WO2020211004A1 - Audio signal processing method and device, and storage medium - Google Patents

Audio signal processing method and device, and storage medium Download PDF

Info

Publication number
WO2020211004A1
WO2020211004A1 PCT/CN2019/083006 CN2019083006W WO2020211004A1 WO 2020211004 A1 WO2020211004 A1 WO 2020211004A1 CN 2019083006 W CN2019083006 W CN 2019083006W WO 2020211004 A1 WO2020211004 A1 WO 2020211004A1
Authority
WO
WIPO (PCT)
Prior art keywords
frequency domain
audio signal
domain data
data
channels
Prior art date
Application number
PCT/CN2019/083006
Other languages
French (fr)
Chinese (zh)
Inventor
莫品西
吴晟
边云锋
薛政
Original Assignee
深圳市大疆创新科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳市大疆创新科技有限公司 filed Critical 深圳市大疆创新科技有限公司
Priority to CN201980005583.8A priority Critical patent/CN111345047A/en
Priority to PCT/CN2019/083006 priority patent/WO2020211004A1/en
Publication of WO2020211004A1 publication Critical patent/WO2020211004A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R3/00—Circuits for transducers, loudspeakers or microphones

Definitions

  • the embodiments of the present invention relate to the field of signal processing, and in particular, to an audio signal processing method, device, and storage medium.
  • the dynamic range of a microphone refers to the ratio of the maximum sound and the minimum sound without distortion that the microphone can record.
  • the dynamic range of an ordinary microphone generally does not exceed 100dBA.
  • circuit noise and quantization noise usually there are slight circuit noise and quantization noise in the signal, and the effective dynamic range will be further compressed.
  • recording requires a higher dynamic range, and ordinary microphones cannot meet the demand.
  • the dynamic range of an ordinary microphone is still limited, and it is difficult to record both large and small signal sounds at the same time. If it meets the requirements for small voice recording, loud voice overload will occur; if it meets loud voice recording, then There will be a low signal-to-noise ratio problem for small sounds.
  • the second method two or more microphones covering different dynamic ranges are required to record, and finally only one sound signal is synthesized. The increase in the number of microphones and the selection, adjustment, and testing of multiple microphones significantly increase the cost.
  • the embodiments of the present invention provide an audio signal processing method, device, and storage medium, so as to effectively increase the dynamic range of the recording system.
  • the first aspect of the embodiments of the present invention is to provide an audio signal processing method, including:
  • a plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals.
  • Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
  • the frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  • the second aspect of the embodiments of the present invention is to provide an audio signal processing device, including: a memory and a processor;
  • the memory is used to store program codes
  • the processor calls the program code, and when the program code is executed, is used to perform the following operations:
  • a plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals.
  • Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
  • the frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  • the third aspect of the embodiments of the present invention is to provide a recording system, including:
  • Microphone used to collect analog audio signals
  • the fourth aspect of the embodiments of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the method described in the first aspect.
  • the audio signal processing method, device and storage medium provided in this embodiment use multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits is preprocessed
  • the processing circuit includes an amplifier and an analog-to-digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; One or at least two channels of target frequency domain data among the channels of frequency domain data determine frequency domain fusion data; the frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  • the method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor and meet the requirement of high signal-to-noise ratio.
  • FIG. 1 is a flowchart of an audio signal processing method provided by an embodiment of the present invention
  • FIG. 2 is a schematic diagram of multiple preprocessing circuits in parallel provided by an embodiment of the present invention
  • FIG. 3 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • FIG. 4 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • FIG. 5 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • FIG. 6 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • FIG. 7 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • FIG. 8 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • Figure 9 is a graph of the mapping function of the nonlinear mapping
  • FIG. 10 is a flowchart of an audio signal processing method provided by another embodiment of the present invention.
  • Figure 11a is a time-domain signal diagram of the digital audio signal of the first frame output by the first preprocessing circuit
  • 11b is a time-domain signal diagram of the digital audio signal of the first frame output by the second preprocessing circuit
  • FIG. 12a is a time-frequency spectrum diagram of the digital audio signal of the first frame output by the first preprocessing circuit
  • Figure 12b is a time-frequency spectrum diagram of the digital audio signal of the first frame output by the second preprocessing circuit
  • Figure 13a is a time-domain signal diagram of the output audio signal of the lth frame
  • Fig. 13b is a time-frequency spectrum diagram of the output audio signal of the first frame
  • Fig. 14 is a structural diagram of an audio signal processing device provided by an embodiment of the present invention.
  • a component When a component is considered to be "connected" to another component, it can be directly connected to another component or there may be a centered component at the same time.
  • FIG. 1 is a flowchart of an audio signal processing method provided by an embodiment of the present invention. As shown in FIG. 1, the audio signal processing method in this embodiment may include:
  • Step S101 Use multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog-to-digital converter, and each preprocessing circuit The analog gains of the amplifiers of the processing circuit are different.
  • each preprocessing circuit includes an amplifier and an analog-to-digital converter.
  • the multiple preprocessing circuits include at least two preprocessing circuits. Processing circuit; where the amplifier can be used to amplify the power of the analog audio signal, the amplification factor is usually expressed by gain, the analog gain of the amplifier of each preprocessing circuit in this embodiment is different, and further, the two adjacent analog gains The dynamic range of the pre-processing circuit at least partially overlaps; and the analog-to-digital converter can be used to convert analog audio signals into digital audio signals for subsequent signal processing.
  • the analog audio signal x to be processed can be collected by a microphone, and then input into a plurality of parallel preprocessing circuits, respectively, through power amplification of different gains, and converted into a digital audio signal x 1 ,...x i ...,x I , where I is the number of preprocessing circuits, that is, the input of each preprocessing circuit is the same analog audio signal to be processed, and the output of each preprocessing circuit is digital audio
  • I is the number of preprocessing circuits, that is, the input of each preprocessing circuit is the same analog audio signal to be processed, and the output of each preprocessing circuit is digital audio
  • the signals are different, resulting in multiple digital audio signals.
  • the analog audio signal to be processed can be intercepted into different segments with a predetermined duration as a frame signal, or a predetermined number of sampled data can be used as a frame signal after analog-to-digital conversion.
  • the subsequent audio signal processing procedures are all It can be processed in units of one frame of signal. In order to ensure the continuity of the signal, there may be a certain overlap between adjacent frame signals, that is, the tail of the previous frame signal and the head of the next frame signal have an overlap amount, thereby establishing the correlation between adjacent frames.
  • Step S102 Perform frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
  • frequency domain conversion is performed on each digital audio signal in the multiple digital audio signals, so as to obtain frequency domain data corresponding to each digital audio signal.
  • the multiple channels of digital audio signals are subjected to frequency domain conversion to obtain multiple channels of frequency domain data, and the fusion of multiple channels of frequency domain data in the frequency domain can be realized.
  • the frequency domain conversion method may adopt Fourier transform (such as discrete Fourier transform), Laplace transform, Z transform, etc. The specific frequency domain conversion process will not be repeated here.
  • Step S103 Determine frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data.
  • each pre-processing circuit since each pre-processing circuit amplifies the same analog audio signal to be processed with different gains, each audio signal has a different maximum and minimum value, and the maximum value of the audio signal with a larger gain And the minimum value are relatively large, and the maximum and minimum values of the audio signal with a smaller gain are relatively small, and the dynamic range is the ratio of the maximum and minimum values of the audio signal without distortion.
  • the louder sound can be adjusted by A pre-processing circuit with a relatively large gain is provided, and a smaller sound can be provided by a pre-processing circuit with a relatively small gain, so that the dynamic system of the recording system can be improved, with higher sensitivity and lower noise at the same time.
  • the size of the sound can be measured by the energy feature information of the audio signal, such as the sound pressure level of the analog audio signal or digital audio signal, or the amplitude of the analog audio signal or digital audio signal.
  • Step S104 Convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal.
  • the frequency domain fusion data can be converted into a time domain audio signal.
  • the conversion method can adopt the inverse transformation of the Fourier transform (such as the discrete Fourier transform), the Lap The inverse transform of the Lass transform, the inverse transform of the Z transform, etc., will not be repeated here.
  • the output audio signal can be obtained according to the time domain audio signal.
  • operations such as compression and noise reduction can be performed on the time domain audio signal.
  • the process of obtaining the output audio signal according to the time domain audio signal also needs to splice the signals of each frame to establish the correlation between adjacent frames. Specifically, the time domain audio signal of the current frame and the time domain audio signal of the previous frame can be superimposed.
  • the audio signal processing method of this embodiment uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog audio signal. Digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data One or at least two channels of target frequency domain data determine frequency domain fusion data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal.
  • the method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor and meet the requirement of high signal-to-noise ratio.
  • the audio signal processing method further includes:
  • the energy characteristic information of the analog audio signal to be processed may be the sound pressure level of the analog audio signal or the amplitude of the analog audio signal, etc., and specifically may be the maximum, minimum, or instantaneous amplitude of the analog audio signal.
  • the intermediate value may also be the maximum, minimum or intermediate value of the average amplitude of the analog audio signal in a short time (preset duration).
  • the step S103 of determining frequency domain fusion data according to one or at least two channels of frequency domain data in the multiple channels of frequency domain data includes:
  • Step S201 Determine one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information
  • Step S202 Determine frequency domain fusion data according to the one or at least two channels of target frequency domain data.
  • one or at least two channels of target frequency domain data to be fused can be determined from multiple channels of frequency domain data according to the energy characteristic information of the analog audio signal to be processed, for example, according to the simulation to be processed
  • the size of the energy feature information of the audio signal determines the number of target frequency domain data. For example, the greater the energy feature information, the greater the number of target frequency domain data; in addition, the reference energy feature parameters of the processing circuit can also be preset based on the energy feature information Compare and then determine one or at least two channels of target frequency domain data to be fused.
  • Each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter.
  • the reference energy characteristic parameter is included by the preprocessing circuit Determined by the analog gain of the amplifier circuit, the reference energy characteristic parameter and the energy characteristic information belong to the same parameter, that is, the sound pressure level or amplitude of the digital audio signal output by the preprocessing circuit, which can be in the case of no distortion
  • the maximum, minimum or intermediate value of the instantaneous amplitude of the digital audio signal, or the maximum, minimum or intermediate value of the average amplitude of the digital audio signal in a short time (preset duration) when the preprocessing circuit includes the amplifier circuit The larger the analog gain of, the larger the corresponding reference energy characteristic parameter.
  • the determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information includes:
  • the first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data.
  • the analog gain of the amplifier circuit included in each preprocessing circuit is determined.
  • At least two channels of target frequency domain data are determined from multiple channels of frequency domain data, where the first target frequency domain data may be only one channel of target frequency domain data, and of course, it may also be more than one channel of target frequency domain data; Similarly, the second target frequency domain data may also be only one channel of target frequency domain data or more than one channel of target frequency domain data.
  • determining the first target frequency domain data and the second target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information may specifically include:
  • Step S301 Determine a first reference energy characteristic parameter and a second reference energy characteristic parameter adjacent to the energy characteristic information from a plurality of reference energy characteristic parameters;
  • Step S302 Determine first target frequency domain data and second target frequency domain data from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter;
  • the first target frequency domain data and the second target frequency domain data are obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, respectively.
  • a digital audio signal and a second digital audio signal are respectively obtained from the analog audio signal to be processed by a first preprocessing circuit and a second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter .
  • Step S303 Determine frequency domain fusion data according to the first target frequency domain data and the second target frequency domain data.
  • the reference energy characteristic parameters (indicated by L 1 , L 2 , ..., L I , where I is the number of preset processing circuits) of multiple preset processing circuits are selected and the analog audio to be processed
  • the energy feature information of the signal (represented by L c ) is adjacent to the first reference energy feature parameter (represented by Li ′ , where 1 ⁇ i′ ⁇ I-1) and the second reference energy feature parameter (represented by L i′+ 1 means), that is, L c is between L i′ and L i′+1 , where the first reference energy characteristic parameter L i′ corresponds to the first preprocessing circuit, and the first digital audio output by the first preprocessing circuit
  • the frequency domain data obtained by frequency domain conversion of the signal is the first target frequency domain data
  • the second reference energy characteristic parameter Li′+1 corresponds to the second preprocessing circuit
  • the frequency domain data obtained by frequency domain conversion of the audio signal is the second target frequency domain data, and the frequency domain fusion data can be obtained
  • the present embodiment may be a second reference parameter L i and L i the energy characteristic parameters determining a first reference energy characteristics' + after 1 to L i 'and L i'-1 (this time need i'> 1)
  • the frequency domain data obtained by the frequency domain conversion of the digital audio signal output by the corresponding preprocessing circuit is the first target frequency domain data, which can also be L i′+1 and L i′+2 (in this case, i′ ⁇ I -1)
  • the frequency domain data obtained by frequency domain conversion of the digital audio signal output by the corresponding preprocessing circuit is the second target frequency domain data.
  • the first target frequency domain data and the second target frequency domain data may also include For more channels of frequency domain data, we will not give an example here.
  • the determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information includes:
  • Step S401 When the energy feature information is less than the smallest third reference energy feature parameter among the multiple reference energy feature parameters, determine from the multiple channels of frequency domain data according to the third reference energy feature parameter Third target frequency domain data;
  • the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is corresponding to a third reference energy characteristic parameter Obtained by the third preprocessing circuit from the analog audio signal to be processed;
  • Step S402 Acquire the frequency domain fusion data according to the third target frequency domain data.
  • the smallest third reference energy characteristic parameter L 1 is greater than that of the analog audio signal to be processed
  • the energy feature information L c that is, L c is less than L 1
  • the third reference energy feature parameter L 1 corresponds to the third digital audio signal output by the third preprocessing circuit and the frequency domain data obtained by frequency domain conversion is the first Three target frequency domain data, and then the frequency domain fusion data can be obtained according to the third target frequency domain data.
  • the present third embodiment with reference to the energy parameters of L 1 may be smaller than in the embodiment determined to be processed analog audio signal energy characteristic information L c, the digital audio signals L 1 and L 2 corresponding to an output of the preprocessing circuit frequency-
  • the frequency domain data obtained by the domain conversion is the third target frequency domain data.
  • the third target frequency domain data may also include more channels of frequency domain data, and no examples are given here.
  • the fusion of the multiple channels of frequency domain data to obtain frequency domain fusion data further includes:
  • Step S501 When the energy feature information is greater than the largest fourth reference energy feature parameter among the multiple reference energy feature parameters, determine from the multiple channels of frequency domain data according to the fourth reference energy feature parameter The fourth target frequency domain data;
  • the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multi-channel digital audio signal, and the fourth digital audio signal is corresponding to a fourth reference energy characteristic parameter Obtained by the fourth preprocessing circuit from the analog audio signal to be processed;
  • Step S502 Acquire the frequency domain fusion data according to the fourth target frequency domain data.
  • the largest fourth reference energy characteristic parameter L I is smaller than that of the analog audio signal to be processed
  • the energy feature information L c that is, L c is greater than L I
  • the fourth reference energy feature parameter L I corresponds to the fourth digital audio signal output by the fourth preprocessing circuit and the frequency domain data obtained by frequency domain conversion is the first Four target frequency domain data, and the frequency domain fusion data can be obtained according to the fourth target frequency domain data.
  • the digital audio signal output by the preprocessing circuit corresponding to L I and L I-1 is the fourth target frequency domain data.
  • the fourth target frequency domain data may also include more channels of frequency domain data respectively, and no examples are given here.
  • the energy feature information L of the analog audio signal to be processed may be first c is compared with the reference energy characteristic parameters of multiple preset processing circuits, if L c is less than (or equal to) the third reference energy characteristic parameter L 1 , then steps 401-402 are executed; if L c is greater than (or equal to) the first For the fourth reference energy characteristic parameter L I , steps 501 to 502 are executed; if L c is between the adjacent third reference energy characteristic parameter L 1 and the fourth reference energy characteristic parameter L I , then steps 301-303 are executed.
  • the determining frequency domain fusion data according to the one or at least two channels of target frequency domain data includes:
  • the frequency domain fusion data is obtained by performing a spectrum superposition operation on the frequency domain data.
  • the acquiring frequency domain fusion data according to the one or more channels of frequency domain data includes:
  • performing the superposition operation can set weights for the first target frequency domain data and the second target frequency domain data, by superposing the first target frequency domain data and the second target frequency domain data with different weights. , You can get frequency domain fusion data with different dynamic ranges.
  • each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter
  • the plurality of reference energy characteristic parameters are determined according to the analog gain of the amplifying circuit included in the plurality of preprocessing circuits , wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the first preprocessing circuit and the first preprocessing circuit corresponding to the first target frequency domain data among the plurality of preset processing circuits
  • the reference energy characteristic parameter of the second preprocessing circuit corresponding to the second target frequency domain data is determined.
