A sound quality testing method and system for sound equipment

Abstract

The present invention relates to the field of audio detection technology, and specifically discloses a sound quality testing method and system for audio. The output audio feature information of the audio is obtained; the output audio feature information is segmented; the frequency features at each preset interception time point are obtained based on multiple segmented audio information, and the audio energy value at each preset interception time point is obtained based on each frequency feature, and the hearing evaluation value is calculated based on each audio energy value. The present application effectively prevents the loss of audio feature information segments by segmenting and supplementing the output audio feature information when obtaining the preset interception time point, thereby ensuring the integrity of the evaluation, and at the same time can reduce the data duplication for frequency domain analysis, thereby improving the efficiency of the test.

Description

Technical Field

[0001] The present invention relates to the technical field of audio detection, and in particular to a method and system for testing the sound quality of an audio system. Background Art

[0002] Speakers are generally composed of speaker units and cabinets. The speaker unit is the sound-generating component, and the cabinet complements the speaker unit and plays a role in correcting the sound. The sound-generating principle of the speaker unit is a conversion of electrical energy into sound. When different electron energies are transmitted to the coil, the coil generates a kind of energy that interacts with the magnetic field of the magnet. This interaction causes the paper plate to vibrate. Because the electron energy changes at any time, the speaker coil will move forward or backward, so the speaker paper plate will move with it. These actions change the density of the air and produce sound.

[0003] The existing audio quality testing methods mainly include frequency response test, distortion test and signal-to-noise ratio test. Before testing using the three test methods, a target audio segment needs to be segmented into multiple small test audio segments. During the target audio segmentation process, the segmentation is generally performed according to approximate positions, such as in the middle of a word or note, rather than in the silence between two words or notes. This can easily lead to inaccurate segmentation, which can have an adverse effect on the entire testing process, such as inaccurate sound quality test results. At the same time, the entire target audio segment needs to be re-segmented, which increases the test time and affects the efficiency of the test. In addition, since the results of the above three test methods cannot fully display the effects of hearing, loudness, etc., it is impossible to judge the quality of the sound more accurately, thereby reducing the accuracy of audio detection. Summary of the invention

[0004] The purpose of the present invention is to provide a sound quality testing method for acoustics to solve the technical problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for testing the sound quality of a sound system, comprising:

[0007] Get the output audio feature information of the speaker;

[0008] Segmenting the output audio feature information to obtain multiple segmented audio information;

[0009] Obtain frequency features at each preset interception time point according to the plurality of segmented audio information, obtain an audio energy value at each preset interception time point according to each frequency feature, and calculate a hearing evaluation value according to each audio energy value;

[0010] Obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value;

[0011] Perform frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculate harmonic distortion evaluation values ​​based on the multiple energy main frequency information, and calculate sound quality evaluation values ​​based on the multiple harmonic distortion evaluation values;

[0012] Calculate the total evaluation value according to the hearing evaluation value, loudness change value and sound quality evaluation value;

[0013] Determining whether the total evaluation value is within a preset range;

[0014] If it is, the sound quality test result of the audio system is determined to be qualified;

[0015] If not, the sound quality test result of the audio system is judged to be unqualified.

[0016] Preferably, the step of segmenting the output audio feature information to obtain a plurality of segmented audio information comprises:

[0017] Get output audio feature information;

[0018] Convert the output audio feature information to obtain mono feature information;

[0019] Obtaining multiple short time frames according to the single channel feature information;

[0020] Acquire multiple short-time sound energy values ​​according to multiple short-time frames;

[0021] Determine whether each short-time sound energy value is within a preset energy range. If the short-time sound energy value is not within the preset energy range and exceeds the upper limit of the preset energy range, use the corresponding short-time sound energy value as the starting segmentation point;

[0022] If the short-time sound energy value is not within the preset energy range, and is within the lower limit of the preset energy range, the corresponding short-time sound energy value is used as the end segmentation point;

[0023] The output audio feature information is segmented based on the start segmentation point and the end segmentation point to obtain a plurality of segmented audio information.

[0024] Preferably, the step of obtaining a frequency feature at each preset interception time point according to a plurality of segmented audio information, obtaining an audio energy value at each preset interception time point according to each frequency feature, and calculating a hearing evaluation value according to each audio energy value comprises:

[0025] Obtain the segmented audio information corresponding to each preset interception time point;

[0026] Obtain multiple frequency features by short-time Fourier transforming multiple segmented audio information;

[0027] Performing spectrum analysis on multiple frequency features to obtain the frequency peak value corresponding to each preset interception time point;

[0028] The audio energy value corresponding to the preset interception time point is calculated according to the frequency peak value. The energy value calculation formula is:

[0029] A=|(logX)| 2 ;

[0030] Where A is the audio energy value, X is the frequency peak;

[0031] Get the audio energy value at each preset interception time point;

[0032] The hearing evaluation value is calculated according to each audio energy value. The hearing evaluation value calculation formula is:

[0033]

[0034] Among them, C is the hearing evaluation value, A1 is the audio energy value at the first preset interception time point, A2 is the audio energy value at the second preset interception time point, and A i is the audio energy value at the i-th preset interception time point.

