A signal processing method for improving the localization performance of a two-speaker crosstalk cancellation system

Abstract

The present invention discloses a signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system. The method includes the following steps: obtaining the azimuth angles of the left and right speakers in the two-speaker crosstalk cancellation system, and calculating the average amplitude of the input signals of the left and right speakers at any low-frequency band under each target angle; obtaining the average perceived angle under free-field conditions based on the speaker azimuth angle and the target angle; calculating the amplitude that needs to be increased for the input signals of the left and right speakers at this low-frequency band under the condition that the target angle is equal to the average perceived angle; adjusting the low-frequency amplitude of the input signals of the left and right speakers, and performing spectral smoothing on the highest frequency of the input signals of the left and right speakers at this low-frequency band. The present invention can perform corresponding signal processing according to the speaker arrangement angle and the target angle of the crosstalk cancellation system, so as to solve the problem of the decline in the system positioning performance caused by the mismatch between the calculated crosstalk cancellation system and the actual application environment.

Description

Technical Field

[0001] The present invention relates to the field of binaural sound signal reproduction based on loudspeakers, and specifically to a signal processing method for improving the positioning performance of a two-loudspeaker crosstalk cancellation system. Background Art

[0002] The binaural sound signals obtained by binaural pickup or binaural synthesis are originally only suitable for headphone reproduction. However, due to the requirements of the actual application environment, binaural sound signals are often reproduced using two loudspeakers. Different from headphone reproduction of binaural sound signals, when using loudspeakers to reproduce binaural sound signals, the listener's left ear not only receives the signal from the loudspeaker on the same side as the ear (simply referred to as the ipsilateral loudspeaker), but also can receive the signal emitted by the loudspeaker on the opposite side of the ear (simply referred to as the contralateral loudspeaker). The same is true for the right ear. The sound received by the listener's two ears from the contralateral loudspeaker is called cross-talk. The existence of cross-talk destroys the spatial information contained in the binaural sound signal, causing distortion in the sound direction and timbre that the listener can perceive. Therefore, the binaural sound signal needs to be filtered by a crosstalk cancellation filter before being fed to the loudspeaker, so as to suppress the influence of the sound signal emitted by the contralateral loudspeaker on the binaural sound pressure positioning information of the listener.

[0003] Generally, the head-related transfer function (HRTF) is used to calculate the crosstalk cancellation filter, and the head-related transfer function is the transfer function from a non-directional sound source measured or calculated under free-field conditions to the listener's two ears. This means that the most suitable application environment for the crosstalk cancellation system should also be an anechoic chamber. However, in the actual application environment, there are often reflected sounds, which leads to a mismatch between the calculated crosstalk cancellation system and the actual application environment. This mismatch will cause the positioning performance of the crosstalk cancellation system to decrease. The intuitive impact of this is that when the crosstalk cancellation system reproduces sound signals at larger target angles (such as ±60° or ±90°), the angle perceived by the listener is often significantly lower than the target angle, resulting in a serious damage to the listener's listening experience.

[0004] In addition, during the process of calculating the crosstalk cancellation filter, a Kemar artificial head or a rigid sphere is often used as an approximate substitute for the listener's head. In the actual application environment, the listener's head has a fixed shape like Kemar, which often varies from person to person and has personalized head characteristics. This also leads to a mismatch between the calculated crosstalk cancellation system and the actual application environment. Then, since the sound source used in the calculation is a non-directional sound source, the actual sound source used is often a directional sound source, and this mismatch in sound source directivity will also affect the positioning performance of the crosstalk cancellation system. Finally, the lack of dynamic factors in the reproduction system, or the listener not being completely located in the central listening area will also lead to a decrease in the positioning performance of the crosstalk cancellation system.

[0005] In recent years, the literature has proposed solutions to the problem of reduced localization performance of crosstalk cancellation systems, as follows:

[0006] ① It is proposed to use personalized head-related transfer functions to eliminate the negative effects caused by the mismatch of the listener's head model (Majdak P, Masiero B, Fels J. Sound localization in individualized and non-individualized crosstalk cancellation systems[J]. The Journal of the Acoustical Society of America, 2013, 133(4): 2055-2068.); ② Use a dynamic crosstalk cancellation system to compensate for the lack of dynamic factors (Literature: Lentz T. Dynamic crosstalk cancellation for binaural synthesis in virtual reality environments[J]. Journal of the Audio Engineering Society, 2006, 54(4): 283-294.); ③ Receive the listener's binaural transfer function in real time and calculate the crosstalk cancellation filter in real time to generate an adaptive crosstalk cancellation system (Kabzinski T, Jax P. An adaptive crosstalk cancellation system using microphones at the ears[C] / / Audio Engineering Society Convention 147. Audio Engineering Society, 2019). However, these methods do not completely solve all the mismatched factors between the calculation and application of the crosstalk cancellation system (such as methods ① and ②), and cannot completely solve the problem that the listener's perception angle is less than the target angle; or they require complicated hardware facilities and huge computational costs, and bring a poor perception experience to the listener, which is not suitable for the actual application environment (such as method ③, the listener needs to wear a microphone to receive sound signals in real time, and the system has a large amount of real-time calculations).

[0007] Generally speaking, due to the mismatch between the calculated crosstalk cancellation system and the actually applied crosstalk cancellation system, when the crosstalk cancellation system replays the sound signal at a relatively large target angle, the perceived angle is much lower than the target angle, and this problem has not been effectively solved so far. In the crosstalk cancellation system in the actual listening environment, the auditory localization effect of the virtual sound image, especially when replaying the virtual sound image at a relatively large angle, the actual listening perceived angle is still small (narrow), and this problem has not been improved well. Summary of the Invention

[0008] The velocity vector synthesis theory proposed by Gerzon holds that the superposition sound field of multiple correlated sound sources can be analyzed based on the synthesis of velocity vectors. The length of the velocity vector is the average amplitude of the sound source signal, and the direction of the velocity vector is from the origin to the sound source or from the sound source to the origin, depending on the phase relationship between the velocity vectors. Therefore, two out-of-phase sound signals can be synthesized into a more lateral azimuth sound signal at low frequencies. The signal processing method proposed in the present invention broadens the perceived angle of the listener based on the velocity vector synthesis theory to improve the positioning performance of the crosstalk cancellation system.

[0009] The present invention relates to a crosstalk cancellation system, and proposes a signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system. First, obtain the azimuth angles of the left and right speakers in the two-speaker crosstalk cancellation system, and calculate the average amplitude of the speaker input signals below 1500 Hz at each target angle. Secondly, based on the speaker arrangement angle and the target angle, obtain the average perceived angle under free field conditions. Then, based on the velocity vector synthesis theory, calculate the amplitude that needs to be increased for the speaker input signals below 1500 Hz under the condition that the target angle is equal to the average perceived angle. Finally, adjust the low-frequency amplitudes of the left and right speaker input signals, and perform spectral smoothing on the left and right speaker input signals at 1500 Hz to eliminate the timbre distortion caused by spectral mutations, improve the positioning performance of the crosstalk cancellation system, and create a better auditory experience for the listener in the crosstalk cancellation system.

[0010] The object of the present invention is achieved by at least one of the following technical solutions.

[0011] A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system, comprising the following steps:

[0012] S1: Obtain the azimuth angles γ L 、γ R of the left and right speakers in the two-speaker crosstalk cancellation system, and calculate the average amplitude of the left and right speaker input signals at any low frequency band at each target angle

[0013] ​S2: Based on the azimuth angle of the speakers and the target angle, obtain the average perceived angle under free-field conditions.

[0014] S3: Based on the velocity vector synthesis theory, at the target angle and the average perceived angle being equal, calculate the amplitudes that need to be increased for the input signals S L and S R of the left and right speakers in this low-frequency band.

[0015] S4: Adjust the low-frequency amplitudes of the input signals S L and S R of the left and right speakers, and perform spectral smoothing on the input signals S L and S R of the left and right speakers in this low-frequency band.

[0016] Furthermore, in step S1, the azimuth angles of the left and right speakers in the crosstalk cancellation system are γ L and γ R respectively;

[0017] The target angle refers to the target azimuth angle of the virtual sound source reproduced in the crosstalk cancellation system, and its value range is [-90°, 90°];

[0018] When the target angle is , the average amplitudes of the input signal S L of the left speaker and the input signal S L of the right speaker in any low-frequency band are respectively defined as and

[0019] Furthermore, in step S2, the perceived angle refers to the azimuth angle of the sound source perceived by the listener. Under free-field conditions, when the target angle of the crosstalk cancellation system is , the average perceived angle is and can be obtained through the following formula:

[0020]

[0021] As shown in formula (1), if the target angle is less than the angle between the two speakers, that is, then no additional signal processing is required;

[0022] If the target angle is outside the angle between the speakers, that is, or then the target angle is The average perceived angle at this time is It can be obtained through formula (1). At this time, the calculated average perceived angle is Less than the target angle It is necessary to further adjust the low-frequency amplitude of the speaker input signal so that the average perceived angle Is widened to the same as the target angle To improve the positioning performance of the crosstalk cancellation system.