  • the weights corresponding to the first target frequency domain data and the second target frequency domain data are determined by the reference energy characteristic parameters of the corresponding preprocessing circuit. More specifically, the weights corresponding to the first target frequency domain data a first reference energy characteristic parameters L i and a second target frequency domain data corresponding to the second reference feature parameter L i + energy magnitude relation between the energy characteristic information L c 1 and the analog audio signal to be processed is determined, wherein the reference energy characteristic parameters of the analog audio signal closely approximates the energy characteristic information L c, the larger the weight, e.g., closer to L i L c, then the analog audio signal is closer to a digital audio signal L i corresponds to the pre-processing circuit, it is necessary to increase
  • the weight of the first target frequency domain data corresponding to L i .
  • the weight a 1 of the first target frequency domain data and the weight a 2 of the second target frequency domain data can be determined by the following formula:
  • the determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data includes:
  • Step S601 Perform compression processing on the one or at least two channels of target frequency domain data according to the compression coefficient corresponding to each of the one or at least two channels of target frequency domain data;
  • Step S602 Acquire frequency domain fusion data according to one or more channels of frequency domain data after the compression processing.
  • the output of the recording system has a digital quantization range, that is, there are restrictions on the maximum amplitude and minimum amplitude of the audio signal, the maximum amplitude of the audio signal cannot be greater than the maximum threshold, and the minimum amplitude of the audio signal The value cannot be less than the minimum threshold, so when acquiring the frequency domain fusion data, it is necessary to compress the target frequency domain data to prevent the fusion frequency domain fusion data from exceeding the digital quantization range.
  • the step of compressing the target frequency domain data can be performed simultaneously with the above-mentioned superposition operation, or it can be completed before the superposition operation.
  • the compression processing is linear compression processing.
  • the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is determined according to the analog gain of the amplifier included in the preprocessing circuit corresponding to each channel of frequency domain data.
  • the compression coefficient corresponding to any channel of frequency domain data may be the product of the channel equalization parameter and the scaling factor of the corresponding preprocessing circuit.
  • a certain preprocessing circuit is used as a reference preprocessing circuit, and the The channel equalization parameter is the ratio of the analog gain of the preprocessing circuit to the analog gain of the reference preprocessing circuit, and the scaling factor is obtained according to the size of the digital audio signal output by the preprocessing circuit.
  • the compression coefficient corresponding to any channel of frequency domain data can be obtained by the following formula:
  • G i′ is the analog gain of the preprocessing circuit
  • G ref is the reference preprocessing
  • ⁇ is the scaling factor, used to scale the frequency domain data.
  • steps S601-602 can be executed only when the frequency domain fusion data exceeds the digital quantization range, and it can also be judged before step S601 whether the frequency domain fusion data has the possibility of exceeding the digital quantization range. When the range is possible, steps S601-602 are executed.
  • the time domain audio signal is the current frame time domain audio signal
  • the step S104 in the foregoing embodiment of obtaining the output audio signal according to the time domain audio signal includes:
  • the output audio signal is determined according to the time-domain fusion audio signal of the current frame.
  • the above-mentioned audio signal processing procedures are performed in units of one frame signal.
  • there may be a certain overlap between adjacent frame signals that is, the tail of the previous frame signal and the next
  • the header of the frame signal has an overlap amount, thereby establishing correlation between adjacent frames. Therefore, after converting the frequency domain fusion data of the current frame into the time domain audio signal of the current frame in S104, the overlapped part of the time domain audio signal of the current frame and the time domain audio signal of the previous frame can be overlapped and superimposed.
  • the non-overlapping part of the time domain audio signal and the previous frame of time domain audio signal is not superimposed to obtain the current frame time domain fused audio signal, and the output audio signal can be determined according to the current frame time domain fused audio signal.
  • the determining the output audio signal according to the time-domain fusion audio signal of the current frame may specifically include:
  • Step S701 Perform compression processing on the time-domain fused audio signal according to the current frame according to a preset compression coefficient
  • Step S702 Determine the output audio signal according to the time-domain fusion audio signal of the current frame after the compression processing.
  • the time-domain fusion audio signal of the current frame may be subjected to dynamic range compression, where the preset compression coefficient may be derived from a preset nonlinear function, that is, the current frame
  • the amplitude of the time-domain fusion audio signal is non-linearly mapped according to the non-linear function, so that the smaller signal remains unchanged or amplified, and the larger signal is compressed to the digital quantization range.
  • a typical nonlinear function is shown in the curve in Figure 9.
  • is input as the abscissa to the nonlinear function, and the ordinate function value of the nonlinear function is the compression
  • f(
  • the performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data includes:
  • an analysis window function is used to perform windowing processing on each channel of digital audio signals.
  • the analysis window function includes but is not limited to sine window function, rectangular window function, triangular window function, Hanning window function (Hanning), Gaussian window function. Window function (Gaussian), Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), Flat top window function (Flap Top), Kaiser window function (Kaiser).
  • windowing is not required, which is equivalent to adding a rectangular window.
  • the obtaining the output audio signal according to the time domain audio signal includes:
  • the output audio signal is determined according to the time domain audio signal after the windowing process.
  • a synthesis window function is used to perform windowing processing on the time domain audio signal.
  • the synthesis window function includes, but is not limited to, a sine window function, a rectangular window function, a triangular window function, a Hanning window function (Hanning), and a Gaussian window. Function (Gaussian), Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser).
  • Windowing is not required, which is equivalent to adding a rectangular window.
  • FIG. 10 is a flowchart of an audio signal processing method provided by another embodiment of the present invention. As shown in FIG. 10, on the basis of the foregoing embodiment, the method in this embodiment may include:
  • Step S801 Use multiple pre-processing circuits to process analog audio signals to be processed to obtain multiple digital audio signals.
  • each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter, and the analog gains of the amplifiers of each preprocessing circuit are different.
  • G i 1, 2, ..., I
  • I is the number of preprocessing circuits.
  • Step S802 Extract N sampling points every M sampling points of each digital audio signal in the multi-channel digital audio signal, and use them as a frame of digital audio signal.
  • the digital audio signal of the first frame of the i-th preprocessing circuit is denoted as:
  • N is the frame length
  • M is the frame shift
  • 0 ⁇ M ⁇ N that is, the last NM sampling points of the l-1th frame are the same as the first NM sampling points of the lth frame, 0.005f s ⁇ N ⁇ f s
  • N is a power of 2.
  • Step S803 Perform windowing processing on the multiple channels of digital audio signals.
  • the first frame of digital audio signal x i (t) l of the i-th preprocessing circuit is windowed, and the windowed digital audio signal is:
  • h ana (t) is the N-point analysis window function.
  • Window functions can include but are not limited to sine window, rectangular window function, triangular window function, Hanning window, Gaussian window function, Black Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser) and so on.
  • Step S804 Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
  • n is the discrete spectrum sequence
  • e is the natural constant
  • It is an imaginary unit.
  • fast Fourier transform can be used to accelerate calculation.
  • Step S805 Determine frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data.
  • multiple channels of frequency domain data can be spectrum fused, and the spectrum signals of multiple channels of frequency domain data can be used for linear superposition with a predetermined weight, and the superposition logic is an analog audio signal with a better dynamic range covering the current frame , The higher the weight.
  • the frequency domain fusion data is X syn (n) l :
  • ⁇ i is the weight corresponding to the frequency domain data of the i-th preprocessing circuit, and ⁇ i ⁇ 0.
  • a 1 and a 2 are the weights of the frequency domain data Xi ′ (n) l and Xi ′+1 (n) l , respectively, which can be determined by the following formula:
  • the frequency domain data in order to prevent the final output audio signal from exceeding the digital quantization range of the recording system, can be linearly compressed in the superposition operation, that is, each frequency domain data is multiplied by the corresponding compression coefficient, where G x ( L i′ ) is the compression coefficient corresponding to the frequency domain data X i′ (n) i , which can be determined by the following formula:
  • G i′ is the analog gain of the preprocessing circuit
  • G ref is the reference preprocessing
  • ⁇ is the scaling factor, which is used to scale the frequency domain data of the i'th preprocessing circuit.
  • each compression factor can be set to 1 in the above formula.
  • Step S806 Convert the frequency domain fusion data into a time domain audio signal.
  • the first frame of frequency domain fusion data X syn (n) l is processed by the inverse transform of the discrete Fourier transform to obtain the first frame of time domain audio signal x'syn (t) l :
  • the inverse transform of the fast Fourier transform can be used to accelerate the calculation.
  • Step S807 Perform windowing processing on the time domain audio signal.
  • the time domain audio signal x'syn (t) l of the l- th frame is windowed to obtain a windowed time domain audio signal
  • h syn (t) is an N-point synthesis window function.
  • Commonly used synthesis window functions can include but are not limited to sine window, rectangular window function, triangular window function, Hanning window, Gaussian window function, Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser) and so on.
  • Step S808 Perform superposition processing on the current frame time domain audio signal and the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal.
  • the overlapped part of the time domain audio signal of the current frame and the time domain audio signal of the previous frame may be overlapped and superimposed to obtain the time domain fusion audio signal of the current frame:
  • Step S809 Perform compression processing on the time-domain fusion audio signal according to the current frame; determine the output audio signal according to the time-domain fusion audio signal of the current frame after the compression processing.
  • step S805 if the dynamic range compression of the frequency domain data is not performed in step S805, or after the dynamic range compression of the frequency domain data in step S805, the amplitude of x syn (t) l still exceeds the digital quantization range, It is necessary to perform dynamic range compression in the time domain to meet the needs of digital quantization.
  • the dynamic range compression is equivalent to the nonlinear mapping of the signal amplitude in the time domain.
  • the basic principle is that the small signal remains unchanged or appropriately amplified, and the large signal is compressed to the digital quantization range L max , where L max is the maximum quantization. Amplitude.
  • a typical mapping function as shown in the curve in Fig. 9 can be selected, and the input abscissa is the current frame time-domain fusion audio signal
  • Output the compressed and processed current frame time-domain fused audio signal y(n) l corresponding to the M points overlapped in the l-th frame by the following formula:
  • y(n) l sign(x syn (t) l )f(
  • ) n 0,1,...,M-1
  • sign( ⁇ ) is a function for judging the sign, a positive number is 1, a negative number is -1, and a zero is 0.
  • step S809 may not be performed.
  • the dynamic range compression in the frequency domain is linear compression, which does not change the timbre of the sound
  • the dynamic range compression in the time domain is non-linear compression, which will cause certain timbre distortion.
  • the current frame time-domain fusion audio signal y(n) l after the compression processing corresponding to the M point overlapped in the l-th frame is obtained, which can be used as an output audio signal, which can be recorded and saved or played in real time.
  • the audio signal processing method of this embodiment uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog audio signal. Digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data One or at least two channels of target frequency domain data determine frequency domain fusion data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal.
  • the method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor, meet the requirements of high signal-to-noise ratio, and can simultaneously take into account the high signal-to-noise ratio recording of small voices and loud voices
  • the non-overload and distortion-free recording meets the recording needs of high dynamic range.
  • the test environment is an ordinary quiet room.
  • the first 3 seconds of the sound source is the voice signal, and the last 1 second is the tapping sound near the microphone.
  • the frame shift M 1024.
  • Both the analysis window and the synthesis window are sine windows:
  • the digital audio signal (time domain signal) of the first frame output by the first preprocessing circuit and the second preprocessing circuit is shown in Figure 11a and Figure 11b, respectively.
  • the output of the first preprocessing circuit and the second preprocessing circuit are
  • the digital audio signal (time spectrum) of the first frame is shown in Figure 12a and Figure 12b, respectively.
  • the time-domain signal and time-frequency spectrum of the output audio signal are shown in Figs. 13a and 13b, respectively.
  • the speech segment signal is basically the same as the digital audio signal of the first frame output by the high gain first preprocessing circuit, which meets the requirements of high signal-to-noise ratio;
  • the percussion segment retains the low gain second pre-processing Processing the frequency spectrum of the digital audio signal of the first frame of the output of the circuit, and dynamically adjusting its amplitude; and a continuous smooth transition between the small sound and the loud sound. Therefore, the method of this embodiment can simultaneously take into account the high signal-to-noise ratio recording of small sounds and the non-overload and distortion-free recording of loud sounds, meeting the recording requirements of high dynamic range.
  • FIG. 14 is a structural diagram of an audio signal processing device provided by an embodiment of the present invention.
  • the audio signal processing device 90 includes a memory 92 and a processor 91.
  • the memory 92 is used to store program codes
  • the processor 91 calls the program code, and when the program code is executed, is used to perform the following operations:
  • a plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals.
  • Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
  • the frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  • the processor 91 is configured to: obtain energy feature information of the analog audio signal to be processed;
  • the determining frequency domain fusion data according to one or at least two channels of frequency domain data among the multiple channels of frequency domain data includes:
  • each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter
  • the processor 91 determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processor 91 is configured to:
  • the first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data.
  • the analog gain of the amplifier circuit included in each preprocessing circuit is determined.
  • the processing The device 91 is configured as:
  • the first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter, where the first target frequency domain data and the second target
  • the frequency domain data is obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, and the first digital audio signal and the second digital audio signal are respectively
  • the analog audio signal to be processed is obtained by the first preprocessing circuit and the second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter.
  • the processor 91 determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processor 91 is configured for:
  • a third target is determined from the multi-channel frequency domain data according to the third reference energy characteristic parameter Frequency domain data, wherein the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is obtained from a third reference energy
  • the third preprocessing circuit corresponding to the characteristic parameter is obtained from the analog audio signal to be processed;
  • the processor 91 when the processor 91 fuses the multiple channels of frequency domain data to obtain frequency domain fusion data, the processor 91 is further configured to:
  • a fourth target is determined from the multiple channels of frequency domain data according to the fourth reference energy feature parameter Frequency domain data, where the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multiple digital audio signals, and the fourth digital audio signal is obtained by a fourth reference energy Obtained by the fourth preprocessing circuit corresponding to the characteristic parameter from the analog audio signal to be processed;
  • the processor 91 determines frequency domain fusion data according to the one or at least two channels of target frequency domain data
  • the processor 91 is configured to:
  • the one or more channels of frequency domain data include first target frequency domain data and second target frequency domain data;
  • the processor 91 When the processor 91 obtains frequency domain fusion data according to the one or more channels of frequency domain data, the processor 91 is configured to:
  • each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter
  • the multiple reference energy characteristic parameters are based on the fact that the multiple preprocessing circuits include
  • the analog gain of the amplification circuit is determined, wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the weights corresponding to the first target frequency domain data in the plurality of preset processing circuits
  • the reference energy characteristic parameters of the first preprocessing circuit and the second preprocessing circuit corresponding to the second target frequency domain data are determined.
  • the processor 91 determines frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data
  • the processor 91 is configured to :
  • the compression processing is linear compression processing.
  • the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is determined according to the analog gain of the amplifier included in the preprocessing circuit corresponding to each channel of frequency domain data of.
  • the time domain audio signal is the current frame time domain audio signal, wherein, when the processor 91 obtains the output audio signal according to the time domain audio signal, the processor 91 Is configured as:
  • the output audio signal is determined according to the time-domain fusion audio signal of the current frame.
  • the processor 91 determines the output audio signal according to the time-domain fusion audio signal of the current frame, the processor 91 is configured to:
  • the output audio signal is determined according to the time-domain fusion audio signal of the current frame after the compression processing.
  • the processor 91 when the processor 91 performs frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data, the processor 91 is configured to:
  • the processor 91 when the processor 91 obtains an output audio signal according to the time domain audio signal, the processor 91 is configured to:
  • the output audio signal is determined according to the time domain audio signal after the windowing process.
  • the audio signal processing device uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, where each of the multiple preprocessing circuits includes an amplifier and The analog-to-digital converter, and the analog gains of the amplifiers of each preprocessing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data Determine the frequency domain fusion data from one or at least two channels of target frequency domain data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal.
  • This embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, can reduce the noise floor, meet the requirements of high signal-to-noise ratio, and can simultaneously take into account the high signal-to-noise ratio recording of small sounds and the absence of loud sounds. Overloading and distortion-free recording meets the recording needs of high dynamic range.
  • An embodiment of the present invention provides a recording system.
  • the recording system includes: a microphone for collecting analog audio signals; and an audio processing device 90 as described in the foregoing embodiment.
  • this embodiment also provides a computer-readable storage medium on which a computer program is stored, and the computer program is executed by a processor to implement the audio processing method described in the foregoing embodiment.
  • the disclosed device and method may be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components can be combined or It can be integrated into another system, or some features can be ignored or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
  • the functional units in the various embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated unit may be implemented in the form of hardware, or may be implemented in the form of hardware plus software functional units.
  • the above-mentioned integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium.
  • the above-mentioned software functional unit is stored in a storage medium and includes several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor execute the method described in the various embodiments of the present invention. Part of the steps.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program code .