[0035] Preferably, the step of obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value comprises:

[0036] Get the audio energy value at each preset interception time point;

[0037] The volume value is calculated according to the audio energy value. The volume value calculation formula is:

[0038]

[0039] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0040] The loudness change value is calculated based on the volume value at each preset interception time point. The loudness change value calculation formula is:

[0041] B i =D i -D i-1 ;

[0042] Among them, Bi is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point;

[0043] The loudness evaluation value is calculated according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0044]

[0045] Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

[0046] Preferably, each audio energy value is subjected to frequency domain analysis to obtain multiple energy main frequency information, and a harmonic distortion evaluation value is calculated according to the multiple energy main frequency information, and the step of calculating the sound quality evaluation value according to the multiple harmonic distortion evaluation values ​​includes:

[0047] Get multiple audio energy values;

[0048] Perform frequency domain analysis on multiple audio energy values ​​to obtain multiple energy main frequency information;

[0049] By analyzing the spectrum of each energy main frequency information, multiple amplitude values ​​are obtained;

[0050] Get the frequency value corresponding to each amplitude value;

[0051] The harmonic distortion value is calculated according to multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is:

[0052]

[0053] Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value;

[0054] Get multiple harmonic distortion values;

[0055] And normalize each harmonic distortion value to obtain the distortion value;

[0056] Get the weight of each distortion value;

[0057] The sound quality evaluation value is calculated for multiple distortion values ​​and weights, and the calculation formula is:

[0058]

[0059] Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and f j is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

[0060] Preferably, the step of calculating the total evaluation value from the hearing evaluation value, the loudness change value and the sound quality evaluation value comprises:

[0061] Acquiring a hearing evaluation value, and normalizing the hearing evaluation value to obtain a normalized hearing value;

[0062] Acquire a loudness change value, and perform normalization processing on the loudness change value to obtain a normalized loudness value;

[0063] Acquiring a sound quality evaluation value, and normalizing the sound quality evaluation value to obtain a normalized sound quality value;

[0064] Obtain weights of normalized hearing perception value, normalized loudness value, and normalized sound quality value;

[0065] The normalized hearing value, normalized loudness value and normalized sound quality value are used to calculate the total evaluation value, where the total evaluation calculation formula is:

[0066]

[0067] Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

[0068] The present application also provides a sound quality testing system for a speaker, comprising:

[0069] An acquisition module, used to obtain output audio feature information of the speaker;

[0070] A segmentation module, used to obtain a plurality of preset interception time points, and obtain the time intervals between the plurality of preset interception time points, and segment the output audio feature information according to the time intervals to obtain a plurality of segmented audio information;

[0071] A first calculation module is used to obtain a frequency feature at each preset interception time point according to the plurality of segmented audio information, obtain an audio energy value at each preset interception time point according to each frequency feature, and calculate a hearing evaluation value according to each audio energy value;

[0072] A second calculation module is used to obtain a volume value corresponding to a preset interception time point according to the audio energy value, obtain a loudness change value at each preset interception time point according to each volume value, and calculate a loudness evaluation value according to each loudness change value;

[0073] A third calculation module is used to perform frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculate harmonic distortion evaluation values ​​according to the multiple energy main frequency information, and calculate sound quality evaluation values ​​according to the multiple harmonic distortion evaluation values;

[0074] A fourth calculation module, used to calculate a total evaluation value according to the hearing evaluation value, the loudness change value and the sound quality evaluation value;

[0075] A judgment module, used to judge whether the total evaluation value is within a preset range;

[0076] If it is, the sound quality test result of the audio system is determined to be qualified;

[0077] If not, the sound quality test result of the audio system is judged to be unqualified.

[0078] Preferably, the second calculation module includes:

[0079] An acquisition unit, used to acquire the audio energy value at each preset interception time point;

[0080] The first calculation unit is used to calculate the volume value according to the audio energy value. The volume value calculation formula is:

[0081]

[0082] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0083] The second acquisition unit is used to calculate the loudness change value according to the volume value at each preset interception time point, and the loudness change value calculation formula is:

[0084] B i =D i -D i-1 ;

[0085] Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point;

[0086] The third acquisition unit is used to calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0087]

[0088] Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

[0089] Preferably, the second calculation module includes:

[0090] A first acquisition unit, used to acquire the audio energy value at each preset interception time point;

[0091] The first calculation unit is used to calculate the volume value according to the audio energy value. The volume value calculation formula is:

[0092]

[0093] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0094] The second calculation unit is used to calculate the loudness change value according to the volume value at each preset interception time point, and the loudness change value calculation formula is:

[0095] B i =D i -D i-1 ;

[0096] Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point;

[0097] The third calculation unit is used to calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0098]

[0099] Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

[0100] Preferably, the third calculation module includes:

[0101] A second acquisition unit, used to acquire multiple audio energy values;

[0102] A first analysis unit, configured to perform frequency domain analysis on a plurality of audio energy values ​​to obtain a plurality of energy main frequency information;

[0103] A second analysis unit is used to obtain multiple amplitude values ​​by analyzing the spectrum of each energy main frequency information;

[0104] A third acquisition unit is used to acquire the frequency value corresponding to each amplitude value;

[0105] The fourth calculation unit is used to calculate the harmonic distortion value according to the multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is:

[0106]

[0107] Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value;

[0108] A fourth acquisition unit, used to acquire a plurality of harmonic distortion values;

[0109] A third analysis unit is used to normalize each harmonic distortion value to obtain a distortion value;

[0110] A fifth acquisition unit, used to acquire a weight of each distortion value;

[0111] The fifth calculation unit is used to calculate the sound quality evaluation value for the multiple distortion values ​​and weights, and the calculation formula is:

[0112]

[0113] Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and f j is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

[0114] Preferably, the fourth calculation module includes:

[0115] A sixth acquisition unit, used to acquire a hearing evaluation value, and normalize the hearing evaluation value to obtain a normalized hearing value;

[0116] a seventh acquisition unit, configured to acquire a loudness change value, and perform normalization processing on the loudness change value to obtain a normalized loudness value;

[0117] an eighth acquisition unit, configured to acquire a sound quality evaluation value, and perform normalization processing on the sound quality evaluation value to obtain a normalized sound quality value;