[0023] Furthermore, in step S3, the left speaker input signal S participating in the velocity vector synthesis L Or the right speaker input signal S R Should be out of phase with the sound signal S of the perceived sound source P In this low-frequency band, that is, the phase difference between the two is between [0.5π, π] or [-0.5π, -π].

[0024] Furthermore, in step S3, the velocity vector r of the perceived sound source under free-field conditions p Is synthesized with the velocity vector r of the left speaker L after the amplitude is increased L Or the velocity vector r of the right speaker R R The synthesized velocity vector r v Points to the target angle. By synthesizing the velocity vector r of the perceived sound source p With the velocity vector r of the left speaker L L Or the velocity vector r of the right speaker R R The vector synthesis can obtain the low-frequency amplitude that needs to be increased in the speaker input signal S L Or S R In this low-frequency band Or The velocity vector synthesis can be expressed by the following formula (2):

[0025]

[0026] Based on the above formula, the low-frequency amplitude that needs to be increased in the speaker input signal can be obtained Where, r v Is the velocity vector of the synthesized sound source after amplitude adjustment, and the direction of r v Is the same as the target angle ; u R , u L Are the unit velocity vectors of the right speaker R and the left speaker L participating in the velocity vector synthesis respectively, and the directions are from the corresponding right speaker R or left speaker L to the origin; Are the average amplitudes that need to be increased in the input signals of the right speaker R and the left speaker L in this low-frequency band respectively; u p Is the average perceived angle under free-field conditions is The unit velocity vector of the perceived sound source, with the direction pointing from the origin to the perceived sound source.

[0027] Further, in step S4, the amplitude of the adjusted input signal of speaker i in this low-frequency band is the original amplitude plus the amplitude that needs to be increased , where i = L or R, that is:

[0028]

[0029] where, are different target angles; according to the obtained the input signals of the right speaker and the left speaker are respectively amplitude-adjusted in this low-frequency band.

[0030] Further, in step S4, after the amplitude adjustment of the speaker input signals S L 、S R , a filter should also be used to smooth the spectrum at the highest frequency of the speaker input signals in this low-frequency band to eliminate the timbre distortion caused by spectrum mutation, and the processed sound signal is fed to the corresponding speaker.

[0031] Further, any of the low-frequency bands includes below 1500 Hz, below 1000 Hz or below 700 Hz.

[0032] Compared with the prior art, the advantages and beneficial effects of the present invention include:

[0033] The present invention can quickly calculate the amplitude that the speaker input signal needs to increase, and the signal processing is simple and convenient;

[0034] After the present invention uses the above signal processing method to adjust the amplitude of the speaker input signal, when the target angle is large, it can significantly increase the actual perceived angle of the listener and effectively improve the positioning performance of the crosstalk cancellation system.

[0035] The present invention uses a spectrum smoothing filter to process the speaker input signal, and the signal processing process will not cause timbre distortion. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is the signal processing flow chart of the crosstalk cancellation system in the embodiment of the present invention;

[0037] Figure 2 is the speaker layout diagram in the crosstalk cancellation system in the embodiment of the present invention;

[0038] Figure 3It is the average perceived angle result graph at different target angles in the embodiment of the present invention;

[0039] Figure 4 It is the phase difference between the input signals of the perceived sound source and the left speaker L below 1500 Hz in the embodiment of the present invention;

[0040] Figure 5 It is the velocity vector synthesis graph at different target angles in the embodiment of the present invention. Detailed implementation manners

[0041] The present invention will be further described below in conjunction with embodiments and the accompanying drawings, but the implementation manners of the present invention are not limited thereto.