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Signal Processing (AREA)
  • Stereophonic System (AREA)

Abstract

Embodiments of the present invention provide an audio signal processing method and device and a storage medium. The method comprises: processing an analog audio signal to be processed using multiple preprocessing circuits, so as to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits comprises an amplifier and an analog-to-digital converter, and analog gains of amplifiers of the respective preprocessing circuits are different from each other; performing frequency domain conversion on the multiple digital audio signals so as to obtain multiple pieces of frequency domain data; determining frequency domain fusion data according to one or at least two pieces of target frequency domain data of the multiple pieces of frequency domain data; and converting the frequency domain fusion data into a time domain audio signal, and obtaining an output audio signal according to the time domain audio signal. The method in the embodiments can effectively improve the dynamic range of a recording system and provide high sensitivity, while also reducing background noise and achieving the required high signal-to-noise ratio.

Description

音频信号处理方法、设备及存储介质Audio signal processing method, equipment and storage medium 技术领域Technical field
本发明实施例涉及信号处理领域，尤其涉及一种音频信号处理方法、设备及存储介质。The embodiments of the present invention relate to the field of signal processing, and in particular, to an audio signal processing method, device, and storage medium.
背景技术Background technique
麦克风的动态范围是指麦克风能录到的无失真最大声音和最小声音的比值，普通麦克风的动态范围一般不超过100dBA。当由模拟信号转为16bit有符号整型的数字信号时，动态范围最多为20lg(32768)＝90.3dB。通常信号中还有轻微的电路底噪和量化噪声，有效的动态范围还会进一步压缩。在一些特设的使用场景中，比如结构冲击、强气流、瞬态强噪声或麦克风附近的强噪声等情况，录音需要更高的动态范围，而普通麦克风不能满足需求。The dynamic range of a microphone refers to the ratio of the maximum sound and the minimum sound without distortion that the microphone can record. The dynamic range of an ordinary microphone generally does not exceed 100dBA. When converting from an analog signal to a 16-bit signed digital signal, the dynamic range is at most 20lg (32768) = 90.3dB. Usually there are slight circuit noise and quantization noise in the signal, and the effective dynamic range will be further compressed. In some special use scenarios, such as structural impact, strong airflow, strong transient noise or strong noise near the microphone, recording requires a higher dynamic range, and ordinary microphones cannot meet the demand.
现有技术中提高麦克风录音动态范围一般有两种方式：一种是改善振膜的特性，更小的振膜尺寸和更大刚度的振膜材料能带来更高的动态范围，并满足更高最大声压级的录音，其副作用则是降低了灵敏度、增强了底噪；另一种是多麦克风的高动态范围方案，即多个麦克风分别覆盖多个动态范围，根据输入信号的特征来切换至某个麦克风或融合多个麦克风，并给出高动态范围的输出信号。In the prior art, there are generally two ways to improve the dynamic range of microphone recording: one is to improve the characteristics of the diaphragm. A smaller diaphragm size and greater rigidity of the diaphragm material can bring a higher dynamic range and meet the needs of more The side effect of recording with high maximum sound pressure level is that the sensitivity is reduced and the noise floor is enhanced; the other is the high dynamic range solution with multiple microphones, that is, multiple microphones cover multiple dynamic ranges, depending on the characteristics of the input signal. Switch to a microphone or merge multiple microphones, and give a high dynamic range output signal.
在第一种方式中，普通麦克风的动态范围仍然受限，很难同时兼顾大信号和小信号声音的录音，若满足小声音录音，则会出现大声音过载问题；若满足大声音录音，则会出现小声音低信噪比问题。而在第二种方式中，需要两个或多个覆盖不同动态范围的麦克风录音，最后只合成一路声信号。而增加麦克风数量、以及多个麦克风的选型、调校、测试等环节显著的增加了成本。In the first method, the dynamic range of an ordinary microphone is still limited, and it is difficult to record both large and small signal sounds at the same time. If it meets the requirements for small voice recording, loud voice overload will occur; if it meets loud voice recording, then There will be a low signal-to-noise ratio problem for small sounds. In the second method, two or more microphones covering different dynamic ranges are required to record, and finally only one sound signal is synthesized. The increase in the number of microphones and the selection, adjustment, and testing of multiple microphones significantly increase the cost.
发明内容Summary of the invention
本发明实施例提供一种音频信号处理方法、设备及存储介质，以有效的提高录音***的动态范围。The embodiments of the present invention provide an audio signal processing method, device, and storage medium, so as to effectively increase the dynamic range of the recording system.
本发明实施例的第一方面是提供一种音频信号处理方法，包括：The first aspect of the embodiments of the present invention is to provide an audio signal processing method, including:
利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；A plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals. Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
对所述多路数字音频信号进行频域转换以获取多路频域数据；Performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data;
根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；Determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data;
将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。The frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
本发明实施例的第二方面是提供一种音频信号处理设备，包括：存储器和处理器；The second aspect of the embodiments of the present invention is to provide an audio signal processing device, including: a memory and a processor;
所述存储器用于存储程序代码；The memory is used to store program codes;
所述处理器，调用所述程序代码，当程序代码被执行时，用于执行以下操作：The processor calls the program code, and when the program code is executed, is used to perform the following operations:
利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；A plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals. Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
对所述多路数字音频信号进行频域转换以获取多路频域数据；Performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data;
根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；Determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data;
将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。The frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
本发明实施例的第三方面是提供一种录音***，包括：The third aspect of the embodiments of the present invention is to provide a recording system, including:
麦克风，用于采集模拟音频信号；以及Microphone, used to collect analog audio signals; and
如第二方面所述的音频处理设备。The audio processing device described in the second aspect.
本发明实施例的第四方面是提供一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行以实现第一方面所述的方法。The fourth aspect of the embodiments of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the method described in the first aspect.
本实施例提供的音频信号处理方法、设备及存储介质，通过利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处 理电路的所述放大器的模拟增益各不相同；对所述多路数字音频信号进行频域转换以获取多路频域数据；根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。本实施例的方法可实现有效的提高录音***的动态范围，具有较高的灵敏度，同时能够降低底噪，满足高信噪比的要求。The audio signal processing method, device and storage medium provided in this embodiment use multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits is preprocessed The processing circuit includes an amplifier and an analog-to-digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; One or at least two channels of target frequency domain data among the channels of frequency domain data determine frequency domain fusion data; the frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal. The method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor and meet the requirement of high signal-to-noise ratio.
附图说明Description of the drawings
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。In order to explain the technical solutions in the embodiments of the present invention more clearly, the following will briefly introduce the drawings used in the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor.
图1为本发明实施例提供的音频信号处理方法的流程图；FIG. 1 is a flowchart of an audio signal processing method provided by an embodiment of the present invention;
图2为本发明实施例提供的多个预处理电路并联的示意图；FIG. 2 is a schematic diagram of multiple preprocessing circuits in parallel provided by an embodiment of the present invention;
图3为本发明另一实施例提供的音频信号处理方法的流程图；FIG. 3 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图4为本发明另一实施例提供的音频信号处理方法的流程图；4 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图5为本发明另一实施例提供的音频信号处理方法的流程图；FIG. 5 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图6为本发明另一实施例提供的音频信号处理方法的流程图；6 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图7为本发明另一实施例提供的音频信号处理方法的流程图；FIG. 7 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图8为本发明另一实施例提供的音频信号处理方法的流程图；FIG. 8 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图9为非线性映射的映射函数的曲线图；Figure 9 is a graph of the mapping function of the nonlinear mapping;
图10为本发明另一实施例提供的音频信号处理方法的流程图；FIG. 10 is a flowchart of an audio signal processing method provided by another embodiment of the present invention;
图11a为第1预处理电路输出的第l帧的数字音频信号的时域信号图；Figure 11a is a time-domain signal diagram of the digital audio signal of the first frame output by the first preprocessing circuit;
图11b为第2预处理电路输出的第l帧的数字音频信号的时域信号图；11b is a time-domain signal diagram of the digital audio signal of the first frame output by the second preprocessing circuit;
图12a为第1预处理电路输出的第l帧的数字音频信号的时频谱图；FIG. 12a is a time-frequency spectrum diagram of the digital audio signal of the first frame output by the first preprocessing circuit;
图12b为第2预处理电路输出的第l帧的数字音频信号的时频谱图；Figure 12b is a time-frequency spectrum diagram of the digital audio signal of the first frame output by the second preprocessing circuit;
图13a为第l帧的输出音频信号的时域信号图；Figure 13a is a time-domain signal diagram of the output audio signal of the lth frame;
图13b为第l帧的输出音频信号的时频谱图；Fig. 13b is a time-frequency spectrum diagram of the output audio signal of the first frame;
图14为本发明实施例提供的音频信号处理设备的结构图。Fig. 14 is a structural diagram of an audio signal processing device provided by an embodiment of the present invention.
具体实施方式detailed description
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The technical solutions in the embodiments of the present invention will be clearly described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.
当一个组件被认为是“连接”另一个组件，它可以是直接连接到另一个组件或者可能同时存在居中组件。When a component is considered to be "connected" to another component, it can be directly connected to another component or there may be a centered component at the same time.
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。本文所使用的术语“及/或”包括一个或多个相关的所列项目的任意的和所有的组合。Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of the present invention. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The term "and/or" as used herein includes any and all combinations of one or more related listed items.
下面结合附图，对本发明的一些实施方式作详细说明。在不冲突的情况下，下述的实施例及实施例中的特征可以相互组合。Hereinafter, some embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the case of no conflict, the following embodiments and features in the embodiments can be combined with each other.
本发明实施例提供一种音频信号处理方法。图1为本发明实施例提供的音频信号处理方法的流程图。如图1所示，本实施例中的音频信号处理方法，可以包括：The embodiment of the present invention provides an audio signal processing method. FIG. 1 is a flowchart of an audio signal processing method provided by an embodiment of the present invention. As shown in FIG. 1, the audio signal processing method in this embodiment may include:
步骤S101、利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同。Step S101: Use multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog-to-digital converter, and each preprocessing circuit The analog gains of the amplifiers of the processing circuit are different.
在本实施例中，如图2所示，将多个预处理电路并联，每一预处理电路包括放大器和模数转换器，需要说明的是，所述多个预处理电路至少包括两个预处理电路；其中放大器可用于对模拟音频信号进行功率放大，放大倍数通常用增益表示，本实施例中各预处理电路的放大器的模拟增益各不相同，进一步的，模拟增益大小相邻的两个预处理电路的动态范围至少存在部分重叠；而模数转换器可用于将模拟音频信号转换成数字音频信号，以便于后续的信号处理。In this embodiment, as shown in Figure 2, multiple preprocessing circuits are connected in parallel, and each preprocessing circuit includes an amplifier and an analog-to-digital converter. It should be noted that the multiple preprocessing circuits include at least two preprocessing circuits. Processing circuit; where the amplifier can be used to amplify the power of the analog audio signal, the amplification factor is usually expressed by gain, the analog gain of the amplifier of each preprocessing circuit in this embodiment is different, and further, the two adjacent analog gains The dynamic range of the pre-processing circuit at least partially overlaps; and the analog-to-digital converter can be used to convert analog audio signals into digital audio signals for subsequent signal processing.
本实施例中，如图2所示，待处理的模拟音频信号x可由麦克风采集，然后分别输入到并联的多个预处理电路中，分别经过不同增益的功率放大，并转换成数字音频信号x 1、…x i…、x I，其中I为预处理电路的个数，也即每 一预处理电路的输入均为同一待处理的模拟音频信号，而每一预处理电路的输出的数字音频信号则各不相同，从而得到多路数字音频信号。 In this embodiment, as shown in FIG. 2, the analog audio signal x to be processed can be collected by a microphone, and then input into a plurality of parallel preprocessing circuits, respectively, through power amplification of different gains, and converted into a digital audio signal x 1 ,...x i …,x I , where I is the number of preprocessing circuits, that is, the input of each preprocessing circuit is the same analog audio signal to be processed, and the output of each preprocessing circuit is digital audio The signals are different, resulting in multiple digital audio signals.
本实施例中可以将待处理的模拟音频信号以预定时长截取为不同片段作为一帧信号，或者在进行模数转换后以预定数量的采样数据作为一帧信号，后续的音频信号处理流程中均可以以一帧信号为单位进行处理。为了保证信号的连续性，相邻帧信号之间可具有一定的重叠，即前一帧信号的尾部与后一帧信号的头部具有重叠量，从而建立相邻帧之间的相关性。In this embodiment, the analog audio signal to be processed can be intercepted into different segments with a predetermined duration as a frame signal, or a predetermined number of sampled data can be used as a frame signal after analog-to-digital conversion. The subsequent audio signal processing procedures are all It can be processed in units of one frame of signal. In order to ensure the continuity of the signal, there may be a certain overlap between adjacent frame signals, that is, the tail of the previous frame signal and the head of the next frame signal have an overlap amount, thereby establishing the correlation between adjacent frames.
步骤S102、对所述多路数字音频信号进行频域转换以获取多路频域数据。Step S102: Perform frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
在本实施例中，对多路数字音频信号中每一路数字音频信号分别进行频域转换，从而得到每一路数字音频信号对应的频域数据。本实施例中将多路数字音频信号进行频域转换以获取多路频域数据，进而可实现多路频域数据在频域上的融合。其中频域转换方法可采用傅里叶变换(如离散傅里叶变换)、拉普拉斯变换、Z变换等等，具体的频域转换过程此处不再赘述。In this embodiment, frequency domain conversion is performed on each digital audio signal in the multiple digital audio signals, so as to obtain frequency domain data corresponding to each digital audio signal. In this embodiment, the multiple channels of digital audio signals are subjected to frequency domain conversion to obtain multiple channels of frequency domain data, and the fusion of multiple channels of frequency domain data in the frequency domain can be realized. The frequency domain conversion method may adopt Fourier transform (such as discrete Fourier transform), Laplace transform, Z transform, etc. The specific frequency domain conversion process will not be repeated here.
步骤S103、根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据。Step S103: Determine frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data.
在本实施例中，由于各预处理电路对同一待处理的模拟音频信号以不同增益进行放大，因此每一路的音频信号具有不同的最大值和最小值，增益较大一路的音频信号的最大值和最小值均相对较大，而增益较小一路的音频信号的最大值和最小值均相对较小，而动态范围是无失真情况下音频信号的最大值和最小值的比值，本实施例中根据多路频域数据中的一路或至少两路目标频域数据确定频域融合数据，则可通过融合实现在融合后最终得到的音频信号的最大值和最小值的调节，例如较大声音可由增益相对较大的预处理电路提供，较小声音可由增益相对较小的预处理电路提供，从而可以提高录音***的动态***，具有较高的灵敏度、同时能够降低底噪。在本实施例中，声音的大小可由音频信号的能量特征信息进行衡量，例如模拟音频信号或数字音频信号的声压级，或者模拟音频信号或数字音频信号的幅值等。In this embodiment, since each pre-processing circuit amplifies the same analog audio signal to be processed with different gains, each audio signal has a different maximum and minimum value, and the maximum value of the audio signal with a larger gain And the minimum value are relatively large, and the maximum and minimum values of the audio signal with a smaller gain are relatively small, and the dynamic range is the ratio of the maximum and minimum values of the audio signal without distortion. In this embodiment Determine the frequency domain fusion data according to one or at least two channels of the target frequency domain data in the multi-channel frequency domain data, and the maximum and minimum values of the audio signal finally obtained after the fusion can be adjusted through the fusion. For example, the louder sound can be adjusted by A pre-processing circuit with a relatively large gain is provided, and a smaller sound can be provided by a pre-processing circuit with a relatively small gain, so that the dynamic system of the recording system can be improved, with higher sensitivity and lower noise at the same time. In this embodiment, the size of the sound can be measured by the energy feature information of the audio signal, such as the sound pressure level of the analog audio signal or digital audio signal, or the amplitude of the analog audio signal or digital audio signal.
步骤S104、将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。Step S104: Convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal.
在本实施例中，在得到频域融合数据后，即可将频域融合数据转换为时域音频信号，转换方法可采用傅里叶变换(如离散傅里叶变换)的逆变换、 拉普拉斯变换的逆变换、Z变换的逆变换等等，此处不再赘述。在完成转换后，可根据时域音频信号获取输出音频信号，其中根据时域音频信号获取输出音频信号过程中，可对时域音频信号进行压缩、降噪等操作。此外，若本实施例中以一帧信号为单位进行上述的音频信号处理流程，则根据时域音频信号获取输出音频信号过程中还需要将各帧信号进行拼接，建立相邻帧之间的相关性，具体的，可将当前帧时域音频信号与前一帧前帧时域音频信号进行叠加处理。In this embodiment, after the frequency domain fusion data is obtained, the frequency domain fusion data can be converted into a time domain audio signal. The conversion method can adopt the inverse transformation of the Fourier transform (such as the discrete Fourier transform), the Lap The inverse transform of the Lass transform, the inverse transform of the Z transform, etc., will not be repeated here. After the conversion is completed, the output audio signal can be obtained according to the time domain audio signal. In the process of obtaining the output audio signal according to the time domain audio signal, operations such as compression and noise reduction can be performed on the time domain audio signal. In addition, if the audio signal processing procedure described above is performed in the unit of a frame signal in this embodiment, the process of obtaining the output audio signal according to the time domain audio signal also needs to splice the signals of each frame to establish the correlation between adjacent frames. Specifically, the time domain audio signal of the current frame and the time domain audio signal of the previous frame can be superimposed.
本实施例的音频信号处理方法，通过利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；对所述多路数字音频信号进行频域转换以获取多路频域数据；根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。本实施例的方法可实现有效的提高录音***的动态范围，具有较高的灵敏度，同时能够降低底噪，满足高信噪比的要求。The audio signal processing method of this embodiment uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog audio signal. Digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data One or at least two channels of target frequency domain data determine frequency domain fusion data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal. The method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor and meet the requirement of high signal-to-noise ratio.
在上述任一实施例的基础上，所述音频信号处理方法还包括：On the basis of any of the foregoing embodiments, the audio signal processing method further includes:
获取所述待处理的模拟音频信号的能量特征信息。Obtain energy feature information of the analog audio signal to be processed.
在本实施例中，待处理的模拟音频信号的能量特征信息可以为模拟音频信号的声压级或者模拟音频信号的幅值等，具体可以为模拟音频信号的瞬时幅度的最大值、最小值或中间值，也可以为模拟音频信号短时(预设时长)平均幅度的最大值、最小值或中间值。In this embodiment, the energy characteristic information of the analog audio signal to be processed may be the sound pressure level of the analog audio signal or the amplitude of the analog audio signal, etc., and specifically may be the maximum, minimum, or instantaneous amplitude of the analog audio signal. The intermediate value may also be the maximum, minimum or intermediate value of the average amplitude of the analog audio signal in a short time (preset duration).
进一步的，如图3所示，步骤S103所述的根据所述多路频域数据中的一路或至少两路频域数据确定频域融合数据，包括：Further, as shown in FIG. 3, the step S103 of determining frequency domain fusion data according to one or at least two channels of frequency domain data in the multiple channels of frequency domain data includes:
步骤S201、根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据；Step S201: Determine one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information;
步骤S202、根据所述一路或至少两路目标频域数据确定频域融合数据。Step S202: Determine frequency domain fusion data according to the one or at least two channels of target frequency domain data.
