shift processing method of the audio signal of each embodiment of the present disclosure above and / or other desired functions.

[0190] In one example, the electronic device 10 may further include: an input device 13 and an output device 14, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0191] The input device 13 may also include, for example, a keyboard, a mouse, a touch screen, a sound pickup device (such as a microphone array), etc.

[0192] The output device 14 can output various information to the outside, and may include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, and the like.

[0193] Of course, to simplify, Fig.10 Only some of the components related to the present disclosure in the electronic device 10 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application situations, the electronic device 10 may also include any other appropriate components.

[0194] Exemplary computer program products and computer-readable storage media

[0195] In addition to the above-mentioned methods and devices, embodiments of the present disclosure may also provide a computer program product, including computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the pitch shifting processing method of the audio signal of various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section.

[0196] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the disclosed embodiments, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0197] In addition, an embodiment of the present disclosure may also be a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the steps of the pitch shift processing method of the audio signal of various embodiments of the present disclosure described in the above-mentioned "Exemplary Method" section.

[0198] Computer readable storage media can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium is, for example, but not limited to, a system, device or device including electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0199] The basic principles of the present disclosure are described above in conjunction with specific embodiments. However, the advantages, strengths, effects, etc. mentioned in the present disclosure are only examples and not limitations, and cannot be considered as necessary for each embodiment of the present disclosure. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, rather than limitation, and the above details do not limit the present disclosure to being implemented by adopting the above specific details.

[0200] Those skilled in the art may make various changes and modifications to the present disclosure without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present disclosure claims and their equivalents, the present disclosure is also intended to include these modifications and variations.

Claims

1. A method for processing pitch change of an audio signal, comprising: Processing a first audio signal according to a first window function to obtain a second audio signal, wherein the first audio signal is a current frame audio signal; According to a preset modulation multiple, performing a first interpolation process on the second audio signal and the first window function respectively, and correspondingly obtaining a third audio signal and a second window function; A fourth audio signal is obtained according to the third audio signal, the second window function and the historical frame data.

2. The method according to claim 1, wherein the historical frame data comprises a first historical frame audio signal and a first historical frame window function corresponding to the first historical frame audio signal; The step of obtaining a fourth audio signal according to the third audio signal, the second window function and the historical frame data comprises: Obtaining a fifth audio signal according to the third audio signal and the first historical frame audio signal; The first historical frame audio signal is an audio signal outputted in a frame before the current frame signal; A third window function is obtained according to the second window function and the second historical frame window function; the second historical frame window function is a window function obtained by performing a second interpolation process on the first historical frame window function; The fourth audio signal is obtained according to the fifth audio signal and the third window function.

3. The method according to claim 2, wherein obtaining the fourth audio signal according to the fifth audio signal and the third window function comprises: Perform envelope compensation processing on the fifth audio signal based on the third window function to obtain the fourth audio signal.

4. The method according to claim 3, wherein the fifth audio signal and the third window function are both time domain signals; The performing envelope compensation processing on the fifth audio signal based on the third window function to obtain the fourth audio signal includes: Determine a third amplitude of the fourth audio signal at each time point based on a first amplitude at each time point in the fifth audio signal and a second amplitude corresponding to each time point in the third window function; The fourth audio signal is obtained based on the third amplitude corresponding to each of the time points.

5. The method according to any one of claims 1 to 4, wherein the step of performing a first interpolation process on the second audio signal according to a preset modulation multiple to obtain a third audio signal comprises: According to the preset modulation multiple, scaling processing is performed on the second audio signal in the time domain to obtain a sixth audio signal; The first interpolation process is performed on the sixth audio signal to obtain the third audio signal.

6. The method according to claim 5, wherein the first interpolation process is performed on the first window function according to the preset modulation multiple to obtain the second window function, comprising: According to the preset modulation multiple, performing the scaling process on the first window function in the time domain to obtain a fourth window function; A first interpolation process is performed on the fourth window function to obtain the second window function.

7. The method according to any one of claims 1 to 6, further comprising: The first audio signal is obtained by performing processing based on the seventh audio signal and the first autocorrelation.

8. The method according to claim 7, wherein the first autocorrelation processing is a weighted autocorrelation processing; The step of obtaining the first audio signal according to the seventh audio signal and the first autocorrelation processing includes: determining a corresponding eighth audio signal based on the seventh audio signal; A weighted autocorrelation process is performed based on the eighth audio signal and the seventh audio signal to obtain the first audio signal.

9. The method according to claim 8, wherein performing weighted autocorrelation processing based on the eighth audio signal and the seventh audio signal to obtain the first audio signal comprises: Determining a plurality of ninth audio signals of preset lengths based on the eighth audio signal; determining a plurality of first similarity values ​​between the plurality of the ninth audio signal and the seventh audio signal; Based on the plurality of first similarity values, the first audio signal is determined.

10. The method according to claim 9, wherein determining a plurality of first similarity values ​​between the plurality of the ninth audio signals and the seventh audio signal comprises: determining a second similarity value based on an autocorrelation function value between each of the ninth audio signal and the seventh audio signal; The multiple first similarity values ​​are obtained based on the multiple second similarity values ​​and the first weight value corresponding to each of the ninth audio signals.

11. The method according to claim 10, wherein obtaining the plurality of first similarity values ​​based on the plurality of second similarities and the first weight value corresponding to each of the ninth audio signals comprises: determining a first weight value corresponding to each of the ninth audio signals based on an interval between each of the ninth audio signals and the seventh audio signal; The plurality of first similarity values ​​are determined based on the plurality of second similarities and the corresponding first weight values.

12. A tone-shifting processing device for an audio signal, comprising: a windowing processing module, configured to process the first audio signal according to a first window function to obtain a second audio signal, wherein the first audio signal is a current frame audio signal; An interpolation processing module, configured to perform a first interpolation processing on the second audio signal and the first window function according to a preset modulation multiple, to obtain a third audio signal and a second window function respectively; A signal synthesis module is used to obtain a fourth audio signal according to the third audio signal, the second window function and the historical frame data.

13. A computer-readable storage medium storing a computer program, wherein the computer program is used to execute the pitch shift processing method for an audio signal according to any one of claims 1 to 11.

14. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the pitch shift processing method of the audio signal described in any one of claims 1-11.