[0118] A ninth acquisition unit, used to acquire weights of the normalized hearing perception value, the normalized loudness value, and the normalized sound quality value;

[0119] The sixth calculation unit is used to calculate the total evaluation value of the normalized hearing value, the normalized loudness value and the normalized sound quality value, wherein the total evaluation calculation formula is:

[0120]

[0121] Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

[0122] The beneficial effects of the present application are as follows: the present invention can accurately identify and divide key sections in audio, segment the audio according to the key sections, and by segmenting long audio into multiple short time frames for analysis, the processing speed and efficiency can be improved, and at the same time, the frequency domain analysis can be reduced for repeated data analysis, thereby improving the efficiency of the test, and the audio energy value is used to calculate the hearing evaluation value, thereby providing a quantitative method to describe the perceptual characteristics of the audio, including volume, audio intensity, audio quality, etc., thereby increasing the accuracy of the evaluation, and by calculating the loudness change value, the loudness change of the audio output can be understood, and the dynamic loudness performance of the audio when playing different audio contents can be tested, thereby improving the accuracy of the test;

[0123] By analyzing the harmonic distortion and calculating the sound quality evaluation value, it provides a quantitative measure of audio quality, making the sound quality test more accurate.

[0124] Finally, by comprehensively considering the hearing evaluation value, loudness change value and sound quality evaluation value, a total evaluation value is obtained, thereby judging the performance of the audio played by the speaker more comprehensively and accurately. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] Figure 1 This is a schematic diagram of a method flow of an embodiment of the present application;

[0126] Figure 2 A schematic diagram of the system structure of an embodiment of the present application.

[0127] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0128] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0129] like Figure 1 and Figure 2 As shown, the present application provides a sound quality testing method for a speaker, comprising:

[0130] S1. Obtaining output audio feature information of the speaker;

[0131] S2. Segment the output audio feature information to obtain a plurality of segmented audio information;

[0132] S3, obtaining a frequency feature at each preset interception time point according to the plurality of segmented audio information, obtaining an audio energy value at each preset interception time point according to each frequency feature, and calculating a hearing evaluation value according to each audio energy value;

[0133] S4, obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value;

[0134] S5, performing frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, and calculating harmonic distortion evaluation values ​​according to the multiple energy main frequency information, and calculating sound quality evaluation values ​​according to the multiple harmonic distortion evaluation values;

[0135] S6. Calculate a total evaluation value based on the hearing evaluation value, the loudness change value, and the sound quality evaluation value;

[0136] S7, determining whether the total evaluation value is within a preset range;

[0137] If it is, the sound quality test result of the audio system is determined to be qualified;

[0138] If not, the sound quality test result of the audio system is judged to be unqualified.

[0139] As described in the above steps S1-S7, the present application receives a sinusoidal wave signal played by a speaker through an audio receiving device, obtains an audio signal after filtering the sinusoidal wave signal, and extracts the audio signal to obtain a piece of output audio feature information, wherein the output audio feature information includes frequency domain feature information, time domain feature information and total harmonic feature information, then obtains a plurality of preset interception time points, and obtains the time interval between the plurality of preset interception time points, divides the output audio feature information according to the time interval, and can supplement the segments of the output audio feature information, and finally outputs the supplemented segments of the output audio feature information to obtain a plurality of segmented audio information, thereby preventing the segmented output audio feature information from being lost. The information fragments are missing, so as to ensure that there is no information loss in the obtained multiple segmented audio information; then, the frequency characteristics at each preset interception time point are obtained according to the multiple segmented audio information, and the audio energy value at each preset interception time point is obtained according to each frequency characteristic, and the audio energy value is obtained by calculating the frequency characteristics at each preset interception time point, and the audio energy value represents the intensity of the audio, and the audio energy value reflects the intensity of the audio signal at each preset interception time point, and then the hearing evaluation value is calculated according to each audio energy value. The hearing evaluation value calculated by the energy value of each audio can provide a quantitative way to describe the perceptual characteristics of the audio, where the perceptual characteristics include volume, audio intensity, audio quality, etc. Analyze various perceptual features of the audio to increase the accuracy of detection; then obtain the volume value corresponding to the preset interception time point according to the audio energy value, the volume value includes the measurement of audio intensity and amplitude, indicating the audio intensity at the preset interception time point, and obtain the loudness change value of each preset interception time point according to each volume value, by calculating the loudness change value, it is possible to understand the loudness change of the audio output, and calculate the loudness evaluation value according to each loudness change value, through the loudness evaluation value, it is possible to understand the change of the intensity of the audio output at the preset interception time point, if the loudness evaluation value changes frequently in a short period of time and the amplitude is large, this indicates that the audio can respond quickly to the preset sine wave signal, and can be used to identify the dynamic changes of the audio when playing different audio content The state loudness performance is tested to improve the accuracy of the test; each audio energy value is analyzed in the frequency domain to obtain multiple energy main frequency information, and the harmonic distortion evaluation value is calculated based on the multiple energy main frequency information. For each energy main frequency information, its harmonic components can be analyzed, including the fundamental frequency and its integer multiple harmonics, which can measure the relationship between the harmonics and the fundamental frequency in the actual signal, test the purity of the audio, and calculate the sound quality evaluation value based on the multiple harmonic distortion evaluation values. The multiple harmonic distortion evaluation values ​​are combined and a weighted average method is used to obtain a comprehensive sound quality evaluation value. The sound quality evaluation value reflects the harmonic distortion level of the overall audio, provides a quantitative measurement of the audio quality, and can accurately test the sound quality;Finally, the total evaluation value is calculated based on the hearing evaluation value, loudness change value and sound quality evaluation value. The audio played by the speaker can be judged more comprehensively and accurately based on the hearing evaluation value, loudness change value and sound quality evaluation value, and the sound quality test can be more accurate. ;