在本实施例中，在进行融合时可根据待处理的模拟音频信号的能量特征信息从多路频域数据中确定待融合的一路或至少两路目标频域数据，例如可根据待处理的模拟音频信号的能量特征信息的大小确定目标频域数据的数量， 如能量特征信息越大、则目标频域数据的数量越多；此外也可根据能量特征信息与预设处理电路的参考能量特征参数进行比较、再确定待融合的一路或至少两路目标频域数据，其中多个预设处理电路的每一个都对应一个各不相同的参考能量特征参数，参考能量特征参数是由预处理电路包括的放大电路的模拟增益决定的，参考能量特征参数与能量特征信息属于相同的参数，也即为预处理电路输出的数字音频信号的声压级或者幅值等，具体可以为在不失真情况下的数字音频信号的瞬时幅度的最大值、最小值或中间值，也可以为数字音频信号短时(预设时长)平均幅度的最大值、最小值或中间值，当预处理电路包括的放大电路的模拟增益越大，则对应的参考能量特征参数越大。In this embodiment, during fusion, one or at least two channels of target frequency domain data to be fused can be determined from multiple channels of frequency domain data according to the energy characteristic information of the analog audio signal to be processed, for example, according to the simulation to be processed The size of the energy feature information of the audio signal determines the number of target frequency domain data. For example, the greater the energy feature information, the greater the number of target frequency domain data; in addition, the reference energy feature parameters of the processing circuit can also be preset based on the energy feature information Compare and then determine one or at least two channels of target frequency domain data to be fused. Each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter. The reference energy characteristic parameter is included by the preprocessing circuit Determined by the analog gain of the amplifier circuit, the reference energy characteristic parameter and the energy characteristic information belong to the same parameter, that is, the sound pressure level or amplitude of the digital audio signal output by the preprocessing circuit, which can be in the case of no distortion The maximum, minimum or intermediate value of the instantaneous amplitude of the digital audio signal, or the maximum, minimum or intermediate value of the average amplitude of the digital audio signal in a short time (preset duration), when the preprocessing circuit includes the amplifier circuit The larger the analog gain of, the larger the corresponding reference energy characteristic parameter.
更具体的，在一种可选实施例中，所述根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据，包括：More specifically, in an optional embodiment, the determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information includes:
根据所述能量特征信息和多个参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data. The analog gain of the amplifier circuit included in each preprocessing circuit is determined.
在本实施例是从多路频域数据中确定至少两路目标频域数据的情况，其中第一目标频域数据可以仅仅为一路目标频域数据，当然也可为不止一路目标频域数据；同样的，第二目标频域数据也可以仅仅为一路目标频域数据，也可为不止一路目标频域数据。其中，如图4所示，所述根据所述能量特征信息从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，具体可包括：In this embodiment, at least two channels of target frequency domain data are determined from multiple channels of frequency domain data, where the first target frequency domain data may be only one channel of target frequency domain data, and of course, it may also be more than one channel of target frequency domain data; Similarly, the second target frequency domain data may also be only one channel of target frequency domain data or more than one channel of target frequency domain data. Wherein, as shown in FIG. 4, determining the first target frequency domain data and the second target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information may specifically include:
步骤S301、从多个参考能量特征参数中确定与所述能量特征信息相邻的第一参考能量特征参数和第二参考能量特征参数；Step S301: Determine a first reference energy characteristic parameter and a second reference energy characteristic parameter adjacent to the energy characteristic information from a plurality of reference energy characteristic parameters;
步骤S302、根据第一参考能量特征参数和第二参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据；Step S302: Determine first target frequency domain data and second target frequency domain data from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter;
其中，第一目标频域数据和第二目标频域数据分别是对所述多路数字音频信号中的第一数字音频信号和第二数字音频信号进行所述频域转换得到的，所述第一数字音频信号和第二数字音频信号分别是由第一参考能量特征参数和第二参考能量特征参数对应的第一预处理电路和第二预处理电路对所述待处理的模拟音频信号得到的。Wherein, the first target frequency domain data and the second target frequency domain data are obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, respectively. A digital audio signal and a second digital audio signal are respectively obtained from the analog audio signal to be processed by a first preprocessing circuit and a second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter .
步骤S303、根据所述第一目标频域数据和所述第二目标频域数据确定频域融合数据。Step S303: Determine frequency domain fusion data according to the first target frequency domain data and the second target frequency domain data.
在本实施例中，从多个预设处理电路的参考能量特征参数(以L 1、L 2、…、L I表示，其中I为预设处理电路的个数)选择与待处理的模拟音频信号的能量特征信息(以L c表示)相邻的第一参考能量特征参数(以L i′表示，其中1≤i′≤I-1)和第二参考能量特征参数(以L i′+1表示)，也即L c处于L i′和L i′+1之间，其中第一参考能量特征参数L i′对应第一预处理电路，该第一预处理电路输出的第一数字音频信号经频域转换得到的频域数据为所述第一目标频域数据，而第二参考能量特征参数L i′+1对应第二预处理电路，该第二预处理电路输出的第二数字音频信号经频域转换得到的频域数据为所述第二目标频域数据，进而可以根据第一目标频域数据和第二目标频域数据进行叠加运算即可获取频域融合数据。 In this embodiment, the reference energy characteristic parameters (indicated by L 1 , L 2 , ..., L I , where I is the number of preset processing circuits) of multiple preset processing circuits are selected and the analog audio to be processed The energy feature information of the signal (represented by L c ) is adjacent to the first reference energy feature parameter (represented by Li ′ , where 1≤i′≤I-1) and the second reference energy feature parameter (represented by L i′+ 1 means), that is, L c is between L i′ and L i′+1 , where the first reference energy characteristic parameter L i′ corresponds to the first preprocessing circuit, and the first digital audio output by the first preprocessing circuit The frequency domain data obtained by frequency domain conversion of the signal is the first target frequency domain data, and the second reference energy characteristic parameter Li′+1 corresponds to the second preprocessing circuit, and the second digital output from the second preprocessing circuit The frequency domain data obtained by frequency domain conversion of the audio signal is the second target frequency domain data, and the frequency domain fusion data can be obtained by performing superposition operation according to the first target frequency domain data and the second target frequency domain data.
当然，本实施例中可在确定第一参考能量特征参数L i和第二参考能量特征参数L i′+1后，以L i′和L i′-1(此时需要i′＞1)对应的预处理电路输出的数字音频信号经频域转换得到的频域数据为所述第一目标频域数据，同样可以L i′+1和L i′+2(此时需要i′＜I-1)对应的预处理电路输出的数字音频信号经频域转换得到的频域数据为所述第二目标频域数据，当然第一目标频域数据和第二目标频域数据也可分别包括更多路的频域数据，此处不再举例。 Of course, the present embodiment may be a second reference parameter L i and L i the energy characteristic parameters determining a first reference energy characteristics' + after 1 to L i 'and L i'-1 (this time need i'> 1) The frequency domain data obtained by the frequency domain conversion of the digital audio signal output by the corresponding preprocessing circuit is the first target frequency domain data, which can also be L i′+1 and L i′+2 (in this case, i′<I -1) The frequency domain data obtained by frequency domain conversion of the digital audio signal output by the corresponding preprocessing circuit is the second target frequency domain data. Of course, the first target frequency domain data and the second target frequency domain data may also include For more channels of frequency domain data, we will not give an example here.
在另一个可选实施例中，如图5所述，所述根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据，包括：In another optional embodiment, as shown in FIG. 5, the determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information includes:
步骤S401、当所述所述能量特征信息小于所述多个参考能量特征参数中最小的第三参考能量特征参数时，根据所述第三参考能量特征参数从所述多路频域数据中确定第三目标频域数据；Step S401: When the energy feature information is less than the smallest third reference energy feature parameter among the multiple reference energy feature parameters, determine from the multiple channels of frequency domain data according to the third reference energy feature parameter Third target frequency domain data;
其中，第三目标频域数据是对所述多路数字音频信号中的第三数字音频信号进行所述频域转换得到的，所述第三数字音频信号是由第三参考能量特征参数对应的第三预处理电路对所述待处理的模拟音频信号得到的；Wherein, the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is corresponding to a third reference energy characteristic parameter Obtained by the third preprocessing circuit from the analog audio signal to be processed;
步骤S402、根据第三目标频域数据获取所述频域融合数据。Step S402: Acquire the frequency domain fusion data according to the third target frequency domain data.
在本实施例中，若多个预设处理电路的参考能量特征参数(L 1、L 2、…、L I)中，最小的第三参考能量特征参数L 1大于与待处理的模拟音频信号的能量特征信息L c，也即L c小于L 1，其中第三参考能量特征参数L 1对应第三预 处理电路输出的第三数字音频信号经频域转换得到的频域数据为所述第三目标频域数据，进而可以根据第三目标频域数据获取所述频域融合数据。 In this embodiment, if among the reference energy characteristic parameters (L 1 , L 2 , ..., L I ) of the plurality of preset processing circuits, the smallest third reference energy characteristic parameter L 1 is greater than that of the analog audio signal to be processed The energy feature information L c , that is, L c is less than L 1 , where the third reference energy feature parameter L 1 corresponds to the third digital audio signal output by the third preprocessing circuit and the frequency domain data obtained by frequency domain conversion is the first Three target frequency domain data, and then the frequency domain fusion data can be obtained according to the third target frequency domain data.
当然，本实施例中可在确定待处理的模拟音频信号的能量特征信息L c小于第三参考能量特征参数L 1后，以L 1及L 2对应的预处理电路输出的数字音频信号经频域转换得到的频域数据为所述第三目标频域数据，当然第三目标频域数据也可分别包括更多路的频域数据，此处不再举例。 Of course, the present third embodiment with reference to the energy parameters of L 1 may be smaller than in the embodiment determined to be processed analog audio signal energy characteristic information L c, the digital audio signals L 1 and L 2 corresponding to an output of the preprocessing circuit frequency- The frequency domain data obtained by the domain conversion is the third target frequency domain data. Of course, the third target frequency domain data may also include more channels of frequency domain data, and no examples are given here.
在另一个可选实施例中，如图6所示，所述对所述多路频域数据进行融合以获取频域融合数据，还包括：In another optional embodiment, as shown in FIG. 6, the fusion of the multiple channels of frequency domain data to obtain frequency domain fusion data further includes:
步骤S501、当所述所述能量特征信息大于所述多个参考能量特征参数中最大的第四参考能量特征参数时，根据所述第四参考能量特征参数从所述多路频域数据中确定第四目标频域数据；Step S501: When the energy feature information is greater than the largest fourth reference energy feature parameter among the multiple reference energy feature parameters, determine from the multiple channels of frequency domain data according to the fourth reference energy feature parameter The fourth target frequency domain data;
其中，第四目标频域数据是对所述多路数字音频信号中的第四数字音频信号进行所述频域转换得到的，所述第四数字音频信号是由第四参考能量特征参数对应的第四预处理电路对所述待处理的模拟音频信号得到的；Wherein, the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multi-channel digital audio signal, and the fourth digital audio signal is corresponding to a fourth reference energy characteristic parameter Obtained by the fourth preprocessing circuit from the analog audio signal to be processed;
步骤S502、根据第四目标频域数据获取所述频域融合数据。Step S502: Acquire the frequency domain fusion data according to the fourth target frequency domain data.
在本实施例中，若多个预设处理电路的参考能量特征参数(L 1、L 2、…、L I)中，最大的第四参考能量特征参数L I小于与待处理的模拟音频信号的能量特征信息L c，也即L c大于L I，其中第四参考能量特征参数L I对应第四预处理电路输出的第四数字音频信号经频域转换得到的频域数据为所述第四目标频域数据，进而可以根据第四目标频域数据获取所述频域融合数据。 In this embodiment, if among the reference energy characteristic parameters (L 1 , L 2 , ..., L I ) of the plurality of preset processing circuits, the largest fourth reference energy characteristic parameter L I is smaller than that of the analog audio signal to be processed The energy feature information L c , that is, L c is greater than L I , where the fourth reference energy feature parameter L I corresponds to the fourth digital audio signal output by the fourth preprocessing circuit and the frequency domain data obtained by frequency domain conversion is the first Four target frequency domain data, and the frequency domain fusion data can be obtained according to the fourth target frequency domain data.
当然，本实施例中可在确定待处理的模拟音频信号的能量特征信息L c大于第四参考能量特征参数L I后，以L I及L I-1对应的预处理电路输出的数字音频信号经频域转换得到的频域数据为所述第四目标频域数据，当然第四目标频域数据也可分别包括更多路的频域数据，此处不再举例。 Of course, in this embodiment, after determining that the energy characteristic information L c of the analog audio signal to be processed is greater than the fourth reference energy characteristic parameter L I , the digital audio signal output by the preprocessing circuit corresponding to L I and L I-1 The frequency domain data obtained by frequency domain conversion is the fourth target frequency domain data. Of course, the fourth target frequency domain data may also include more channels of frequency domain data respectively, and no examples are given here.
在上述实施例的基础上，在根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据时，可首先将待处理的模拟音频信号的能量特征信息L c与多个预设处理电路的参考能量特征参数进行比对，若L c小于(或等于)第三参考能量特征参数L 1，则执行步骤401-402；若L c大于(或等于)第四参考能量特征参数L I，则执行步骤501-502；若L c处于相邻的第三参考能量特征参数L 1和第四参考能量特征参数L I之间，则执行步骤 301-303。 On the basis of the foregoing embodiment, when determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy feature information, the energy feature information L of the analog audio signal to be processed may be first c is compared with the reference energy characteristic parameters of multiple preset processing circuits, if L c is less than (or equal to) the third reference energy characteristic parameter L 1 , then steps 401-402 are executed; if L c is greater than (or equal to) the first For the fourth reference energy characteristic parameter L I , steps 501 to 502 are executed; if L c is between the adjacent third reference energy characteristic parameter L 1 and the fourth reference energy characteristic parameter L I , then steps 301-303 are executed.
在上述任一实施例的基础上，所述根据所述一路或至少两路目标频域数据确定频域融合数据，包括：On the basis of any of the foregoing embodiments, the determining frequency domain fusion data according to the one or at least two channels of target frequency domain data includes:
对所述至少两路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the at least two channels of frequency domain data to obtain frequency domain fusion data.
在本实施例中，对于至少两路目标频域数据进行融合时，采用对频域数据进行频谱叠加运算的方式，获取频域融合数据。In this embodiment, when at least two channels of target frequency domain data are fused, the frequency domain fusion data is obtained by performing a spectrum superposition operation on the frequency domain data.
进一步的，当所述一路或者多路频域数据包括第一目标频域数据和第二目标频域数据时，则所述根据所述一路或者多路频域数据获取频域融合数据，包括：Further, when the one or more channels of frequency domain data include the first target frequency domain data and the second target frequency domain data, the acquiring frequency domain fusion data according to the one or more channels of frequency domain data includes:
根据第一目标频域数据和第二目标频域数据对应的权重对所述第一路和第二路路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the first channel and the second channel of frequency domain data according to the weights corresponding to the first target frequency domain data and the second target frequency domain data to obtain frequency domain fusion data.
在本实施例中，进行叠加运算是可对第一目标频域数据和第二目标频域数据设置权重，通过将第一目标频域数据和第二目标频域数据以不同的权重进行叠加运算，则可得到不同动态范围的频域融合数据。In this embodiment, performing the superposition operation can set weights for the first target frequency domain data and the second target frequency domain data, by superposing the first target frequency domain data and the second target frequency domain data with different weights. , You can get frequency domain fusion data with different dynamic ranges.
进一步的，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的，其中，所述第一目标频域数据和第二目标频域数据对应的权重是根据所述多个预设处理电路中与所述第一目标频域数据对应的第一预处理电路和与所述第二目标频域数据对应的第二预处理电路的参考能量特征参数确定。Further, each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter, and the plurality of reference energy characteristic parameters are determined according to the analog gain of the amplifying circuit included in the plurality of preprocessing circuits , Wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the first preprocessing circuit and the first preprocessing circuit corresponding to the first target frequency domain data among the plurality of preset processing circuits The reference energy characteristic parameter of the second preprocessing circuit corresponding to the second target frequency domain data is determined.
在本实施例中，第一目标频域数据和第二目标频域数据对应的权重由对应的预处理电路的参考能量特征参数来确定，更具体的，可根据第一目标频域数据对应的第一参考能量特征参数L i以及第二目标频域数据对应的第二参考能量特征参数L i+1与待处理的模拟音频信号的能量特征信息L c之间的大小关系确定，其中参考能量特征参数约接近模拟音频信号的能量特征信息L c的，权重越大，例如，L c更接近L i，则说明模拟音频信号更接近L i对应的预处理电路的数字音频信号，则需要增加L i对应的第一目标频域数据的权重。本实施例中可通过如下公式确定第一目标频域数据的权重a 1和第二目标频域数据的权重a 2： In this embodiment, the weights corresponding to the first target frequency domain data and the second target frequency domain data are determined by the reference energy characteristic parameters of the corresponding preprocessing circuit. More specifically, the weights corresponding to the first target frequency domain data a first reference energy characteristic parameters L i and a second target frequency domain data corresponding to the second reference feature parameter L i + energy magnitude relation between the energy characteristic information L c 1 and the analog audio signal to be processed is determined, wherein the reference energy characteristic parameters of the analog audio signal closely approximates the energy characteristic information L c, the larger the weight, e.g., closer to L i L c, then the analog audio signal is closer to a digital audio signal L i corresponds to the pre-processing circuit, it is necessary to increase The weight of the first target frequency domain data corresponding to L i . In this embodiment, the weight a 1 of the first target frequency domain data and the weight a 2 of the second target frequency domain data can be determined by the following formula:
Figure PCTCN2019083006-appb-000001
Figure PCTCN2019083006-appb-000001
Figure PCTCN2019083006-appb-000002
Figure PCTCN2019083006-appb-000002
在上述任一实施例的基础上，如图7所示，所述根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据，包括：On the basis of any of the foregoing embodiments, as shown in FIG. 7, the determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data includes:
步骤S601、根据所述一路或至少两路目标频域数据中每一路对应的压缩系数对所述一路或至少两路目标频域数据进行压缩处理；Step S601: Perform compression processing on the one or at least two channels of target frequency domain data according to the compression coefficient corresponding to each of the one or at least two channels of target frequency domain data;
步骤S602、根据所述压缩处理之后的一路或者多路频域数据获取频域融合数据。Step S602: Acquire frequency domain fusion data according to one or more channels of frequency domain data after the compression processing.
在本实施例中，由于录音***的输出是存在数字量化范围的，也即对于音频信号的最大幅值和最小幅值具有限制，音频信号的最大幅值不能大于最大阈值，音频信号的最小幅值不能小于最小阈值，因此在获取频域融合数据时需要对目标频域数据进行压缩处理，以避免融合后的频域融合数据超过数字量化范围。本实施例中，对目标频域数据进行压缩的步骤可与上述的叠加运算同时进行，也可在叠加运算前完成。In this embodiment, since the output of the recording system has a digital quantization range, that is, there are restrictions on the maximum amplitude and minimum amplitude of the audio signal, the maximum amplitude of the audio signal cannot be greater than the maximum threshold, and the minimum amplitude of the audio signal The value cannot be less than the minimum threshold, so when acquiring the frequency domain fusion data, it is necessary to compress the target frequency domain data to prevent the fusion frequency domain fusion data from exceeding the digital quantization range. In this embodiment, the step of compressing the target frequency domain data can be performed simultaneously with the above-mentioned superposition operation, or it can be completed before the superposition operation.
进一步的，所述压缩处理为线性压缩处理。Further, the compression processing is linear compression processing.
在上述实施例的基础上，所述一路或者多路频域数据每一路频域数据对应的压缩系数是根据与所述每一路频域数据对应的预处理电路包括的放大器的模拟增益确定的。On the basis of the foregoing embodiment, the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is determined according to the analog gain of the amplifier included in the preprocessing circuit corresponding to each channel of frequency domain data.
在本实施例中，任一路频域数据对应的压缩系数可以为对应的预处理电路的通道均衡参数和缩放系数的乘积，其中某一预处理电路作为参考预处理电路，任一路预处理电路的通道均衡参数为该预处理电路模拟增益与参考预处理电路模拟增益的比值，缩放系数则根据该预处理电路输出的数字音频信号的大小获取。本实施例中可通过如下公式获取任一路频域数据对应的压缩系数：In this embodiment, the compression coefficient corresponding to any channel of frequency domain data may be the product of the channel equalization parameter and the scaling factor of the corresponding preprocessing circuit. A certain preprocessing circuit is used as a reference preprocessing circuit, and the The channel equalization parameter is the ratio of the analog gain of the preprocessing circuit to the analog gain of the reference preprocessing circuit, and the scaling factor is obtained according to the size of the digital audio signal output by the preprocessing circuit. In this embodiment, the compression coefficient corresponding to any channel of frequency domain data can be obtained by the following formula:
Figure PCTCN2019083006-appb-000003
Figure PCTCN2019083006-appb-000003
其中，
Figure PCTCN2019083006-appb-000004
为通道均衡参数，用于将第i′路预处理电路的频域数据与参考预处理电路的频域数据进行幅值均衡，G i′为预处理电路的模拟增益，G ref为参考预处理电路的模拟增益，α为缩放系数，用于对频域数据进行幅值缩放，一般而言，对于小信号，α≥1，使得信号保持或放大；对于大信号，α＜1， 使得信号缩小，从而达到压缩动态范围的目的。 among them,
Figure PCTCN2019083006-appb-000004