[0042] Embodiment 1:

[0043] A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system, as Figure 1 shown, includes the following steps:

[0044] S1: Obtain the azimuth angles γ L , γ R of the left and right speakers in the two-speaker crosstalk cancellation system, and calculate the average amplitudes of the input signals of the left and right speakers below 1500 Hz at each target angle

[0045] The azimuth angles of the left and right speakers in the crosstalk cancellation system are γ L and γ R respectively;

[0046] The target angle refers to the target azimuth angle of the virtual sound source reproduced in the crosstalk cancellation system, and its value range is [-90°, 90°];

[0047] When the target angle is , the average amplitudes of the input signal S L of the left speaker and the input signal S L of the right speaker in the crosstalk cancellation system below 1500 Hz are respectively defined as and

[0048] In this embodiment, taking a two-speaker crosstalk cancellation system with a classic stereo layout as an example, the speaker layout in the system is as Figure 2 shown. In the figure, the coordinate origin O is the center position of the listener's head, the x-axis points vertically from the coordinate origin to the listener's front, the y-axis is horizontal to the right, and the azimuth angles γ R = 30°, γ L = -30° of the two speakers in the crosstalk cancellation system, and P represents the sound source perceived by the listener.

[0049] Target angle It refers to the target azimuth angle at which the crosstalk cancellation system reproduces the virtual sound source. After the binaural sound signal is processed by the crosstalk cancellation system, the input signal S of the speaker when the target angle is and S L 、S R The average amplitude below 1500 Hz is It can be any angle within [-90°, 90°]. Table 1 shows the average amplitudes of the left and right speaker input signals below 1500 Hz at different target angles

[0050] Table 1

[0051]

[0052] S2: Based on the speaker azimuth angle and the target angle, obtain the average perceived angle under free-field conditions

[0053] The perceived angle refers to the azimuth angle of the sound source perceived by the listener. Under free-field conditions, when the target angle of the crosstalk cancellation system is the average perceived angle is It can be obtained through the following formula:

[0054]

[0055] As shown in formula (1), if the target angle is less than the angle between the two speakers, that is then no additional signal processing is required;

[0056] If the target angle is outside the angle between the speakers, that is or then the target angle is when the average perceived angle is It can be obtained through formula (1), and the calculated average perceived angle at this time is less than the target angle It is necessary to further adjust the low-frequency amplitude of the speaker input signal so that the average perceived angle is widened to be consistent with the target angle to improve the positioning performance of the crosstalk cancellation system.

[0057] From Figure 3 it can be seen that when the target angle is small ​The average perceived angle is almost exactly the same as the target angle, and the localization performance of the crosstalk cancellation system does not degrade at this time. When the target angle is a relatively large angle ( or ), the average perceived angle is significantly lower than the target angle. For example, when the target angle is 90°, the average perceived angle is only about 60°, which is caused by the degradation of the localization performance of the crosstalk cancellation system in the actual environment. Therefore, in this embodiment, only the cases of relatively large target angles ( or ) need to be subjected to subsequent signal processing.

[0058] S3: Based on the velocity vector synthesis theory, under the condition that the target angle is equal to the average perceived angle , calculate the amplitudes L S R that the input signals of the left and right speakers need to increase below 1500 Hz

[0059] The input signal S L of the left speaker or the input signal S R of the right speaker participating in the velocity vector synthesis should be out of phase with the sound signal S P of the perceived sound source below 1500 Hz, that is, the phase difference between the two is between [0.5π, π] or [-0.5π, -π].

[0060] In a two-speaker crosstalk cancellation system, when reproducing a virtual sound source with a relatively large target angle, the sound signal S P of the perceived sound source is only out of phase with the input signal of the speaker on the opposite side of the target angle below 1500 Hz. For example, when the target angle is 60° or 90°, the input signal S L of the left speaker L with an azimuth angle of -30° is out of phase with the sound signal S P of the perceived sound source. When the target angle is -60° or -90°, the input signal S R of the speaker R with an azimuth angle of 30° is out of phase with the sound signal S P of the perceived sound source. Figure 4 is the phase difference between the perceived sound source and the input signal S L of the left speaker L below 1500 Hz.