[0140] In one embodiment, the step of segmenting the output audio feature information to obtain a plurality of segmented audio information includes:

[0141] S201, obtaining output audio feature information;

[0142] S202, converting the output audio feature information to obtain mono feature information;

[0143] S203, obtaining a plurality of short time frames according to the monophonic feature information;

[0144] S204, acquiring multiple short-time sound energy values ​​according to multiple short-time frames;

[0145] S205, determining whether each short-time sound energy value is within a preset energy range; if the short-time sound energy value is not within the preset energy range and exceeds an upper limit of the preset energy range, taking the corresponding short-time sound energy value as a starting segmentation point;

[0146] S206, if the short-time sound energy value is not within the preset energy range, but within the lower limit of the preset energy range, the corresponding short-time sound energy value is used as the end segmentation point;

[0147] S207: Segment the output audio feature information based on the start segmentation point and the end segmentation point to obtain a plurality of segmented audio information.

[0148] As described in the above steps S201-S207, the present application simplifies the audio data and reduces the processing complexity by converting the output audio feature information into a mono channel, while retaining the main features of the audio, and divides the continuous mono feature information into a series of short time frames. The division into short time frames makes the audio data easier to process, and a fast Fourier transform can be performed on each short time frame to convert the audio signal in the time domain into a frequency domain representation, and the power spectrum data of each short time frame is obtained in the frequency domain. The power spectrum data represents the power energy of each frequency domain, and the power energy is output to obtain a short-time sound energy value; by traversing By going through all short-time frames, the energy changes of the audio signal can be fully analyzed to ensure that no possible segmentation points are missed; the sound start point is determined based on the short-time sound energy value exceeding the preset threshold, which can effectively identify the valid sound segments in the audio; and the sound end point is determined based on the short-time sound energy value dropping below the preset threshold, which helps to accurately divide the audio segments and avoid mistaking silence or background noise for valid sounds; valid speech signals (such as human speech, vocal music, etc.) usually have higher energy values, because the sound will produce stronger sound waves. In contrast, the energy values ​​of silent segments or segments containing only background noise are usually The energy of background noise is relatively weak, and it is not as concentrated and continuous as the effective speech signal. By setting a preset energy range, it is possible to distinguish between high-energy effective speech and low-energy silence or background noise. When it is detected that the short-term sound energy value suddenly rises from below the preset energy range and exceeds the upper limit, it can be determined as the beginning of the sound. This change indicates that effective speech activity has begun. The previous low-energy stage may be silence or only contain background noise. Similarly, when the short-term sound energy value drops and is below the lower limit of the preset energy range, it can be determined as the end of the sound, which indicates that there is The effective speech activity has ended, and the subsequent segment may be silence or background noise. By determining the start and end points of the sound based on a preset energy range, the audio signal can be effectively divided into segments containing effective sound and segments containing silence or background noise; it is used to solve the problem that background noise may affect the accuracy of the segmentation algorithm, especially when the speech or music in the audio is similar to the noise level; the application defines the start and end segmentation points of the audio through short-time sound energy, which can effectively identify significant events in the audio, such as the beginning and end of speech, so as to determine the segmentation points of the audio and improve the accuracy of audio segmentation.

[0149] In one embodiment, the steps of obtaining a frequency feature at each preset interception time point according to a plurality of segmented audio information, obtaining an audio energy value at each preset interception time point according to each frequency feature, and calculating a hearing evaluation value according to each audio energy value include:

[0150] S301, obtaining the segmented audio information corresponding to each preset interception time point;

[0151] S302, obtaining multiple frequency features by performing short-time Fourier transform on multiple segmented audio information;

[0152] S303, performing spectrum analysis on multiple frequency features to obtain a frequency peak value corresponding to each preset interception time point;

[0153] S304: Calculate the audio energy value corresponding to the preset interception time point according to the frequency peak value. The audio energy value calculation formula is:

[0154] A=|(logX)| 2 ;

[0155] Where A is the audio energy value, X is the frequency peak;

[0156] S305, obtaining the audio energy value at each preset interception time point;

[0157] S306: Calculate a hearing evaluation value according to each audio energy value. The calculation formula of the hearing evaluation value is:

[0158]

[0159] Among them, C is the hearing evaluation value, A1 is the audio energy value at the first preset interception time point, A2 is the audio energy value at the second preset interception time point, and A i is the audio energy value at the i-th preset interception time point.

[0160] As described in the above steps S301-S306, the present application improves the efficiency of analyzing audio data in time by segmenting the audio into multiple audio segments. The segmentation can be based on the time interval, and the frequency characteristics of the audio in different time periods can be provided by short-time Fourier transform, which can capture the frequency changes of the audio in time and accurately detect the rapid changes of the tone or audio characteristics in a short period of time. The main frequency components of each preset time point can be identified through spectrum analysis and frequency peak detection, and the main frequency of the audio in a specific time period can be determined, thereby improving the accuracy of detection. By calculating the frequency peak corresponding to the The audio energy value can quantify the audio intensity or signal intensity at each preset time point, and is used to measure the intensity of the audio at a specific time point. The audio energy value corresponds to the preset interception time point, and can reflect the intensity changes of the audio in different time periods. For example, 10 segmented audio information is obtained, and each segmented audio information is short-time Fourier transformed to obtain the frequency domain information of each segmented audio information. The frequency domain information is subjected to spectrum analysis to obtain the frequency peak value at each preset interception time point (for example, every 2 seconds), and then the energy value is calculated based on the frequency peak value. For example, if the frequency peak value of an audio clip is 1000Hz, its energy value is Then, the audio energy value of each preset interception time point is obtained, and the hearing evaluation value is calculated according to the audio energy value of each audio. For example, if the audio energy values ​​of 10 preset interception time points of an audio are 20, 21, 22, ..., 30 respectively, then its hearing evaluation value is

[0161] The hearing evaluation value of the audio can be obtained, thereby making a quantitative evaluation of the quality or effect of the audio, and can effectively display the dynamic range of the audio signal and the intensity changes over time. The calculation of the hearing evaluation value takes into account the audio energy value at each time point to quantify the user's auditory perception of the audio, which can solve the problem of unquantifiable problems in hearing tests and the inability to accurately conduct hearing tests. By calculating the hearing evaluation value, the accuracy of judging the quality of the sound can be improved.