Is the channel equalization parameter, used to perform amplitude equalization between the frequency domain data of the i′th preprocessing circuit and the frequency domain data of the reference preprocessing circuit, G i′ is the analog gain of the preprocessing circuit, and G ref is the reference preprocessing The analog gain of the circuit, α is the scaling factor, used to scale the frequency domain data. Generally speaking, for small signals, α≥1, so that the signal is maintained or amplified; for large signals, α<1, so that the signal is reduced , So as to achieve the purpose of compressing the dynamic range.
需要说明的是，步骤S601-602可以仅在频域融合数据超出数字量化范围时执行，也即可在步骤S601前判断频域融合数据是否存在超出数字量化范围的可能性，若存在超出数字量化范围的可能性时，才执行步骤S601-602。It should be noted that steps S601-602 can be executed only when the frequency domain fusion data exceeds the digital quantization range, and it can also be judged before step S601 whether the frequency domain fusion data has the possibility of exceeding the digital quantization range. When the range is possible, steps S601-602 are executed.
在上述任一实施例的基础上，所述时域音频信号为当前帧时域音频信号，其中，上述实施例中步骤S104所述的根据所述时域音频信号获取输出音频信号，包括：On the basis of any of the foregoing embodiments, the time domain audio signal is the current frame time domain audio signal, wherein the step S104 in the foregoing embodiment of obtaining the output audio signal according to the time domain audio signal includes:
将所述当前帧时域音频信号与当前帧时域音频信号之前获取的历史帧时域音频信号进行叠加处理以获取当前帧时域融合音频信号；Superimposing the current frame time domain audio signal with the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal;
根据所述当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame.
在本实施例中，以一帧信号为单位进行上述的各音频信号处理流程，为了保证信号的连续性，相邻帧信号之间可具有一定的重叠，即前一帧信号的尾部与后一帧信号的头部具有重叠量，从而建立相邻帧之间的相关性。因此在通过S104中将当前帧频域融合数据转换为当前帧时域音频信号后，可将当前帧时域音频信号与前一帧时域音频信号中重叠部分进行重叠叠加运算，而当前帧时域音频信号与前一帧时域音频信号中未重叠部分则不进行叠加运算，从而得到当前帧时域融合音频信号，进而可根据所述当前帧时域融合音频信号确定所述输出音频信号。In this embodiment, the above-mentioned audio signal processing procedures are performed in units of one frame signal. In order to ensure the continuity of the signal, there may be a certain overlap between adjacent frame signals, that is, the tail of the previous frame signal and the next The header of the frame signal has an overlap amount, thereby establishing correlation between adjacent frames. Therefore, after converting the frequency domain fusion data of the current frame into the time domain audio signal of the current frame in S104, the overlapped part of the time domain audio signal of the current frame and the time domain audio signal of the previous frame can be overlapped and superimposed. The non-overlapping part of the time domain audio signal and the previous frame of time domain audio signal is not superimposed to obtain the current frame time domain fused audio signal, and the output audio signal can be determined according to the current frame time domain fused audio signal.
在上述实施例的基础上，如图8所示，所述根据所述当前帧时域融合音频信号确定所述输出音频信号，具体可包括：On the basis of the foregoing embodiment, as shown in FIG. 8, the determining the output audio signal according to the time-domain fusion audio signal of the current frame may specifically include:
步骤S701、根据预设的压缩系数对所述根据所述当前帧时域融合音频信号进行压缩处理；Step S701: Perform compression processing on the time-domain fused audio signal according to the current frame according to a preset compression coefficient;
步骤S702、根据所述压缩处理之后的当前帧时域融合音频信号确定所述输出音频信号。Step S702: Determine the output audio signal according to the time-domain fusion audio signal of the current frame after the compression processing.
在本实施例中，为了避免输出音频信号超出数字量化范围，可对当前帧时域融合音频信号进行动态范围压缩，其中预设的压缩系数可来源于预设的非线性函数，也即将当前帧时域融合音频信号的幅值根据非线性函数进行非线性映射，从而使较小的信号保持不变或放大，较大的信号压缩到数字量化范围。典型的非线性函数如图9曲线所示，将当前帧时域融合音频信号的幅值的绝对值|x syn|作为横坐标输入非线性函数，而非线性函数的纵坐标函数值 即为压缩后的幅值的绝对值|y|＝f(|x syn|)，从而使输出音频信号不超出数字量化范围L max。 In this embodiment, in order to prevent the output audio signal from exceeding the digital quantization range, the time-domain fusion audio signal of the current frame may be subjected to dynamic range compression, where the preset compression coefficient may be derived from a preset nonlinear function, that is, the current frame The amplitude of the time-domain fusion audio signal is non-linearly mapped according to the non-linear function, so that the smaller signal remains unchanged or amplified, and the larger signal is compressed to the digital quantization range. A typical nonlinear function is shown in the curve in Figure 9. The absolute value of the amplitude of the current frame time domain fusion audio signal |x syn | is input as the abscissa to the nonlinear function, and the ordinate function value of the nonlinear function is the compression The absolute value of the subsequent amplitude |y|=f(|x syn |), so that the output audio signal does not exceed the digital quantization range L max .
在上述任一实施例的基础上，所述对所述多路数字音频信号进行频域转换以获取多路频域数据，包括：On the basis of any of the foregoing embodiments, the performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data includes:
对所述多路数字音频信号进行加窗处理；Performing windowing processing on the multi-channel digital audio signal;
根据加窗后的多路数字音频信号进行频域转换以获取多路频域数据。Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
在本实施例中，采用分析窗函数对各路数字音频信号进行加窗处理，其中分析窗函数包括但不限于正弦窗函数，矩形窗函数，三角窗函数，汉宁窗函数(Hanning)，高斯窗函数(Gaussian)，布莱克曼窗函数(Blackman)，切比雪夫窗函数(Chebyshev)，汉明窗函数(Hamming)，平顶窗函数(Flap Top)，凯塞窗函数(Kaiser)。当然，也可不进行加窗处理，也即相当于加了矩形窗。In this embodiment, an analysis window function is used to perform windowing processing on each channel of digital audio signals. The analysis window function includes but is not limited to sine window function, rectangular window function, triangular window function, Hanning window function (Hanning), Gaussian window function. Window function (Gaussian), Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), Flat top window function (Flap Top), Kaiser window function (Kaiser). Of course, windowing is not required, which is equivalent to adding a rectangular window.
进一步的，所述根据所述时域音频信号获取输出音频信号，包括：Further, the obtaining the output audio signal according to the time domain audio signal includes:
对所述时域音频信号进行加窗处理；Windowing the time domain audio signal;
根据所述加窗处理之后的时域音频信号确定所述输出音频信号。The output audio signal is determined according to the time domain audio signal after the windowing process.
在本实施例中，采用合成窗函数对时域音频信号进行加窗处理，其中合成窗函数包括但不限于正弦窗函数，矩形窗函数，三角窗函数，汉宁窗函数(Hanning)，高斯窗函数(Gaussian)，布莱克曼窗函数(Blackman)，切比雪夫窗函数(Chebyshev)，汉明窗函数(Hamming)，平顶窗函数(Flap Top)，凯塞窗函数(Kaiser)。当然，也可不进行加窗处理，也即相当于加了矩形窗。In this embodiment, a synthesis window function is used to perform windowing processing on the time domain audio signal. The synthesis window function includes, but is not limited to, a sine window function, a rectangular window function, a triangular window function, a Hanning window function (Hanning), and a Gaussian window. Function (Gaussian), Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser). Of course, windowing is not required, which is equivalent to adding a rectangular window.
本发明实施例提供一种音频信号处理方法。图10为本发明另一实施例提供的音频信号处理方法的流程图。如图10所示，在上述实施例的基础上，本实施例中的方法，可以包括：The embodiment of the present invention provides an audio signal processing method. FIG. 10 is a flowchart of an audio signal processing method provided by another embodiment of the present invention. As shown in FIG. 10, on the basis of the foregoing embodiment, the method in this embodiment may include:
步骤S801、利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号。Step S801: Use multiple pre-processing circuits to process analog audio signals to be processed to obtain multiple digital audio signals.
其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同。Wherein, each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter, and the analog gains of the amplifiers of each preprocessing circuit are different.
预处理的模拟音频信号x(t)，t为采样离散时间序列t＝1,2,…，采样频率为f s，将预处理的模拟音频信号x(t)输入并联的多路数字音频信号中，其中第i路预处理电路包括的放大器的模拟增益分别为G i，其中，i＝1,2,…,I，I为 预处理电路的个数，在模拟音频信号传输、放大、模数转换过程中引入综合噪声，其中各路的噪声分别为n i(t)，也即各通道输出的数字音频信号x i(t)如下： The preprocessed analog audio signal x(t), t is the sampling discrete time sequence t=1, 2,..., the sampling frequency is f s , and the preprocessed analog audio signal x(t) is input to the parallel multi-channel digital audio signal Among them, the analog gains of the amplifiers included in the i-th preprocessing circuit are respectively G i , where i = 1, 2, ..., I, and I is the number of preprocessing circuits. In analog audio signal transmission, amplification, and analog The integrated noise is introduced in the digital conversion process, and the noise of each channel is n i (t), that is, the digital audio signal x i (t) output by each channel is as follows:
x i(t)＝G ix(t)+n i(t)   i＝1,2,…,I x i (t)=G i x(t)+n i (t) i=1,2,...,I
步骤S802、对多路数字音频信号中的每一数字音频信号每间隔M个采样点提取N个采样点，作为一帧数字音频信号。Step S802: Extract N sampling points every M sampling points of each digital audio signal in the multi-channel digital audio signal, and use them as a frame of digital audio signal.
在本实施例中，第i路预处理电路的第l帧数字音频信号记作：In this embodiment, the digital audio signal of the first frame of the i-th preprocessing circuit is denoted as:
x i(t) l，t＝1,2,…,N。 x i (t) l , t=1, 2,...,N.
其中，N称为帧长，M称为帧移，0＜M＜N，也即第l-1帧的最后N-M个采样点与第l帧的前N-M个采样点相同，0.005f s＜N＜f s，且N为2的幂次方。 Among them, N is the frame length, M is the frame shift, 0＜M＜N, that is, the last NM sampling points of the l-1th frame are the same as the first NM sampling points of the lth frame, 0.005f s ＜N <f s , and N is a power of 2.
步骤S803、对所述多路数字音频信号进行加窗处理。Step S803: Perform windowing processing on the multiple channels of digital audio signals.
本实施例中，对第i路预处理电路的第l帧数字音频信号x i(t) l进行加窗处理，加窗后的数字音频信号为： In this embodiment, the first frame of digital audio signal x i (t) l of the i-th preprocessing circuit is windowed, and the windowed digital audio signal is:
x′ i(t) l＝x i(t) lh ana(t)   t＝1,2,…,N x′ i (t) l = x i (t) l h ana (t) t=1,2,...,N
其中，h ana(t)为N点分析窗函数，常用的窗函数可以包括但不限于如正弦窗，矩形窗函数，三角窗函数，汉宁窗(Hanning)，高斯窗函数(Gaussian)，布莱克曼窗函数(Blackman)，切比雪夫窗函数(Chebyshev)，汉明窗函数(Hamming)，平顶窗函数(Flap Top)，凯塞窗函数(Kaiser)等。 Among them, h ana (t) is the N-point analysis window function. Commonly used window functions can include but are not limited to sine window, rectangular window function, triangular window function, Hanning window, Gaussian window function, Black Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser) and so on.
步骤S804、根据加窗后的多路数字音频信号进行频域转换以获取多路频域数据。Step S804: Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
本实施例中，对第i路预处理电路的第l帧加窗后的数字音频信号x′ i(t) l进行离散傅里叶变换，得到第i路频域数据X i(n) l： In this embodiment, the digital audio signal of the l-frame windowing preprocessing circuit path i x 'i (t) l discrete Fourier transform, the i-th channel frequency domain data X i (n) l :
Figure PCTCN2019083006-appb-000005
Figure PCTCN2019083006-appb-000005
其中，n为离散频谱序列，e为自然常数，
Figure PCTCN2019083006-appb-000006
为虚数单位。本实施例中可用快速傅里叶变换进行加速计算。 Among them, n is the discrete spectrum sequence, e is the natural constant,
Figure PCTCN2019083006-appb-000006
It is an imaginary unit. In this embodiment, fast Fourier transform can be used to accelerate calculation.
步骤S805、根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据。Step S805: Determine frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data.
在本实施例中，可将多路频域数据进行谱融合，可采用多路频域数据的 频谱信号以预定的权重进行线性叠加，叠加逻辑为动态范围较好的覆盖当前帧的模拟音频信号的，则权重越高。频域融合数据为X syn(n) l： In this embodiment, multiple channels of frequency domain data can be spectrum fused, and the spectrum signals of multiple channels of frequency domain data can be used for linear superposition with a predetermined weight, and the superposition logic is an analog audio signal with a better dynamic range covering the current frame , The higher the weight. The frequency domain fusion data is X syn (n) l :
Figure PCTCN2019083006-appb-000007
Figure PCTCN2019083006-appb-000007
其中，α i为第i路预处理电路的频域数据对应的权重，α i≥0。 Among them, α i is the weight corresponding to the frequency domain data of the i-th preprocessing circuit, and α i ≥0.
作为一种可选实施例，根据待处理的模拟音频信号的能量特征信息L c以及各预设处理电路的参考能量特征参数L 1、L 2、…、L I，从多路频域数据中确定目标频域数据。假设G 1＞G 2＞…＞G I，则0＜L 1＜L 2＜…＜L I。本实施例中采用如下公式获取频域融合数据X syn(n) l： As an optional embodiment, according to the energy characteristic information L c of the analog audio signal to be processed and the reference energy characteristic parameters L 1 , L 2 , ..., L I of each preset processing circuit, from multiple channels of frequency domain data Determine the target frequency domain data. Assuming G 1 >G 2 >...>G I , then 0<L 1 <L 2 ＜...<L I. In this embodiment, the following formula is adopted to obtain frequency domain fusion data X syn (n) l :
Figure PCTCN2019083006-appb-000008
Figure PCTCN2019083006-appb-000008
其中，a 1和a 2分别为频域数据X i′(n) l和X i′+1(n) l的权重，可通过如下公式确定： Among them, a 1 and a 2 are the weights of the frequency domain data Xi ′ (n) l and Xi ′+1 (n) l , respectively, which can be determined by the following formula:
Figure PCTCN2019083006-appb-000009
Figure PCTCN2019083006-appb-000009
Figure PCTCN2019083006-appb-000010
Figure PCTCN2019083006-appb-000010
在上述公式中，为避免最终的输出音频信号超出录音***的数字量化范围可在叠加运算中对频域数据进行线性压缩，也即每一频域数据乘以对应的压缩系数，其中G x(L i′)为为频域数据X i′(n) i对应的压缩系数，可通过如下公式确定： In the above formula, in order to prevent the final output audio signal from exceeding the digital quantization range of the recording system, the frequency domain data can be linearly compressed in the superposition operation, that is, each frequency domain data is multiplied by the corresponding compression coefficient, where G x ( L i′ ) is the compression coefficient corresponding to the frequency domain data X i′ (n) i , which can be determined by the following formula:
Figure PCTCN2019083006-appb-000011
Figure PCTCN2019083006-appb-000011
其中，
Figure PCTCN2019083006-appb-000012
为通道均衡参数，用于将第i′路预处理电路的频域数据与参考预处理电路的频域数据进行幅值均衡，G i′为预处理电路的模拟增益，G ref为参考预处理电路的模拟增益，α为缩放系数，用于对第i′路预处理电路的频域数据进行幅值缩放，一般而言，对于小信号，α≥1，使得信号保持或放大；对于大信号，α＜1，使得信号缩小，从而达到压缩动态范围的目的。 among them,
Figure PCTCN2019083006-appb-000012
Is the channel equalization parameter, used to perform amplitude equalization between the frequency domain data of the i′th preprocessing circuit and the frequency domain data of the reference preprocessing circuit, G i′ is the analog gain of the preprocessing circuit, and G ref is the reference preprocessing The analog gain of the circuit, α is the scaling factor, which is used to scale the frequency domain data of the i'th preprocessing circuit. Generally speaking, for small signals, α≥1, so that the signal is maintained or amplified; for large signals ,Α<1, so that the signal is reduced, so as to achieve the purpose of compressing the dynamic range.
需要说明的是，对频域数据进行线性压缩并非必须执行的，也即上述公式中可令各压缩系数为1。It should be noted that it is not necessary to perform linear compression on frequency domain data, that is, each compression factor can be set to 1 in the above formula.
步骤S806、将所述频域融合数据转换为时域音频信号。Step S806: Convert the frequency domain fusion data into a time domain audio signal.
本实施例中，将第l帧频域融合数据X syn(n) l通过离散傅里叶变换的逆变换进行处理，得到第l帧时域音频信号x′ syn(t) l： In this embodiment, the first frame of frequency domain fusion data X syn (n) l is processed by the inverse transform of the discrete Fourier transform to obtain the first frame of time domain audio signal x'syn (t) l :
Figure PCTCN2019083006-appb-000013
Figure PCTCN2019083006-appb-000013
具体可采用快速傅里叶变换的逆变换进行加速计算。Specifically, the inverse transform of the fast Fourier transform can be used to accelerate the calculation.
步骤S807、对所述时域音频信号进行加窗处理。Step S807: Perform windowing processing on the time domain audio signal.
在本实施例中，将第l帧时域音频信号x′ syn(t) l进行加窗处理，得到加窗后的时域音频信号
Figure PCTCN2019083006-appb-000014
In this embodiment, the time domain audio signal x'syn (t) l of the l- th frame is windowed to obtain a windowed time domain audio signal
Figure PCTCN2019083006-appb-000014
Figure PCTCN2019083006-appb-000015
Figure PCTCN2019083006-appb-000015
其中，h syn(t)为N点合成窗函数，常用的合成窗函数可以包括但不限于如正弦窗，矩形窗函数，三角窗函数，汉宁窗(Hanning)，高斯窗函数(Gaussian)，布莱克曼窗函数(Blackman)，切比雪夫窗函数(Chebyshev)，汉明窗函数(Hamming)，平顶窗函数(Flap Top)，凯塞窗函数(Kaiser)等。 Among them, h syn (t) is an N-point synthesis window function. Commonly used synthesis window functions can include but are not limited to sine window, rectangular window function, triangular window function, Hanning window, Gaussian window function, Blackman window function (Blackman), Chebyshev window function (Chebyshev), Hamming window function (Hamming), flat top window function (Flap Top), Kaiser window function (Kaiser) and so on.
步骤S808、将所述当前帧时域音频信号与当前帧时域音频信号之前获取的历史帧时域音频信号进行叠加处理以获取当前帧时域融合音频信号。Step S808: Perform superposition processing on the current frame time domain audio signal and the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal.
在本实施例中，可将当前帧时域音频信号与前一帧时域音频信号中重叠部分进行重叠叠加运算，从而得到当前帧时域融合音频信号：In this embodiment, the overlapped part of the time domain audio signal of the current frame and the time domain audio signal of the previous frame may be overlapped and superimposed to obtain the time domain fusion audio signal of the current frame:
Figure PCTCN2019083006-appb-000016
Figure PCTCN2019083006-appb-000016
其中，
Figure PCTCN2019083006-appb-000017
为第l-1帧重叠累加N点时域信号；若l＝1，
Figure PCTCN2019083006-appb-000018
于是，第l帧重叠累加M点后，输出该重叠的M点对应的当前帧时域融合音频信号x syn(t) l为： among them,
Figure PCTCN2019083006-appb-000017
To overlap and accumulate N-point time-domain signals for the l-1th frame; if l=1,
Figure PCTCN2019083006-appb-000018
Therefore, after the l-th frame overlaps and accumulates M points, the current frame time-domain fusion audio signal x syn (t) l corresponding to the overlapped M points is output as:
Figure PCTCN2019083006-appb-000019
Figure PCTCN2019083006-appb-000019
步骤S809、对所述根据所述当前帧时域融合音频信号进行压缩处理；根据所述压缩处理之后的当前帧时域融合音频信号确定所述输出音频信号。Step S809: Perform compression processing on the time-domain fusion audio signal according to the current frame; determine the output audio signal according to the time-domain fusion audio signal of the current frame after the compression processing.
在本实施例中，若在步骤S805中未对频域数据进行动态范围压缩，或者经过步骤S805对频域数据的动态范围压缩后，x syn(t) l的幅值仍然超过数字量化范围，需要进行在时域上的动态范围压缩，以满足数字量化的需求。 In this embodiment, if the dynamic range compression of the frequency domain data is not performed in step S805, or after the dynamic range compression of the frequency domain data in step S805, the amplitude of x syn (t) l still exceeds the digital quantization range, It is necessary to perform dynamic range compression in the time domain to meet the needs of digital quantization.
本实施例中，动态范围压缩等效于时域信号幅值的非线性映射，基本原则是小信号保持不变或适当放大，大信号压缩到数字量化范围L max以内，其中L max为最大量化幅值。本实施例中可选择如图9曲线所示的典型的映射函数，输入横坐标为第l帧第l帧重叠的M点对应的当前帧时域融合音频信号|x syn(t) l|，输出纵坐标为f(|x syn(t) l|)。通过如下公式输出第l帧重叠的M点对应的压缩处理后的当前帧时域融合音频信号y(n) l： In this embodiment, the dynamic range compression is equivalent to the nonlinear mapping of the signal amplitude in the time domain. The basic principle is that the small signal remains unchanged or appropriately amplified, and the large signal is compressed to the digital quantization range L max , where L max is the maximum quantization. Amplitude. In this embodiment, a typical mapping function as shown in the curve in Fig. 9 can be selected, and the input abscissa is the current frame time-domain fusion audio signal |x syn (t) l | The output ordinate is f(|x syn (t) l |). Output the compressed and processed current frame time-domain fused audio signal y(n) l corresponding to the M points overlapped in the l-th frame by the following formula:
y(n) l＝sign(x syn(t) l)f(|x syn(t) l|)   n＝0,1,…,M-1 y(n) l = sign(x syn (t) l )f(|x syn (t) l |) n=0,1,…,M-1
其中，sign(·)为判断符号函数，正数为1，负数为‐1，零则为0。Among them, sign(·) is a function for judging the sign, a positive number is 1, a negative number is -1, and a zero is 0.
在本实施例中，若x syn(t) l的幅值未超过数字量化范围，即达到了量化范围的要求，可不执行步骤S809。需要说明的是，为确保输出信号满足量化范围，可以至少执行步骤S805中在频域的动态范围压缩和步骤S809中在时域的动态范围压缩中的任一种动态范围压缩。一般而言，频域的动态范围压缩为线性压缩，不改变声音的音色，而时域的动态范围压缩为非线性压缩，会有一定的音色失真。 In this embodiment, if the amplitude of x syn (t) l does not exceed the digital quantization range, that is, the quantization range requirement is met, step S809 may not be performed. It should be noted that, in order to ensure that the output signal meets the quantization range, at least any one of the dynamic range compression in the frequency domain in step S805 and the dynamic range compression in the time domain in step S809 may be performed. Generally speaking, the dynamic range compression in the frequency domain is linear compression, which does not change the timbre of the sound, while the dynamic range compression in the time domain is non-linear compression, which will cause certain timbre distortion.
本实施例中，在得到第l帧重叠的M点对应的压缩处理后的当前帧时域融合音频信号y(n) l，即可作为输出音频信号，录音保存或实时播放。 In this embodiment, the current frame time-domain fusion audio signal y(n) l after the compression processing corresponding to the M point overlapped in the l-th frame is obtained, which can be used as an output audio signal, which can be recorded and saved or played in real time.
本发明实施例提供的音频信号处理方法中的具体原理和实现方式均与上述实施例类似，此处不再赘述。The specific principles and implementation manners in the audio signal processing method provided by the embodiment of the present invention are similar to the foregoing embodiment, and will not be repeated here.