[0061] Synthesize the velocity vector r p of the perceived sound source under free-field conditions with the velocity vector r L of the left speaker L after the amplitude increase or the velocity vector r R of the right speaker R, and the synthesized velocity vector r vPoint to the target angle, and by sensing the velocity vector r of the sound source p and the velocity vector r of the left speaker L L or the velocity vector r of the right speaker R R perform vector synthesis to obtain the speaker input signal S L or S R The amplitude to be increased below 1500 Hz or The velocity vector synthesis can be expressed by the following formula (2):

[0062]

[0063] Based on the above formula, the increased low-frequency amplitude of the speaker input signal can be obtained where r v is the velocity vector of the synthesized sound source after amplitude adjustment, and the direction of r v is consistent with the target angle ; u R , u L are the unit velocity vectors of the right speaker R and the left speaker L participating in the velocity vector synthesis respectively, and the direction is from the corresponding right speaker R or left speaker L to the origin; are the average amplitudes to be increased below 1500 Hz of the input signals of the right speaker R and the left speaker L respectively; u p is the unit velocity vector of the perceived sound source with an average perceived angle of in the free field condition, and the direction is from the origin to the perceived sound source.

[0064] The velocity vector synthesis formula can be more intuitively expressed through a vector synthesis diagram, Figure 5 which is the velocity vector synthesis diagram for some target angles. Substituting the data in Table 1 and formula (3) into formula (4), the increased low-frequency amplitude of the speaker input signal can be obtained In this embodiment, the calculated is shown in Table 2.

[0065] Table 2

[0066]

[0067] S4: Adjust the low-frequency amplitudes of the input signals S L , S R of the left and right speakers, and perform spectral smoothing on the input signals S L , S R of the left and right speakers at 1500 Hz.

[0068] The amplitude of the adjusted input signal of speaker i below 1500 Hz is the original amplitude and the amplitude to be increased The sum, i = L or R, that is:

[0069]

[0070] wherein are different target angles; the solution gives as shown in Table 3. According to the input signal S to the speaker L 、S R The amplitude is adjusted at frequencies below 1500 Hz.

[0071] Table 3

[0072]

[0073] After the amplitude of the input signal S to the speaker L 、S R is adjusted, a filter should also be used to smooth the spectrum at 1500 Hz of the input signal to the speaker to eliminate the timbre distortion caused by the spectrum mutation, and the processed sound signal is fed to the corresponding speaker.

[0074] Embodiment 2:

[0075] In this embodiment, taking the crosstalk cancellation system of two speakers arranged at ±45° angles as an example, the azimuth angles γ of the two speakers in the crosstalk cancellation system R = 45°, γ L = -45°. The target angle refers to the target azimuth angle of the crosstalk cancellation system to reproduce the virtual sound source. After the binaural sound signal is processed by the crosstalk cancellation system, the input signals S of the speakers at the target angle of are obtained L 、S R The average amplitude below 1500 Hz is It can be any angle within [-90°, 90°]. Table 4 shows the average amplitudes below 1500 Hz of the left and right speaker input signals at different target angles

[0076] Table 4

[0077]

[0078] The other implementation steps in this embodiment are exactly the same as those in Embodiment 1, and based on the data in Table 4 and the speaker azimuth angles γ R = 45°, γ L ​=-45° gives the low-frequency amplitude below 1500 Hz that the speaker needs to increase

[0079] Embodiment 3:

[0080] In this embodiment, taking the crosstalk cancellation system of two speakers arranged at an angle of ±60° as an example, the azimuth angles γ of the two speakers in the crosstalk cancellation system R = 60°, γ L = -60°. The target angle refers to the target azimuth angle at which the crosstalk cancellation system reproduces the virtual sound source. After the binaural sound signal is processed by the crosstalk cancellation system, the input signals S of the speakers when the target angle is have an average amplitude below 1500 Hz of L 、S R The average amplitude below 1500 Hz is It can be any angle within [-90°, 90°]. Table 5 shows the average amplitudes below 1500 Hz of the left and right speaker input signals at different target angles

[0081] Table 5

[0082]

[0083] The other implementation steps in this embodiment are exactly the same as those in Embodiment 1. Based on the data in Table 5 and the speaker azimuth angles γ R = 60°, γ L = -60° gives the low-frequency amplitude that the speaker needs to increase

[0084] As can be seen from the above examples, the present invention can perform signal processing on the crosstalk cancellation system with different speaker arrangement schemes according to the above steps. Based on the method of vector synthesis, the positioning performance of the crosstalk cancellation system is improved by adjusting the low-frequency amplitude of the speaker input signal. Compared with the traditional method, the present invention has a low calculation cost, fast speed, a significant improvement in positioning performance, and no timbre distortion, and can create a better auditory experience for listeners under the crosstalk cancellation system.