[0162] In one embodiment, the steps of obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value include:

[0163] S401, obtaining the audio energy value at each preset interception time point;

[0164] S402: Calculate the volume value according to the audio energy value. The volume value calculation formula is:

[0165]

[0166] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0167] S403, calculating the loudness change value according to the volume value at each preset interception time point, the loudness change value calculation formula is:

[0168] B i =D i -D i-1 ;

[0169] Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point;

[0170] S404: Calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0171]

[0172] Where L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, B2 is the loudness change value at the second preset interception time point, and B n The loudness change value at the nth preset cutoff time point.

[0173] As described in the above steps S401-S404, the present application calculates the volume value at each preset time point through the audio energy value. For example, in audio signal processing, the audio signal can be intercepted at a preset time interval (such as 100 sample points per second), and the audio energy value at each time point is calculated. The volume value is used to represent the volume intensity at different preset interception time points, which is helpful for analyzing the dynamic range of the audio, and the loudness change value represents the change in audio energy between adjacent time points. By calculating the loudness change value, the loudness change at the preset interception time point can be judged, such as the sudden increase or decrease of the audio segment. After obtaining the volume value at each preset interception time point, the loudness change value at each time point can be calculated by the difference between the volume values ​​at adjacent time points. Specifically, if the audio energy value A at the i-th time point has been calculated i and the audio energy value A at the i-1th time point i-1 , then the loudness change value B at the i-th time point i It can be calculated by the following formula: i =A i -A i-1 , which helps to identify the dynamic characteristics in the audio and provide more detailed volume intensity information. By calculating the loudness change value, the change in volume intensity output by the sound system at different time points can be captured. For example, after obtaining the loudness change value of each preset interception time point, the loudness change value can be calculated based on these loudness change values. Specifically, the loudness change value can be calculated by adding up all the loudness change values ​​and integrating the result. If all the loudness change values ​​{B1, B2, ..., B ... n}, the loudness change value L can be calculated by the following formula:

[0174]

[0175] Where n is the number of preset interception time points. By calculating the loudness change value, the dynamic loudness characteristics of the audio signal can be further analyzed, which can solve the problem that the volume changes at multiple preset interception time points cannot be judged in the existing test. After obtaining the loudness change value of each preset interception time point, the loudness of each preset interception time point can be accurately detected. Through loudness detection, the accuracy of the sound quality test can be increased.

[0176] In one embodiment, each audio energy value is subjected to frequency domain analysis to obtain multiple energy main frequency information, and a harmonic distortion evaluation value is calculated according to the multiple energy main frequency information, and the step of calculating the sound quality evaluation value according to the multiple harmonic distortion evaluation values ​​includes:

[0177] S501, obtaining multiple audio energy values;

[0178] S502, performing frequency domain analysis on multiple audio energy values ​​to obtain multiple energy main frequency information;

[0179] S503, obtaining multiple amplitude values ​​by analyzing the spectrum of each energy main frequency information;

[0180] S504, obtaining the frequency value corresponding to each amplitude value;

[0181] S505. Calculate the harmonic distortion value according to the multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is:

[0182]

[0183] Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value;

[0184] S506, obtaining multiple harmonic distortion values;

[0185] S507, normalizing each harmonic distortion value to obtain a distortion value;

[0186] S508, obtaining a weight of each distortion value;

[0187] S509: Calculate a sound quality evaluation value for multiple distortion values ​​and weights, using the following calculation formula:

[0188]

[0189] Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and f j is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

[0190] As described in the above steps S501-S506, the present application obtains multiple audio energy values ​​and performs frequency domain analysis on the multiple audio energy values ​​to obtain multiple energy main frequency information. The energy main frequency information includes frequency domain information. The energy main frequency information can reflect the sound characteristics of the audio at different moments. Spectral analysis can decompose the audio energy value in the frequency domain to obtain the amplitude information of different frequency components. By performing spectrum analysis on each audio energy value, the amplitude values ​​at all preset interception time points can be obtained. The amplitude values ​​correspond to the amplitudes of the frequency components. The frequency components at the preset interception time points are analyzed by obtaining the frequency values ​​corresponding to each amplitude value. At the same time, because distortion usually causes changes in frequency components, harmonic distortion values ​​are calculated for different amplitude values. The harmonic distortion values ​​can effectively evaluate the degree of harmonic distortion in the signal. Harmonic distortion values ​​of different ranges can be mapped to a unified scale through normalization, so that multiple harmonic distortion values ​​are comparable, simplifying subsequent processing steps, and more effectively integrating multiple harmonic distortion values. Distortion values ​​are obtained, and sound quality evaluation values ​​are calculated for multiple distortion values. The sound quality evaluation value can more comprehensively evaluate the overall quality of the audio signal. Taking into account the weights of different distortion sources, it can more accurately reflect the user's subjective perception of the audio and provide more accurate results for the quantitative evaluation of audio quality. For example, based on obtaining 10 segmented audio information about a piece of audio, frequency domain information is obtained based on the segmented audio information, and then the frequency domain information is disassembled through spectrum analysis technology to obtain the amplitude information of each frequency component. By analyzing the amplitude values ​​of different frequency components, 10 amplitude values ​​are obtained. According to the formula: The harmonic distortion values ​​of each segmented audio information are calculated and normalized to map them to a uniform scale, so that multiple distortion values ​​are comparable and subsequent comprehensive evaluation can be more convenient.