本实施例的音频信号处理方法，通过利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器 的模拟增益各不相同；对所述多路数字音频信号进行频域转换以获取多路频域数据；根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。本实施例的方法可实现有效的提高录音***的动态范围，具有较高的灵敏度，同时能够降低底噪，满足高信噪比的要求，能够同时兼顾小声音的高信噪比录音和大声音的无过载无失真录音，满足了高动态范围的录音需求。The audio signal processing method of this embodiment uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, wherein each of the multiple preprocessing circuits includes an amplifier and an analog audio signal. Digital converter, and the analog gains of the amplifiers of each pre-processing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data One or at least two channels of target frequency domain data determine frequency domain fusion data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal. The method of this embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, and can reduce the noise floor, meet the requirements of high signal-to-noise ratio, and can simultaneously take into account the high signal-to-noise ratio recording of small voices and loud voices The non-overload and distortion-free recording meets the recording needs of high dynamic range.
下面以一个具体实例进行举例说明，在该实例中，预处理电路为两个(即I＝2)，高增益的第1预处理电路比低增益的第2预处理电路的模拟增益大30dB，即G 1＝31.6G 2。测试环境为普通安静房间，音源前3秒为语音信号，后1秒为麦克风附近的敲击声，信号采样率为f s＝48kHz，分析帧长N＝2048，帧移M＝1024。分析窗和合成窗分别皆为正弦窗： Let’s take a specific example to illustrate. In this example, there are two preprocessing circuits (i.e. I=2). The first preprocessing circuit with high gain has an analog gain that is 30dB larger than the second preprocessing circuit with low gain. That is, G 1 =31.6G 2 . The test environment is an ordinary quiet room. The first 3 seconds of the sound source is the voice signal, and the last 1 second is the tapping sound near the microphone. The signal sampling rate is f s =48kHz, the analysis frame length is N = 2048, and the frame shift M = 1024. Both the analysis window and the synthesis window are sine windows:
Figure PCTCN2019083006-appb-000020
Figure PCTCN2019083006-appb-000020
在获取频域融合数据时，选择L 1＝0.3L max，L 2＝0.8L max，L c＝max(|x 1(t) l|)，G x(L 1)＝1，G x(L 2)＝L max/L c。其中，数字量化最大幅值L max＝32768，max(|x 1(t) l|)表示第1预处理电路的输出的第l帧的数字音频信号的最大幅值。由于第l帧重叠累加输出信号x syn(t) l的幅值未超出数字量化范围，故无需在时域的动态范围压缩，即 When acquiring frequency-domain fusion data, choose L 1 =0.3L max , L 2 =0.8L max , L c =max(|x 1 (t) l |), G x (L 1 )=1, G x ( L 2 )=L max /L c . Among them, the maximum amplitude of digital quantization L max =32768, and max(|x 1 (t) l |) represents the maximum amplitude of the digital audio signal of the first frame output by the first preprocessing circuit. Since the amplitude of the l- th frame overlap accumulation output signal x syn (t) l does not exceed the digital quantization range, there is no need for dynamic range compression in the time domain, that is
y(n) l＝x syn(t) l   t＝1,2,…,M y(n) l = x syn (t) l t = 1, 2,...,M
第1预处理电路和第2预处理电路的输出的第l帧的数字音频信号(时域信号)分别如图11a和图11b所示，第1预处理电路和第2预处理电路的输出的第l帧的数字音频信号(时频谱)分别如12a和图12b所示，观察高增益的第1预处理电路的时域信号和时频谱可知，语音段的录音效果可观，但敲击声段时域信号过载，频谱失真严重；观察低增益的第2预处理电路的时域信号和时频谱可知，敲击声的录音效果可观，但语音段声音小且信噪比极低， 也即说明了普通的麦克风录音***(采用单一的预处理电路)无法保证小声音的高信噪比和大声音的不过载两者兼得。The digital audio signal (time domain signal) of the first frame output by the first preprocessing circuit and the second preprocessing circuit is shown in Figure 11a and Figure 11b, respectively. The output of the first preprocessing circuit and the second preprocessing circuit are The digital audio signal (time spectrum) of the first frame is shown in Figure 12a and Figure 12b, respectively. Observing the time domain signal and time spectrum of the first preprocessing circuit with high gain, we can see that the recording effect of the voice segment is considerable, but the percussion segment The time-domain signal is overloaded and the frequency spectrum is severely distorted. Observing the time-domain signal and time-frequency spectrum of the second preprocessing circuit with low gain, it can be seen that the recording effect of the percussion sound is considerable, but the voice segment has a small sound and a very low signal-to-noise ratio, which means that The ordinary microphone recording system (using a single pre-processing circuit) cannot guarantee the high signal-to-noise ratio of small sounds and the overdrive of loud sounds.
而通过本实施例的音频信号处理方法进行处理后，输出音频信号的时域信号和时频谱分别如图13a和13b所示。由图可知，语音段信号与高增益的第1预处理电路的输出的第l帧的数字音频信号基本一致，满足了高信噪比的要求；敲击声段保留了低增益的第2预处理电路的输出的第l帧的数字音频信号的频谱，并动态调整其幅值；且小声音和大声音之间连续平滑过渡。因此，本实施例的方法能同时兼顾小声音的高信噪比录音和大声音的无过载无失真录音，满足了高动态范围的录音需求。After processing by the audio signal processing method of this embodiment, the time-domain signal and time-frequency spectrum of the output audio signal are shown in Figs. 13a and 13b, respectively. It can be seen from the figure that the speech segment signal is basically the same as the digital audio signal of the first frame output by the high gain first preprocessing circuit, which meets the requirements of high signal-to-noise ratio; the percussion segment retains the low gain second pre-processing Processing the frequency spectrum of the digital audio signal of the first frame of the output of the circuit, and dynamically adjusting its amplitude; and a continuous smooth transition between the small sound and the loud sound. Therefore, the method of this embodiment can simultaneously take into account the high signal-to-noise ratio recording of small sounds and the non-overload and distortion-free recording of loud sounds, meeting the recording requirements of high dynamic range.
本发明实施例提供一种音频信号处理设备。图14为本发明实施例提供的音频信号处理设备的结构图，如图14所示，音频信号处理设备90包括存储器92和处理器91。The embodiment of the present invention provides an audio signal processing device. FIG. 14 is a structural diagram of an audio signal processing device provided by an embodiment of the present invention. As shown in FIG. 14, the audio signal processing device 90 includes a memory 92 and a processor 91.
所述存储器92用于存储程序代码；The memory 92 is used to store program codes;
所述处理器91，调用所述程序代码，当程序代码被执行时，用于执行以下操作：The processor 91 calls the program code, and when the program code is executed, is used to perform the following operations:
利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；A plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals. Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
对所述多路数字音频信号进行频域转换以获取多路频域数据；Performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data;
根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；Determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data;
将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。The frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
在上述任一实施例的基础上，所述处理器91被配置为：获取所述待处理的模拟音频信号的能量特征信息；On the basis of any of the foregoing embodiments, the processor 91 is configured to: obtain energy feature information of the analog audio signal to be processed;
所述根据所述多路频域数据中的一路或至少两路频域数据确定频域融合数据，包括：The determining frequency domain fusion data according to one or at least two channels of frequency domain data among the multiple channels of frequency domain data includes:
根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标 频域数据；Determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information;
根据所述一路或至少两路目标频域数据确定频域融合数据。Determine the frequency domain fusion data according to the one or at least two channels of target frequency domain data.
在上述任一实施例的基础上，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数；On the basis of any of the foregoing embodiments, each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter;
在所述处理器91根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据时，所述处理器91被配置为：When the processor 91 determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processor 91 is configured to:
根据所述能量特征信息和多个参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data. The analog gain of the amplifier circuit included in each preprocessing circuit is determined.
在上述任一实施例的基础上，在所述处理器91根据所述能量特征信息从所述多路频域数据中确定第一目标频域数据和第二目标频域数据时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 determines the first target frequency domain data and the second target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processing The device 91 is configured as:
从多个参考能量特征参数中确定与所述能量特征信息相邻的第一参考能量特征参数和第二参考能量特征参数；Determining a first reference energy characteristic parameter and a second reference energy characteristic parameter adjacent to the energy characteristic information from a plurality of reference energy characteristic parameters;
根据第一参考能量特征参数和第二参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，第一目标频域数据和第二目标频域数据分别是对所述多路数字音频信号中的第一数字音频信号和第二数字音频信号进行所述频域转换得到的，所述第一数字音频信号和第二数字音频信号分别是由第一参考能量特征参数和第二参考能量特征参数对应的第一预处理电路和第二预处理电路对所述待处理的模拟音频信号得到的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter, where the first target frequency domain data and the second target The frequency domain data is obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, and the first digital audio signal and the second digital audio signal are respectively The analog audio signal to be processed is obtained by the first preprocessing circuit and the second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter.
在上述任一实施例的基础上，在所述处理器91根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processor 91 is configured for:
当所述所述能量特征信息小于所述多个参考能量特征参数中最小的第三参考能量特征参数时，根据所述第三参考能量特征参数从所述多路频域数据中确定第三目标频域数据，其中，第三目标频域数据是对所述多路数字音频信号中的第三数字音频信号进行所述频域转换得到的，所述第三数字音频信号是由第三参考能量特征参数对应的第三预处理电路对所述待处理的模拟音频信号得到的；When the energy characteristic information is less than the smallest third reference energy characteristic parameter among the plurality of reference energy characteristic parameters, a third target is determined from the multi-channel frequency domain data according to the third reference energy characteristic parameter Frequency domain data, wherein the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is obtained from a third reference energy The third preprocessing circuit corresponding to the characteristic parameter is obtained from the analog audio signal to be processed;
根据第三目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the third target frequency domain data.
在上述任一实施例的基础上，在所述处理器91对所述多路频域数据进行融合以获取频域融合数据时，所述处理器91还被配置为：On the basis of any of the foregoing embodiments, when the processor 91 fuses the multiple channels of frequency domain data to obtain frequency domain fusion data, the processor 91 is further configured to:
当所述所述能量特征信息大于所述多个参考能量特征参数中最大的第四参考能量特征参数时，根据所述第四参考能量特征参数从所述多路频域数据中确定第四目标频域数据，其中，第四目标频域数据是对所述多路数字音频信号中的第四数字音频信号进行所述频域转换得到的，所述第四数字音频信号是由第四参考能量特征参数对应的第四预处理电路对所述待处理的模拟音频信号得到的；When the energy feature information is greater than the largest fourth reference energy feature parameter among the multiple reference energy feature parameters, a fourth target is determined from the multiple channels of frequency domain data according to the fourth reference energy feature parameter Frequency domain data, where the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multiple digital audio signals, and the fourth digital audio signal is obtained by a fourth reference energy Obtained by the fourth preprocessing circuit corresponding to the characteristic parameter from the analog audio signal to be processed;
根据第四目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the fourth target frequency domain data.
在上述任一实施例的基础上，在所述处理器91根据所述一路或至少两路目标频域数据确定频域融合数据时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 determines frequency domain fusion data according to the one or at least two channels of target frequency domain data, the processor 91 is configured to:
对所述至少两路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the at least two channels of frequency domain data to obtain frequency domain fusion data.
在上述任一实施例的基础上，所述一路或者多路频域数据包括第一目标频域数据和第二目标频域数据；On the basis of any of the foregoing embodiments, the one or more channels of frequency domain data include first target frequency domain data and second target frequency domain data;
在所述处理器91根据所述一路或者多路频域数据获取频域融合数据时，所述处理器91被配置为：When the processor 91 obtains frequency domain fusion data according to the one or more channels of frequency domain data, the processor 91 is configured to:
根据第一目标频域数据和第二目标频域数据对应的权重对所述第一路和第二路路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the first channel and the second channel of frequency domain data according to the weights corresponding to the first target frequency domain data and the second target frequency domain data to obtain frequency domain fusion data.
在上述任一实施例的基础上，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的，其中，所述第一目标频域数据和第二目标频域数据对应的权重是根据所述多个预设处理电路中与所述第一目标频域数据对应的第一预处理电路和与所述第二目标频域数据对应的第二预处理电路的参考能量特征参数确定。On the basis of any of the foregoing embodiments, each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter, and the multiple reference energy characteristic parameters are based on the fact that the multiple preprocessing circuits include The analog gain of the amplification circuit is determined, wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the weights corresponding to the first target frequency domain data in the plurality of preset processing circuits The reference energy characteristic parameters of the first preprocessing circuit and the second preprocessing circuit corresponding to the second target frequency domain data are determined.
在上述任一实施例的基础上，在所述处理器91根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 determines frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data, the processor 91 is configured to :
根据所述一路或至少两路目标频域数据中每一路对应的压缩系数对所述一路或至少两路目标频域数据进行压缩处理；Performing compression processing on the one or at least two channels of target frequency domain data according to a compression coefficient corresponding to each of the one or at least two channels of target frequency domain data;
根据所述压缩处理之后的一路或者多路频域数据获取频域融合数据。Acquire frequency domain fusion data according to one or more channels of frequency domain data after the compression processing.
在上述任一实施例的基础上，所述压缩处理为线性压缩处理。On the basis of any of the foregoing embodiments, the compression processing is linear compression processing.
在上述任一实施例的基础上，所述一路或者多路频域数据每一路频域数据对应的压缩系数是根据与所述每一路频域数据对应的预处理电路包括的放大器的模拟增益确定的。On the basis of any of the foregoing embodiments, the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is determined according to the analog gain of the amplifier included in the preprocessing circuit corresponding to each channel of frequency domain data of.
在上述任一实施例的基础上，所述时域音频信号为当前帧时域音频信号，其中，在所述处理器91根据所述时域音频信号获取输出音频信号时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, the time domain audio signal is the current frame time domain audio signal, wherein, when the processor 91 obtains the output audio signal according to the time domain audio signal, the processor 91 Is configured as:
将所述当前帧时域音频信号与当前帧时域音频信号之前获取的历史帧时域音频信号进行叠加处理以获取当前帧时域融合音频信号；Superimposing the current frame time domain audio signal with the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal;
根据所述当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame.
在上述任一实施例的基础上，在所述处理器91根据所述当前帧时域融合音频信号确定所述输出音频信号时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 determines the output audio signal according to the time-domain fusion audio signal of the current frame, the processor 91 is configured to:
根据预设的压缩系数对所述根据所述当前帧时域融合音频信号进行压缩处理；Performing compression processing on the time-domain fusion audio signal according to the current frame according to a preset compression coefficient;
根据所述压缩处理之后的当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame after the compression processing.
在上述任一实施例的基础上，在所述处理器91对所述多路数字音频信号进行频域转换以获取多路频域数据时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 performs frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data, the processor 91 is configured to:
对所述多路数字音频信号进行加窗处理；Performing windowing processing on the multi-channel digital audio signal;
根据加窗后的多路数字音频信号进行频域转换以获取多路频域数据。Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
在上述任一实施例的基础上，在所述处理器91根据所述时域音频信号获取输出音频信号时，所述处理器91被配置为：On the basis of any of the foregoing embodiments, when the processor 91 obtains an output audio signal according to the time domain audio signal, the processor 91 is configured to:
对所述时域音频信号进行加窗处理；Windowing the time domain audio signal;
根据所述加窗处理之后的时域音频信号确定所述输出音频信号。The output audio signal is determined according to the time domain audio signal after the windowing process.
本发明实施例提供的音频信号处理设备的具体原理和实现方式均与上述实施例类似，此处不再赘述。The specific principles and implementation manners of the audio signal processing device provided in the embodiment of the present invention are similar to those in the foregoing embodiment, and will not be repeated here.
本实施例提供的音频信号处理设备，通过利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；对所述多路数字音频信号进行频域转换以获取多 路频域数据；根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。本实施例可实现有效的提高录音***的动态范围，具有较高的灵敏度，同时能够降低底噪，满足高信噪比的要求，能够同时兼顾小声音的高信噪比录音和大声音的无过载无失真录音，满足了高动态范围的录音需求。The audio signal processing device provided in this embodiment uses multiple preprocessing circuits to process analog audio signals to be processed to obtain multiple digital audio signals, where each of the multiple preprocessing circuits includes an amplifier and The analog-to-digital converter, and the analog gains of the amplifiers of each preprocessing circuit are different; frequency domain conversion is performed on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data; according to the multiple channels of frequency domain data Determine the frequency domain fusion data from one or at least two channels of target frequency domain data; convert the frequency domain fusion data into a time domain audio signal, and obtain an output audio signal according to the time domain audio signal. This embodiment can effectively improve the dynamic range of the recording system, has high sensitivity, can reduce the noise floor, meet the requirements of high signal-to-noise ratio, and can simultaneously take into account the high signal-to-noise ratio recording of small sounds and the absence of loud sounds. Overloading and distortion-free recording meets the recording needs of high dynamic range.
本发明实施例提供一种录音***，所述录音***包括：麦克风，用于采集模拟音频信号；以及如上述实施例所述的音频处理设备90。An embodiment of the present invention provides a recording system. The recording system includes: a microphone for collecting analog audio signals; and an audio processing device 90 as described in the foregoing embodiment.
另外，本实施例还提供一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行以实现上述实施例所述的音频处理方法。In addition, this embodiment also provides a computer-readable storage medium on which a computer program is stored, and the computer program is executed by a processor to implement the audio processing method described in the foregoing embodiment.
在本发明所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个***，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。In the several embodiments provided by the present invention, it should be understood that the disclosed device and method may be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other divisions in actual implementation, for example, multiple units or components can be combined or It can be integrated into another system, or some features can be ignored or not implemented. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。In addition, the functional units in the various embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware, or may be implemented in the form of hardware plus software functional units.
上述以软件功能单元的形式实现的集成的单元，可以存储在一个计算机可读取存储介质中。上述软件功能单元存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或处理器(processor)执行本发明各个实施例所述方法的部分步骤。而前述 的存储介质包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。The above-mentioned integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes several instructions to make a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor execute the method described in the various embodiments of the present invention. Part of the steps. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other media that can store program code .
本领域技术人员可以清楚地了解到，为描述的方便和简洁，仅以上述各功能模块的划分进行举例说明，实际应用中，可以根据需要而将上述功能分配由不同的功能模块完成，即将装置的内部结构划分成不同的功能模块，以完成以上描述的全部或者部分功能。上述描述的装置的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。Those skilled in the art can clearly understand that, for the convenience and conciseness of the description, only the division of the above-mentioned functional modules is used as an example. In practical applications, the above-mentioned functions can be allocated by different functional modules as required, namely, the device The internal structure is divided into different functional modules to complete all or part of the functions described above. For the specific working process of the device described above, reference may be made to the corresponding process in the foregoing method embodiment, which is not repeated here.
最后应说明的是：以上各实施例仅用以说明本发明的技术方案，而非对其限制；尽管参照前述各实施例对本发明进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand: It is still possible to modify the technical solutions described in the foregoing embodiments, or equivalently replace some or all of the technical features; these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present invention range.