[0085] The above embodiments are the implementation manners with obvious effects of the present invention, but the implementation manners of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement manners and are all included in the protection scope of the present invention.​

Claims

1. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system, characterized in that including the following steps: S1: Obtain the azimuth angles γ L L and γ R R of the left and right speakers in the two-speaker crosstalk cancellation system, and calculate the average amplitudes of the input signals of the left and right speakers at each target angle in any low-frequency band S2: Obtain the average perceived angle under free-field conditions based on the loudspeaker azimuth angle and the target angle S3: Based on the velocity vector synthesis theory, under the condition that the target angle is equal to the average perceived angle , calculate the increased amplitudes of the input signals S L and S R required in this low frequency band The velocity vector r of the perceived sound source under free-field conditions p is combined with the velocity vector r of the left speaker L after the amplitude is increased L or the velocity vector r of the right speaker R R to obtain a combined velocity vector r v pointing to the target angle. By combining the velocity vector r of the perceived sound source p with the velocity vector r of the left speaker L L or the velocity vector r of the right speaker R R the speaker input signal S can be obtained through vector synthesis L or S R The amplitude to be increased in this low-frequency band or The velocity vector synthesis is represented by the following formula (2): Based on the above formula, the low-frequency amplitude by which the input signal of the speaker needs to be increased can be obtained where r v is the velocity vector of the synthesized sound source after amplitude adjustment, and the direction of r v is consistent with the target angle ; u R , u L are the unit velocity vectors of the right speaker R and the left speaker L participating in the synthesis of the velocity vector, respectively, and the directions point from the corresponding right speaker R or left speaker L to the origin; are the average amplitudes by which the input signals of the right speaker R and the left speaker L need to be increased in this low-frequency band, respectively; u p is the unit velocity vector of the perceived sound source with an average perceived angle of under free-field conditions, and the direction points from the origin to the perceived sound source; S4: Adjust the low-frequency amplitudes of the input signals S L , S R to the left and right speakers, and perform spectral smoothing on the input signals S L , S R to the left and right speakers at the highest frequency in this low-frequency band.

2. The signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 1, wherein In step S1, the azimuth angles of the left speaker and the right speaker in the crosstalk cancellation system are γ L and γ R ; Target Angle Refers to the target azimuth angle of the virtual sound source reproduced in the crosstalk cancellation system, and its value range is [-90°, 90°]; When the target angle is , the average amplitudes of the left speaker input signal S L and the right speaker input signal S L in any low frequency band are respectively defined as and 3. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 2, characterized in that, In step S2, the perception angle refers to the azimuth angle of the sound source perceived by the listener. Under free-field conditions, the average perception angle of the crosstalk cancellation system when the target angle is is obtained through the following formula: As shown in formula (1), if the target angle is less than the angle between the two speakers, that is then no additional signal processing is required; If the target angle is outside the angular spread of the speaker, that is or then the average perceived angle when the target angle is is obtained through formula (1), and the calculated average perceived angle at this time is less than the target angle It is necessary to further adjust the low-frequency amplitude of the input signal of the speaker so that the average perceived angle is widened to be consistent with the target angle to improve the positioning performance of the crosstalk cancellation system.​ 4. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 3, characterized in that, In step S3, the left speaker input signal S participating in the speed vector synthesis L or the right speaker input signal S R should be in antiphase with the sound signal S of the perceived sound source P in this low-frequency band, that is, the phase difference between the two is between [0.5π, π] or [-0.5π, -π].

5. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 1, characterized in that In step S4, the amplitude of the input signal of the adjusted speaker i in this low frequency band is the original amplitude plus the amplitude that needs to be increased and i = L or R, that is: Among them, are different target angles; according to the obtained the input signals of the right speaker and the left speaker are respectively amplitude-adjusted in this low-frequency band.

6. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 1, characterized in that In step S4, for the speaker input signals S L and S R after amplitude adjustment, a filter should also be used to smooth the spectrum at the highest frequency in the low frequency band of the speaker input signal, so as to eliminate the timbre distortion caused by spectrum mutation, and feed the processed sound signal to the corresponding speaker.

7. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to any one of claims 1 to 6, characterized in that The frequency band range of any of the low frequency bands does not exceed 1500 Hz.

8. A signal processing method for improving the positioning performance of a two-speaker crosstalk cancellation system according to claim 1, characterized in that, Unit velocity vector u R and u L are in the same direction as the velocity vectors r R and r L respectively, and their magnitudes are both unity (1).

Images