[0191] Finally, multiple distortion values ​​and weights are input into the sound quality evaluation value calculation formula:

[0192] The calculation takes into account the weights of different distortion sources. For example, the weights can be adjusted according to people's sensitivity to pitch, timbre and dynamic range. By combining multiple distortion values ​​and considering the weights of different distortion sources, more accurate results can be provided for the quality assessment of music recordings. At the same time, it can avoid repeatedly obtaining the amplitude value of each preset interception time point, reduce data redundancy in the entire test process, and improve test efficiency.

[0193] In one embodiment, the step of calculating the total evaluation value from the hearing evaluation value, the loudness change value and the sound quality evaluation value comprises:

[0194] S601, obtaining a hearing evaluation value, and normalizing the hearing evaluation value to obtain a normalized hearing value;

[0195] S602, obtaining a loudness change value, and normalizing the loudness change value to obtain a normalized loudness value;

[0196] S603, obtaining a sound quality evaluation value, and normalizing the sound quality evaluation value to obtain a normalized sound quality value;

[0197] S604, obtaining weights of a normalized hearing perception value, a normalized loudness value, and a normalized sound quality value;

[0198] S605: Calculate the total evaluation value by using the normalized hearing value, the normalized loudness value, and the normalized sound quality value, wherein the total evaluation calculation formula is:

[0199]

[0200] Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

[0201] As described in steps S601-S605, by comprehensively considering the hearing, loudness and sound quality, the overall evaluation calculation formula can more comprehensively evaluate the overall output audio information of the audio system and provide a more comprehensive and objective audio evaluation. For example, for an audio system, we obtain the following values ​​through the hearing evaluation, loudness change and sound quality evaluation:

[0202] Normalized hearing value: E1 = 80; Normalized loudness value: E2 = 70; Normalized sound quality value: E3 = 90; Weight of normalized hearing value: R1 = 0.3; Weight of normalized loudness value: R2 = 0.2; Weight of normalized sound quality value: R3 = 0.5;

[0203] Substituting these values ​​into the overall evaluation formula, we get:

[0204] S=80*0.3+70*0.2+90*0.5

[0205] =24+14+45

[0206] =83

[0207] Therefore, the total evaluation value of the audio equipment is 83, which can reflect the overall output audio information of the audio equipment and provide a more comprehensive and objective audio evaluation. It can avoid the situation where the hearing evaluation value, loudness change value and sound quality evaluation value are not on a unified scale. In this way, the hearing evaluation value, loudness change value and sound quality evaluation value are comparable, making it easier to conduct subsequent comprehensive evaluation.

[0208] The present application also provides a sound quality testing system for a speaker, comprising:

[0209] Acquisition module 1, used to obtain output audio feature information of the speaker;

[0210] Segmentation module 2, used to obtain multiple preset interception time points, and obtain the time intervals between the multiple preset interception time points, and segment the output audio feature information according to the time interval to obtain multiple segmented audio information;

[0211] A first calculation module 3 is used to obtain a frequency feature at each preset interception time point according to the multiple segmented audio information, obtain an audio energy value at each preset interception time point according to each frequency feature, and calculate a hearing evaluation value according to each audio energy value;

[0212] A second calculation module 4 is used to obtain a volume value corresponding to a preset interception time point according to the audio energy value, obtain a loudness change value at each preset interception time point according to each volume value, and calculate a loudness evaluation value according to each loudness change value;

[0213] The third calculation module 5 is used to perform frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculate harmonic distortion evaluation values ​​according to the multiple energy main frequency information, and calculate sound quality evaluation values ​​according to the multiple harmonic distortion evaluation values;

[0214] A fourth calculation module 6, used to calculate a total evaluation value according to the hearing evaluation value, the loudness change value and the sound quality evaluation value;

[0215] A judging module 7 is used to judge whether the total evaluation value is within a preset range;

[0216] If it is, the sound quality test result of the audio system is determined to be qualified;

[0217] If not, the sound quality test result of the audio system is judged to be unqualified.

[0218] In one embodiment, the second calculation module includes:

[0219] An acquisition unit, used to acquire the audio energy value at each preset interception time point;

[0220] The first calculation unit is used to calculate the volume value according to the audio energy value. The volume value calculation formula is:

[0221]

[0222] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0223] The second acquisition unit is used to calculate the loudness change value according to the volume value at each preset interception time point, and the loudness change value calculation formula is:

[0224] B i =D i -D i-1 ;

[0225] Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point;

[0226] The third acquisition unit is used to calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0227]

[0228] Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

[0229] In one embodiment, the second calculation module includes:

[0230] A first acquisition unit, used to acquire the audio energy value at each preset interception time point;

[0231] The first calculation unit is used to calculate the volume value according to the audio energy value. The volume value calculation formula is:

[0232]

[0233] Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point;

[0234] The second calculation unit is used to calculate the loudness change value according to the volume value at each preset interception time point, and the loudness change value calculation formula is:

[0235] B i =D i -D i-1 ;

[0236] Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1The audio energy value at the i-1th preset interception time point;

[0237] The third calculation unit is used to calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is:

[0238]

[0239] Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

[0240] In one embodiment, the third calculation module includes:

[0241] A second acquisition unit, used to acquire multiple audio energy values;

[0242] A first analysis unit, configured to perform frequency domain analysis on a plurality of audio energy values ​​to obtain a plurality of energy main frequency information;