Claims (34)

  1. 一种音频信号处理方法，其特征在于，包括：An audio signal processing method, characterized by comprising:
    利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；A plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals. Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
    对所述多路数字音频信号进行频域转换以获取多路频域数据；Performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data;
    根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；Determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data;
    将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。The frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  2. 根据权利要求1所述的方法，其特征在于，所述方法还包括：获取所述待处理的模拟音频信号的能量特征信息；The method according to claim 1, wherein the method further comprises: obtaining energy characteristic information of the analog audio signal to be processed;
    所述根据所述多路频域数据中的一路或至少两路频域数据确定频域融合数据，包括：The determining frequency domain fusion data according to one or at least two channels of frequency domain data among the multiple channels of frequency domain data includes:
    根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据；Determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information;
    根据所述一路或至少两路目标频域数据确定频域融合数据。Determine the frequency domain fusion data according to the one or at least two channels of target frequency domain data.
  3. 根据权利要求2所述的方法，其特征在于，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数；The method according to claim 2, wherein each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter;
    所述根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据，包括：The determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information includes:
    根据所述能量特征信息和多个参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data. The analog gain of the amplifier circuit included in each preprocessing circuit is determined.
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述能量特征信息从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，包括：The method according to claim 3, wherein the determining the first target frequency domain data and the second target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information comprises:
    从多个参考能量特征参数中确定与所述能量特征信息相邻的第一参考能量特征参数和第二参考能量特征参数；Determining a first reference energy characteristic parameter and a second reference energy characteristic parameter adjacent to the energy characteristic information from a plurality of reference energy characteristic parameters;
    根据第一参考能量特征参数和第二参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，第一目标频域数据 和第二目标频域数据分别是对所述多路数字音频信号中的第一数字音频信号和第二数字音频信号进行所述频域转换得到的，所述第一数字音频信号和第二数字音频信号分别是由第一参考能量特征参数和第二参考能量特征参数对应的第一预处理电路和第二预处理电路对所述待处理的模拟音频信号得到的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter, where the first target frequency domain data and the second target The frequency domain data is obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, and the first digital audio signal and the second digital audio signal are respectively The analog audio signal to be processed is obtained by the first preprocessing circuit and the second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter.
  5. 根据权利要求3或4所述的方法，其特征在于，所述根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据，包括：The method according to claim 3 or 4, wherein the determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information comprises:
    当所述所述能量特征信息小于所述多个参考能量特征参数中最小的第三参考能量特征参数时，根据所述第三参考能量特征参数从所述多路频域数据中确定第三目标频域数据，其中，第三目标频域数据是对所述多路数字音频信号中的第三数字音频信号进行所述频域转换得到的，所述第三数字音频信号是由第三参考能量特征参数对应的第三预处理电路对所述待处理的模拟音频信号得到的；When the energy characteristic information is less than the smallest third reference energy characteristic parameter among the plurality of reference energy characteristic parameters, a third target is determined from the multi-channel frequency domain data according to the third reference energy characteristic parameter Frequency domain data, wherein the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is obtained from a third reference energy The third preprocessing circuit corresponding to the characteristic parameter is obtained from the analog audio signal to be processed;
    根据第三目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the third target frequency domain data.
  6. 根据权利要求3-5任一项所述的方法，其特征在于，所述对所述多路频域数据进行融合以获取频域融合数据，还包括：The method according to any one of claims 3-5, wherein the fusing the multiple channels of frequency domain data to obtain frequency domain fusion data further comprises:
    当所述所述能量特征信息大于所述多个参考能量特征参数中最大的第四参考能量特征参数时，根据所述第四参考能量特征参数从所述多路频域数据中确定第四目标频域数据，其中，第四目标频域数据是对所述多路数字音频信号中的第四数字音频信号进行所述频域转换得到的，所述第四数字音频信号是由第四参考能量特征参数对应的第四预处理电路对所述待处理的模拟音频信号得到的；When the energy feature information is greater than the largest fourth reference energy feature parameter among the multiple reference energy feature parameters, a fourth target is determined from the multiple channels of frequency domain data according to the fourth reference energy feature parameter Frequency domain data, where the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multiple digital audio signals, and the fourth digital audio signal is obtained by a fourth reference energy Obtained by the fourth preprocessing circuit corresponding to the characteristic parameter from the analog audio signal to be processed;
    根据第四目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the fourth target frequency domain data.
  7. 根据权利要求1-6任一项所述的方法，其特征在于，所述根据所述一路或至少两路目标频域数据确定频域融合数据，包括：The method according to any one of claims 1-6, wherein the determining frequency domain fusion data according to the one or at least two channels of target frequency domain data comprises:
    对所述至少两路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the at least two channels of frequency domain data to obtain frequency domain fusion data.
  8. 根据权利要求7所述的方法，其特征在于，所述一路或者多路频域数据包括第一目标频域数据和第二目标频域数据；The method according to claim 7, wherein the one or more channels of frequency domain data comprise first target frequency domain data and second target frequency domain data;
    所述根据所述一路或者多路频域数据获取频域融合数据，包括：The acquiring frequency domain fusion data according to the one or more channels of frequency domain data includes:
    根据第一目标频域数据和第二目标频域数据对应的权重对所述第一路和第二路路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the first channel and the second channel of frequency domain data according to the weights corresponding to the first target frequency domain data and the second target frequency domain data to obtain frequency domain fusion data.
  9. 根据权利要求8所述的方法，其特征在于，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的，其中，所述第一目标频域数据和第二目标频域数据对应的权重是根据所述多个预设处理电路中与所述第一目标频域数据对应的第一预处理电路和与所述第二目标频域数据对应的第二预处理电路的参考能量特征参数确定。The method according to claim 8, wherein each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter, and the multiple reference energy characteristic parameters are based on the multiple preset processing circuits. The analog gain of the amplifying circuit included in the processing circuit is determined, wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the weights corresponding to the first target frequency in the plurality of preset processing circuits. The reference energy characteristic parameters of the first preprocessing circuit corresponding to the domain data and the second preprocessing circuit corresponding to the second target frequency domain data are determined.
  10. 根据权利要求1-9所述的方法，其特征在于，所述根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据，包括：The method according to claims 1-9, wherein the determining frequency domain fusion data according to one or at least two channels of target frequency domain data in the multiple channels of frequency domain data comprises:
    根据所述一路或至少两路目标频域数据中每一路对应的压缩系数对所述一路或至少两路目标频域数据进行压缩处理；Performing compression processing on the one or at least two channels of target frequency domain data according to a compression coefficient corresponding to each of the one or at least two channels of target frequency domain data;
    根据所述压缩处理之后的一路或者多路频域数据获取频域融合数据。Acquire frequency domain fusion data according to one or more channels of frequency domain data after the compression processing.
  11. 根据权利要求10所述的方法，其特征在于，所述压缩处理为线性压缩处理。The method according to claim 10, wherein the compression process is a linear compression process.
  12. 根据权利要求10或11所述的方法，其特征在于，所述一路或者多路频域数据每一路频域数据对应的压缩系数是根据与所述每一路频域数据对应的预处理电路包括的放大器的模拟增益确定的。The method according to claim 10 or 11, wherein the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is included according to the preprocessing circuit corresponding to each channel of frequency domain data. The analog gain of the amplifier is determined.
  13. 根据权利要求1-12任一项所述的方法，其特征在于，所述时域音频信号为当前帧时域音频信号，其中，所述根据所述时域音频信号获取输出音频信号，包括：The method according to any one of claims 1-12, wherein the time domain audio signal is a current frame time domain audio signal, and wherein the obtaining the output audio signal according to the time domain audio signal comprises:
    将所述当前帧时域音频信号与当前帧时域音频信号之前获取的历史帧时域音频信号进行叠加处理以获取当前帧时域融合音频信号；Superimposing the current frame time domain audio signal with the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal;
    根据所述当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame.
  14. 根据权利要求13所述的方法，其特征在于，所述根据所述当前帧时域融合音频信号确定所述输出音频信号，包括：The method according to claim 13, wherein the determining the output audio signal according to the time-domain fusion audio signal of the current frame comprises:
    根据预设的压缩系数对所述根据所述当前帧时域融合音频信号进行压缩处理；Performing compression processing on the time-domain fusion audio signal according to the current frame according to a preset compression coefficient;
    根据所述压缩处理之后的当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame after the compression processing.
  15. 根据权利要求1-14任一项所述的方法，其特征在于，所述对所述多路数字音频信号进行频域转换以获取多路频域数据，包括：The method according to any one of claims 1-14, wherein the performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data comprises:
    对所述多路数字音频信号进行加窗处理；Performing windowing processing on the multi-channel digital audio signal;
    根据加窗后的多路数字音频信号进行频域转换以获取多路频域数据。Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
  16. 根据权利要求1-15任一项所述的方法，其特征在于，所述根据所述时域音频信号获取输出音频信号，包括：The method according to any one of claims 1-15, wherein the obtaining an output audio signal according to the time domain audio signal comprises:
    对所述时域音频信号进行加窗处理；Windowing the time domain audio signal;
    根据所述加窗处理之后的时域音频信号确定所述输出音频信号。The output audio signal is determined according to the time domain audio signal after the windowing process.
  17. 一种音频信号处理设备，其特征在于，包括：存储器和处理器；An audio signal processing device, which is characterized by comprising: a memory and a processor;
    所述存储器用于存储程序代码；The memory is used to store program codes;
    所述处理器，调用所述程序代码，当程序代码被执行时，用于执行以下操作：The processor calls the program code, and when the program code is executed, is used to perform the following operations:
    利用多个预处理电路对待处理的模拟音频信号进行处理以获取多路数字音频信号，其中，多个预处理电路中的每一个预处理电路包括放大器和模数转换器，且各预处理电路的所述放大器的模拟增益各不相同；A plurality of preprocessing circuits are used to process the analog audio signal to be processed to obtain multiple digital audio signals. Each of the plurality of preprocessing circuits includes an amplifier and an analog-to-digital converter. The analog gains of the amplifiers are different;
    对所述多路数字音频信号进行频域转换以获取多路频域数据；Performing frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data;
    根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据；Determining frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data;
    将所述频域融合数据转换为时域音频信号，并根据所述时域音频信号获取输出音频信号。The frequency domain fusion data is converted into a time domain audio signal, and an output audio signal is obtained according to the time domain audio signal.
  18. 根据权利要求17所述的设备，其特征在于，所述处理器还被配置为：获取所述待处理的模拟音频信号的能量特征信息；The device according to claim 17, wherein the processor is further configured to: obtain energy characteristic information of the analog audio signal to be processed;
    所述根据所述多路频域数据中的一路或至少两路频域数据确定频域融合数据，包括：The determining frequency domain fusion data according to one or at least two channels of frequency domain data among the multiple channels of frequency domain data includes:
    根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据；Determining one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information;
    根据所述一路或至少两路目标频域数据确定频域融合数据。Determine the frequency domain fusion data according to the one or at least two channels of target frequency domain data.
  19. 根据权利要求18所述的设备，其特征在于，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数；The device according to claim 18, wherein each of the plurality of preset processing circuits corresponds to a different reference energy characteristic parameter;
    在所述处理器根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据时，所述处理器被配置为：When the processor determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processor is configured to:
    根据所述能量特征信息和多个参考能量特征参数从所述多路频域数据中 确定第一目标频域数据和第二目标频域数据，其中，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the energy feature information and multiple reference energy feature parameters, wherein the multiple reference energy feature parameters are based on the multiple frequency domain data. The analog gain of the amplifier circuit included in each preprocessing circuit is determined.
  20. 根据权利要求19所述的设备，其特征在于，在所述处理器根据所述能量特征信息从所述多路频域数据中确定第一目标频域数据和第二目标频域数据时，所述处理器被配置为：The device according to claim 19, wherein when the processor determines the first target frequency domain data and the second target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, The processor is configured as:
    从多个参考能量特征参数中确定与所述能量特征信息相邻的第一参考能量特征参数和第二参考能量特征参数；Determining a first reference energy characteristic parameter and a second reference energy characteristic parameter adjacent to the energy characteristic information from a plurality of reference energy characteristic parameters;
    根据第一参考能量特征参数和第二参考能量特征参数从所述多路频域数据中确定第一目标频域数据和第二目标频域数据，其中，第一目标频域数据和第二目标频域数据分别是对所述多路数字音频信号中的第一数字音频信号和第二数字音频信号进行所述频域转换得到的，所述第一数字音频信号和第二数字音频信号分别是由第一参考能量特征参数和第二参考能量特征参数对应的第一预处理电路和第二预处理电路对所述待处理的模拟音频信号得到的。The first target frequency domain data and the second target frequency domain data are determined from the multiple channels of frequency domain data according to the first reference energy characteristic parameter and the second reference energy characteristic parameter, where the first target frequency domain data and the second target The frequency domain data is obtained by performing the frequency domain conversion on the first digital audio signal and the second digital audio signal in the multi-channel digital audio signal, and the first digital audio signal and the second digital audio signal are respectively The analog audio signal to be processed is obtained by the first preprocessing circuit and the second preprocessing circuit corresponding to the first reference energy characteristic parameter and the second reference energy characteristic parameter.
  21. 根据权利要求19或20所述的设备，其特征在于，在所述处理器根据所述能量特征信息从所述多路频域数据中确定一路或至少两路目标频域数据时，所述处理器被配置为：The device according to claim 19 or 20, wherein when the processor determines one or at least two channels of target frequency domain data from the multiple channels of frequency domain data according to the energy characteristic information, the processing The device is configured as:
    当所述所述能量特征信息小于所述多个参考能量特征参数中最小的第三参考能量特征参数时，根据所述第三参考能量特征参数从所述多路频域数据中确定第三目标频域数据，其中，第三目标频域数据是对所述多路数字音频信号中的第三数字音频信号进行所述频域转换得到的，所述第三数字音频信号是由第三参考能量特征参数对应的第三预处理电路对所述待处理的模拟音频信号得到的；When the energy characteristic information is less than the smallest third reference energy characteristic parameter among the plurality of reference energy characteristic parameters, a third target is determined from the multi-channel frequency domain data according to the third reference energy characteristic parameter Frequency domain data, wherein the third target frequency domain data is obtained by performing the frequency domain conversion on a third digital audio signal in the multi-channel digital audio signal, and the third digital audio signal is obtained from a third reference energy The third preprocessing circuit corresponding to the characteristic parameter is obtained from the analog audio signal to be processed;
    根据第三目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the third target frequency domain data.
  22. 根据权利要求19-21任一项所述的设备，其特征在于，在所述处理器对所述多路频域数据进行融合以获取频域融合数据时，所述处理器还被配置为：The device according to any one of claims 19-21, wherein when the processor fusions the multiple channels of frequency domain data to obtain frequency domain fusion data, the processor is further configured to:
    当所述所述能量特征信息大于所述多个参考能量特征参数中最大的第四参考能量特征参数时，根据所述第四参考能量特征参数从所述多路频域数据中确定第四目标频域数据，其中，第四目标频域数据是对所述多路数字音频信号中的第四数字音频信号进行所述频域转换得到的，所述第四数字音频信 号是由第四参考能量特征参数对应的第四预处理电路对所述待处理的模拟音频信号得到的；When the energy feature information is greater than the largest fourth reference energy feature parameter among the multiple reference energy feature parameters, a fourth target is determined from the multiple channels of frequency domain data according to the fourth reference energy feature parameter Frequency domain data, where the fourth target frequency domain data is obtained by performing the frequency domain conversion on a fourth digital audio signal in the multiple digital audio signals, and the fourth digital audio signal is obtained by a fourth reference energy Obtained by the fourth preprocessing circuit corresponding to the characteristic parameter from the analog audio signal to be processed;
    根据第四目标频域数据获取所述频域融合数据。Acquire the frequency domain fusion data according to the fourth target frequency domain data.
  23. 根据权利要求17-22任一项所述的设备，其特征在于，在所述处理器根据所述一路或至少两路目标频域数据确定频域融合数据时，所述处理器被配置为：The device according to any one of claims 17-22, wherein when the processor determines frequency domain fusion data according to the one or at least two channels of target frequency domain data, the processor is configured to:
    对所述至少两路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the at least two channels of frequency domain data to obtain frequency domain fusion data.
  24. 根据权利要求23所述的设备，其特征在于，所述一路或者多路频域数据包括第一目标频域数据和第二目标频域数据；The device according to claim 23, wherein the one or more channels of frequency domain data comprise first target frequency domain data and second target frequency domain data;
    在所述处理器根据所述一路或者多路频域数据获取频域融合数据时，所述处理器被配置为：When the processor acquires frequency domain fusion data according to the one or more channels of frequency domain data, the processor is configured to:
    根据第一目标频域数据和第二目标频域数据对应的权重对所述第一路和第二路路频域数据进行叠加运算以获取频域融合数据。Perform a superposition operation on the first channel and the second channel of frequency domain data according to the weights corresponding to the first target frequency domain data and the second target frequency domain data to obtain frequency domain fusion data.
  25. 根据权利要求24所述的设备，其特征在于，所述多个预设处理电路中的每一个都对应一个各不相同的参考能量特征参数，多个参考能量特征参数是根据所述多个预处理电路包括的放大电路的模拟增益确定的，其中，所述第一目标频域数据和第二目标频域数据对应的权重是根据所述多个预设处理电路中与所述第一目标频域数据对应的第一预处理电路和与所述第二目标频域数据对应的第二预处理电路的参考能量特征参数确定。The device according to claim 24, wherein each of the multiple preset processing circuits corresponds to a different reference energy characteristic parameter, and the multiple reference energy characteristic parameters are based on the multiple preset processing circuits. The analog gain of the amplifying circuit included in the processing circuit is determined, wherein the weights corresponding to the first target frequency domain data and the second target frequency domain data are based on the weights corresponding to the first target frequency in the plurality of preset processing circuits. The reference energy characteristic parameters of the first preprocessing circuit corresponding to the domain data and the second preprocessing circuit corresponding to the second target frequency domain data are determined.
  26. 根据权利要求17-25所述的设备，其特征在于，在所述处理器根据所述多路频域数据中的一路或至少两路目标频域数据确定频域融合数据时，所述处理器被配置为：The device according to claims 17-25, wherein when the processor determines frequency domain fusion data according to one or at least two channels of target frequency domain data among the multiple channels of frequency domain data, the processor Is configured as:
    根据所述一路或至少两路目标频域数据中每一路对应的压缩系数对所述一路或至少两路目标频域数据进行压缩处理；Performing compression processing on the one or at least two channels of target frequency domain data according to a compression coefficient corresponding to each of the one or at least two channels of target frequency domain data;
    根据所述压缩处理之后的一路或者多路频域数据获取频域融合数据。Acquire frequency domain fusion data according to one or more channels of frequency domain data after the compression processing.
  27. 根据权利要求26所述的设备，其特征在于，所述压缩处理为线性压缩处理。The device according to claim 26, wherein the compression process is a linear compression process.
  28. 根据权利要求26或27所述的设备，其特征在于，所述一路或者多路频域数据每一路频域数据对应的压缩系数是根据与所述每一路频域数据对应的预处理电路包括的放大器的模拟增益确定的。The device according to claim 26 or 27, wherein the compression coefficient corresponding to each channel of frequency domain data of the one or more channels of frequency domain data is included according to the preprocessing circuit corresponding to each channel of frequency domain data. The analog gain of the amplifier is determined.
  29. 根据权利要求17-28任一项所述的设备，其特征在于，所述时域音频信号为当前帧时域音频信号，其中，在所述处理器根据所述时域音频信号获取输出音频信号时，所述处理器被配置为：The device according to any one of claims 17-28, wherein the time domain audio signal is a current frame time domain audio signal, wherein the processor obtains the output audio signal according to the time domain audio signal When the processor is configured to:
    将所述当前帧时域音频信号与当前帧时域音频信号之前获取的历史帧时域音频信号进行叠加处理以获取当前帧时域融合音频信号；Superimposing the current frame time domain audio signal with the historical frame time domain audio signal obtained before the current frame time domain audio signal to obtain the current frame time domain fusion audio signal;
    根据所述当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame.
  30. 根据权利要求29所述的设备，其特征在于，在所述处理器根据所述当前帧时域融合音频信号确定所述输出音频信号时，所述处理器被配置为：The device according to claim 29, wherein when the processor determines the output audio signal according to the time domain fusion audio signal of the current frame, the processor is configured to:
    根据预设的压缩系数对所述根据所述当前帧时域融合音频信号进行压缩处理；Performing compression processing on the time-domain fusion audio signal according to the current frame according to a preset compression coefficient;
    根据所述压缩处理之后的当前帧时域融合音频信号确定所述输出音频信号。The output audio signal is determined according to the time-domain fusion audio signal of the current frame after the compression processing.
  31. 根据权利要求17-30任一项所述的设备，其特征在于，在所述处理器对所述多路数字音频信号进行频域转换以获取多路频域数据时，所述处理器被配置为：The device according to any one of claims 17-30, wherein when the processor performs frequency domain conversion on the multiple channels of digital audio signals to obtain multiple channels of frequency domain data, the processor is configured for:
    对所述多路数字音频信号进行加窗处理；Performing windowing processing on the multi-channel digital audio signal;
    根据加窗后的多路数字音频信号进行频域转换以获取多路频域数据。Perform frequency domain conversion according to the windowed multiple channels of digital audio signals to obtain multiple channels of frequency domain data.
  32. 根据权利要求17-31任一项所述的设备，其特征在于，在所述处理器根据所述时域音频信号获取输出音频信号时，所述处理器被配置为：The device according to any one of claims 17-31, wherein when the processor obtains the output audio signal according to the time domain audio signal, the processor is configured to:
    对所述时域音频信号进行加窗处理；Windowing the time domain audio signal;
    根据所述加窗处理之后的时域音频信号确定所述输出音频信号。The output audio signal is determined according to the time domain audio signal after the windowing process.
  33. 一种录音***，其特征在于，包括：A recording system, characterized in that it comprises:
    麦克风，用于采集模拟音频信号；以及Microphone, used to collect analog audio signals; and
    如权利要求17-32任一项所述的音频处理设备。The audio processing device according to any one of claims 17-32.
  34. 一种计算机可读存储介质，其特征在于，其上存储有计算机程序，所述计算机程序被处理器执行以实现如权利要求1-16任一项所述的方法。A computer-readable storage medium, characterized in that a computer program is stored thereon, and the computer program is executed by a processor to implement the method according to any one of claims 1-16.
PCT/CN2019/083006 2019-04-17 2019-04-17 Audio signal processing method and device, and storage medium WO2020211004A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201980005583.8A CN111345047A (en) 2019-04-17 2019-04-17 Audio signal processing method, apparatus and storage medium
PCT/CN2019/083006 WO2020211004A1 (en) 2019-04-17 2019-04-17 Audio signal processing method and device, and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/083006 WO2020211004A1 (en) 2019-04-17 2019-04-17 Audio signal processing method and device, and storage medium

Publications (1)

Publication Number Publication Date
WO2020211004A1 true WO2020211004A1 (en) 2020-10-22

Family

ID=71186612

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/083006 WO2020211004A1 (en) 2019-04-17 2019-04-17 Audio signal processing method and device, and storage medium

Country Status (2)

Country Link
CN (1) CN111345047A (en)
WO (1) WO2020211004A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114812789B (en) * 2021-01-28 2024-03-08 漳州立达信光电子科技有限公司 Sound detection method, device and equipment
CN113079440B (en) * 2021-03-22 2022-12-06 Oppo广东移动通信有限公司 Audio signal processing method and device, terminal and storage medium
CN113823314B (en) * 2021-08-12 2022-10-28 北京荣耀终端有限公司 Voice processing method and electronic equipment
CN113643719A (en) * 2021-08-26 2021-11-12 Oppo广东移动通信有限公司 Audio signal processing method and device, storage medium and terminal equipment
CN117761393B (en) * 2024-02-22 2024-05-07 南京派格测控科技有限公司 Time domain signal acquisition method and device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104066036A (en) * 2014-06-19 2014-09-24 华为技术有限公司 Pick-up device and method
US20170033754A1 (en) * 2015-07-29 2017-02-02 Invensense, Inc. Multipath digital microphones
CN108683978A (en) * 2018-05-30 2018-10-19 中南大学 A kind of audio-frequency processing method based on frequency domain adaptive equilibrium

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
FR2566658A1 (en) * 1984-06-28 1986-01-03 Inst Nat Sante Rech Med Multichannel auditory prosthesis.
US8085954B2 (en) * 2005-07-21 2011-12-27 Freescale Semiconductor, Inc. Microphone amplification arrangement and integrated circuit therefor
WO2009039897A1 (en) * 2007-09-26 2009-04-02 Fraunhofer - Gesellschaft Zur Förderung Der Angewandten Forschung E.V. Apparatus and method for extracting an ambient signal in an apparatus and method for obtaining weighting coefficients for extracting an ambient signal and computer program
US9380381B2 (en) * 2014-03-18 2016-06-28 Infineon Technologies Ag Microphone package and method for providing a microphone package
CN104954020B (en) * 2014-03-28 2018-07-24 意法半导体股份有限公司 Multichannel transducer device and its operating method
US9609451B2 (en) * 2015-02-12 2017-03-28 Dts, Inc. Multi-rate system for audio processing
CN106162432A (en) * 2015-04-03 2016-11-23 吴法功 A kind of audio process device and sound thereof compensate framework and process implementation method
CN107895580B (en) * 2016-09-30 2021-06-01 华为技术有限公司 Audio signal reconstruction method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104066036A (en) * 2014-06-19 2014-09-24 华为技术有限公司 Pick-up device and method
US20170033754A1 (en) * 2015-07-29 2017-02-02 Invensense, Inc. Multipath digital microphones
CN108683978A (en) * 2018-05-30 2018-10-19 中南大学 A kind of audio-frequency processing method based on frequency domain adaptive equilibrium

Also Published As

Publication number Publication date
CN111345047A (en) 2020-06-26

Similar Documents

Publication Publication Date Title
WO2020211004A1 (en) Audio signal processing method and device, and storage medium
US9210506B1 (en) FFT bin based signal limiting
EP2695394B1 (en) Integrated psychoacoustic bass enhancement (pbe) for improved audio
US10109288B2 (en) Dynamic range and peak control in audio using nonlinear filters
JP5507596B2 (en) Speech enhancement
JP5984943B2 (en) Improving stability and ease of listening to sound in hearing devices
US9311933B2 (en) Method of processing a voice segment and hearing aid
US20070104335A1 (en) Acoustic feedback suppression for audio amplification systems
WO2010089976A1 (en) Hearing aid
JP4249729B2 (en) Automatic gain control method, automatic gain control device, automatic gain control program, and recording medium recording the same
JP6698125B2 (en) Audio processing apparatus and method
WO2020211017A1 (en) Audio signal processing method and device, and storage medium
EP3603106B1 (en) Dynamically extending loudspeaker capabilities
US11863946B2 (en) Method, apparatus and computer program for processing audio signals
JPH09311696A (en) Automatic gain control device
WO2018027583A1 (en) Microphone and method for processing audio signals
CN104793525A (en) Amplifier arrangement with limiting module
US20210329387A1 (en) Systems and methods for a hearing assistive device
Girisha et al. Implementation of novel algorithm for auditory compensation in hearing aids using STFT algorithm
CN117912485A (en) Speech band extension method, noise reduction audio device, and storage medium
Girisha et al. IMPLEMENTATION OF NOVEL ALGORITHM FOR AUDITORY COMPENSATION IN HEARING AIDS USING STFT
WO2023249957A1 (en) Speech enhancement and interference suppression
CN114094964A (en) Dynamic range control method and device and electronic equipment
CN115623385A (en) Audio dynamic range control circuit and audio equipment
Chen et al. Research and Optimization of Multichannel Wide Dynamic Compression Algorithm in Hearing Aids

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19925363

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19925363

Country of ref document: EP

Kind code of ref document: A1