[0243] A second analysis unit is used to obtain multiple amplitude values ​​by analyzing the spectrum of each energy main frequency information;

[0244] A third acquisition unit is used to acquire the frequency value corresponding to each amplitude value;

[0245] The fourth calculation unit is used to calculate the harmonic distortion value according to the multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is:

[0246]

[0247] Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value;

[0248] A fourth acquisition unit, used to acquire a plurality of harmonic distortion values;

[0249] A third analysis unit is used to normalize each harmonic distortion value to obtain a distortion value;

[0250] A fifth acquisition unit, used to acquire a weight of each distortion value;

[0251] The fifth calculation unit is used to calculate the sound quality evaluation value for the multiple distortion values ​​and weights, and the calculation formula is:

[0252]

[0253] Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and fj is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

[0254] In one embodiment, the fourth calculation module includes:

[0255] A sixth acquisition unit, used to acquire a hearing evaluation value, and normalize the hearing evaluation value to obtain a normalized hearing value;

[0256] a seventh acquisition unit, configured to acquire a loudness change value, and perform normalization processing on the loudness change value to obtain a normalized loudness value;

[0257] an eighth acquisition unit, configured to acquire a sound quality evaluation value, and perform normalization processing on the sound quality evaluation value to obtain a normalized sound quality value;

[0258] A ninth acquisition unit, used to acquire weights of the normalized hearing perception value, the normalized loudness value, and the normalized sound quality value;

[0259] The sixth calculation unit is used to calculate the total evaluation value of the normalized hearing value, the normalized loudness value and the normalized sound quality value, wherein the total evaluation calculation formula is:

[0260]

[0261] Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

[0262] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0263] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0264] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent results or equivalent process changes made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for testing the sound quality of an acoustic device, characterized in that: include: Get the output audio feature information of the speaker; Segmenting the output audio feature information to obtain multiple segmented audio information; Obtain frequency features at each preset interception time point according to the plurality of segmented audio information, obtain an audio energy value at each preset interception time point according to each frequency feature, and calculate a hearing evaluation value according to each audio energy value; Obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value; Perform frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculate harmonic distortion evaluation values ​​based on the multiple energy main frequency information, and calculate sound quality evaluation values ​​based on the multiple harmonic distortion evaluation values; Calculate the total evaluation value according to the hearing evaluation value, loudness change value and sound quality evaluation value; Determining whether the total evaluation value is within a preset range; If it is, the sound quality test result of the audio system is determined to be qualified; If not, the sound quality test result of the audio system is judged to be unqualified; The steps of obtaining a frequency feature at each preset interception time point according to a plurality of segmented audio information, obtaining an audio energy value at each preset interception time point according to each frequency feature, and calculating a hearing evaluation value according to each audio energy value include: Obtain the segmented audio information corresponding to each preset interception time point; Multiple frequency features are obtained by short-time Fourier transforming multiple segmented audio information; Performing spectrum analysis on multiple frequency features to obtain the frequency peak value corresponding to each preset interception time point; The audio energy value corresponding to the preset interception time point is calculated according to the frequency peak value. The energy value calculation formula is: A=|(logX)| 2 ; Where A is the audio energy value, and X is the frequency peak; Get the audio energy value at each preset interception time point; The hearing evaluation value is calculated according to each audio energy value. The hearing evaluation value calculation formula is: Among them, C is the hearing evaluation value, A1 is the audio energy value at the first preset interception time point, A2 is the audio energy value at the second preset interception time point, and A i is the audio energy value at the i-th preset interception time point; The step of obtaining a volume value corresponding to a preset interception time point according to the audio energy value, obtaining a loudness change value at each preset interception time point according to each volume value, and calculating a loudness evaluation value according to each loudness change value includes: Get the audio energy value at each preset interception time point; The volume value is calculated according to the audio energy value. The volume value calculation formula is: Among them, D i is the volume value at the i-th preset interception time point, A i is the audio energy value at the i-th preset interception time point; The loudness change value is calculated based on the volume value at each preset interception time point. The loudness change value calculation formula is: B i =D i -D i-1 ; Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point; The loudness evaluation value is calculated according to the multiple loudness change values. The loudness evaluation value calculation formula is: Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point; The steps of performing frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculating harmonic distortion evaluation values ​​according to the multiple energy main frequency information, and calculating sound quality evaluation values ​​according to the multiple harmonic distortion evaluation values ​​include: Get multiple audio energy values; Perform frequency domain analysis on multiple audio energy values ​​to obtain multiple energy main frequency information; By analyzing the spectrum of each energy main frequency information, multiple amplitude values ​​are obtained; Get the frequency value corresponding to each amplitude value; The harmonic distortion value is calculated based on multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is: Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value; Get multiple harmonic distortion values; And normalize each harmonic distortion value to obtain the distortion value; Get the weight of each distortion value; The sound quality evaluation value is calculated for multiple distortion values ​​and weights, and the calculation formula is: Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and f j is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

2. The sound quality testing method of the acoustics according to claim 1, characterized in that: The step of segmenting the output audio feature information to obtain a plurality of segmented audio information comprises: Get output audio feature information; Convert the output audio feature information to obtain mono feature information; Obtaining multiple short time frames according to the single channel feature information; Acquire multiple short-time sound energy values ​​according to multiple short-time frames; Determine whether each short-time sound energy value is within a preset energy range. If the short-time sound energy value is not within the preset energy range and exceeds the upper limit of the preset energy range, use the corresponding short-time sound energy value as the starting segmentation point; If the short-time sound energy value is not within the preset energy range, and is within the lower limit of the preset energy range, the corresponding short-time sound energy value is used as the end segmentation point; The output audio feature information is segmented based on the start segmentation point and the end segmentation point to obtain a plurality of segmented audio information.

3. The sound quality testing method of the acoustics according to claim 1, characterized in that: The step of calculating the total evaluation value by the hearing evaluation value, the loudness change value and the sound quality evaluation value comprises: Acquiring a hearing evaluation value, and normalizing the hearing evaluation value to obtain a normalized hearing value; Acquire a loudness change value, and perform normalization processing on the loudness change value to obtain a normalized loudness value; Acquiring a sound quality evaluation value, and normalizing the sound quality evaluation value to obtain a normalized sound quality value; Obtain weights of normalized hearing perception value, normalized loudness value, and normalized sound quality value; The normalized hearing value, normalized loudness value and normalized sound quality value are used to calculate the total evaluation value, where the total evaluation calculation formula is: Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

4. A sound quality testing system for an acoustic device is used to execute the sound quality testing method for an acoustic device as claimed in any one of claims 1 to 3, characterized in that: include: An acquisition module, used to obtain output audio feature information of the speaker; A segmentation module, used to obtain a plurality of preset interception time points, and obtain the time intervals between the plurality of preset interception time points, and segment the output audio feature information according to the time intervals to obtain a plurality of segmented audio information; A first calculation module is used to obtain a frequency feature at each preset interception time point according to the plurality of segmented audio information, obtain an audio energy value at each preset interception time point according to each frequency feature, and calculate a hearing evaluation value according to each audio energy value; A second calculation module is used to obtain a volume value corresponding to a preset interception time point according to the audio energy value, obtain a loudness change value at each preset interception time point according to each volume value, and calculate a loudness evaluation value according to each loudness change value; A third calculation module is used to perform frequency domain analysis on each audio energy value to obtain multiple energy main frequency information, calculate harmonic distortion evaluation values ​​according to the multiple energy main frequency information, and calculate sound quality evaluation values ​​according to the multiple harmonic distortion evaluation values; A fourth calculation module, used to calculate a total evaluation value according to the hearing evaluation value, the loudness change value and the sound quality evaluation value; A judgment module, used to judge whether the total evaluation value is within a preset range; If it is, the sound quality test result of the audio system is determined to be qualified; If not, the sound quality test result of the audio system is judged to be unqualified.

5. The sound quality testing system of the sound system according to claim 4, characterized in that: The second calculation module includes: A first acquisition unit, used to acquire the audio energy value at each preset interception time point; The first calculation unit is used to calculate the volume value according to the audio energy value. The volume value calculation formula is: Among them, D i is the volume value at the i-th preset interception time point, A i The audio energy value at the i-th preset interception time point The second calculation unit is used to calculate the loudness change value according to the volume value at each preset interception time point, and the loudness change value calculation formula is: B i =D i -D i-1 ; Among them, B i is the loudness change value at the i-th preset interception time point, D i is the audio energy value at the i-th preset interception time point, D i-1 The audio energy value at the i-1th preset interception time point; The third calculation unit is used to calculate a loudness evaluation value according to the multiple loudness change values. The loudness evaluation value calculation formula is: Wherein, L is the loudness evaluation value, n is the number of preset interception time points, B1 is the loudness change value at the first preset interception time point, and B2 is the loudness change value at the second preset interception time point; B i is the loudness change value at the i-th preset interception time point.

6. The sound quality testing system of the sound system according to claim 4, characterized in that: The third computing module comprises: A second acquisition unit, used to acquire multiple audio energy values; A first analysis unit, configured to perform frequency domain analysis on a plurality of audio energy values ​​to obtain a plurality of energy main frequency information; A second analysis unit is used to obtain multiple amplitude values ​​by analyzing the spectrum of each energy main frequency information; A third acquisition unit is used to acquire the frequency value corresponding to each amplitude value; The fourth calculation unit is used to calculate the harmonic distortion value according to the multiple amplitude values ​​and frequency values, wherein the harmonic distortion value calculation formula is: Where W is the harmonic distortion value, H1 is the first amplitude value, H2 is the second amplitude value, H m is the mth amplitude value; A fourth acquisition unit, used to acquire a plurality of harmonic distortion values; A third analysis unit is used to normalize each harmonic distortion value to obtain a distortion value; A fifth acquisition unit, used to acquire a weight of each distortion value; The fifth calculation unit is used to calculate the sound quality evaluation value for the multiple distortion values ​​and weights, and the calculation formula is: Among them, Q is the sound quality evaluation value, f1 is the weight of the first distortion value, f2 is the weight of the second distortion value, and f j is the weight of the jth distortion value, W1 is the first distortion value, W2 is the second distortion value, and W j is the jth distortion value.

7. The sound quality testing system of an acoustic device according to claim 4, characterized in that: The fourth computing module comprises: A sixth acquisition unit, used to acquire a hearing evaluation value, and normalize the hearing evaluation value to obtain a normalized hearing value; a seventh acquisition unit, configured to acquire a loudness change value, and perform normalization processing on the loudness change value to obtain a normalized loudness value; an eighth acquisition unit, configured to acquire a sound quality evaluation value, and perform normalization processing on the sound quality evaluation value to obtain a normalized sound quality value; A ninth acquisition unit, used to acquire weights of the normalized hearing perception value, the normalized loudness value, and the normalized sound quality value; The sixth calculation unit is used to calculate the total evaluation value of the normalized hearing value, the normalized loudness value and the normalized sound quality value, wherein the total evaluation calculation formula is: Among them, S is the total evaluation value, E1 is the normalized listening value, E2 is the normalized loudness value, E3 is the normalized sound quality value, R1 is the weight of the normalized listening value, R2 is the weight of the normalized loudness value, and R3 is the weight of the normalized sound quality value.

Images