A temporal broadband optimization algorithm for speaker placement in a personal sound zone system

Abstract

This invention relates to the field of sound field reproduction technology, specifically, to a temporal broadband optimization algorithm for speaker placement in a personal audio zone system, comprising the following steps: assuming L0 speakers are located at L0 candidate positions, the l-th speaker is muted, and the temporal filter coefficients of the remaining L0-1 speakers are calculated; based on the obtained optimal speaker filter coefficients, the performance index of the remaining L0-1 speakers is calculated when the l-th speaker is muted; the performance index corresponding to l is calculated for l taking values ​​of 1, 2, ..., L0, and the speaker position corresponding to the minimum performance index is found and denoted as l0; the l0-th speaker is removed from the candidate positions, and the remaining L0-1 speaker positions are taken as new candidate positions; for the new L0-1 speaker candidate positions, steps one to three are repeated for iterative optimization until the pre-set L speaker positions are selected.

Description

Technical Field

[0001] This invention relates to the field of sound field reproduction technology, specifically, to a time-domain broadband optimization algorithm for loudspeaker placement in a personal audio zone system. Background Technology

[0002] A personal zone system refers to a system that uses a speaker array to create two or more independent sound zones that do not interfere with each other. This allows users in different locations within the same physical space to hear only the music played by their own device without being disturbed by other users' devices. Personal zone systems avoid the discomfort caused by prolonged headphone use and overcome the drawbacks of traditional speakers that disturb surrounding users. They have broad application prospects in personal audio systems, in-vehicle audio systems, and public venues.

[0003] Currently, the main methods for realizing a personal vocal zone system include acoustic contrast control (Joung-Woo Choi, Yang-Hann Kim. Generation of an acoustically bright zone with an illuminated region using multiple sources. [J]. J. Acoust. Soc. Am., 2002, 111(4): 1695-1700.) and sound pressure matching (Kirkeby, O., and Nelson, PA (1993). Reproduction of plane wave sound fields. [J]. J. Acoust. Soc. Am. 94, 2992–3000.) and a combination of the two (Chang, J.-H., and Jacobsen, F.. Sound field control with a circular double-layer array of loudspeakers. [J]. J. Acoust. Soc. Am., 2012, 131, 4518–4525.).The above methods are usually based on experience and employ specific loudspeaker array placement forms, such as linear arrays (Sipei Zhao, Xiaojun Qiu, Ian Burnett. Acoustic contrast control in an arc-shaped area using a linear loudspeaker array.[J]. J. Acoust. Soc. Am., 2015, 137(2): 1036-1039.), circular arrays (Mincheol Shin, Sung Q. Lee, Filippo M. Fazi, Philip A. Nelson, Daesung Kim, Semyung Wang, Kang Ho Park, Jeongil Seo. Maximization of acoustic energy difference between two spaces.[J]. J. Acoust. Soc. Am., 2010, 128(1): 121-131.), and arc arrays (Qiaoxi Zhu, Philip Coleman, Ming Wu, Jun Yang. Robust acoustic contrast control with reduced in-situ measurement by acoustic modelling. J. Audio Eng.Soc., 2017, 65:460-473, etc.

[0004] Invention patent ZL202011391492.3 discloses an iterative method for optimizing speaker placement in a multi-region sound field playback system based on acoustic contrast control. This method can optimize speaker placement in the frequency domain, resulting in different speaker placements at different frequency points, which makes it difficult to apply in practice. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention discloses an algorithm for speaker placement in a time-domain broadband optimization personal audio zone system. The difference between this algorithm and the method disclosed in the aforementioned invention patent ZL202011391492.3 is that: (1) the proposed method operates in the time domain and optimizes all frequencies at once to obtain a unified speaker placement optimization result, overcoming the shortcomings of the aforementioned method where different optimization results are obtained for each frequency; (2) the broadband acoustic contrast ratio (BAC) defined by the following formula (2) is proposed as the performance index of the optimization algorithm, replacing the single-frequency acoustic contrast ratio index in the aforementioned patent disclosure method.

[0006] The algorithm disclosed in the present invention optimally selects L positions (L < L0) from L0 candidate positions for deploying a speaker array for a personal sound zone system.

[0007] The specific algorithm process is as follows:

[0008] Step 1: Assume that L0 speakers are respectively located at L0 candidate positions. Mute the l-th speaker (1 ≤ l ≤ L0), and calculate the time-domain filter coefficients of the remaining L0 - 1 speakers. The calculation method is:

[0009] w ,

[0012] ,

[0011] , ,

[0013] , , ,

[0014] = Φ[(γR D + αC RTE + βC SUC + λI) -1 R B (1)

[0010] Where, w o represents the optimal filter coefficients of the L0 - 1 speakers, R B = H B T H B / M B , R D = H D T H D / M D ,, H B and H D are respectively the room impulse response matrices from the speaker array to the bright area and the dark area, M B and M D are respectively the number of microphones in the bright area and the dark area, I is the identity matrix, λ is the regularization coefficient, α, β, and γ are weight coefficients, and Φ[*] represents taking the eigenvector corresponding to the largest eigenvalue of the corresponding matrix.

[0011] Step 2: According to the optimal filter coefficients of the speakers obtained in Step 1, calculate the performance index: broadband sound contrast BAC(l) obtained by the remaining L0 - 1 speakers when the l-th speaker is muted. The calculation method is:

[0012]

[0013] Step 3: Repeat Step 1 and Step 2, calculate the performance index BAC(l) corresponding to l taking 1, 2,..., L0 in sequence, and find the speaker position corresponding to the minimum performance index, denoted as l0, that is, l0 = arg min[BAC(l)]. Remove the l0-th speaker from the candidate positions, and take the remaining L0 - 1 speaker positions as the new candidate positions.

[0014] Step 4: For the new L0-1 candidate speaker positions, repeat steps 1 to 3 for iterative optimization until the pre-defined L speaker positions (L...) are selected. <L0)。

[0015] The beneficial effects of this invention are as follows: This invention optimizes the placement of loudspeakers in the time domain with broadband. The personal sound zone system optimized by the proposed iterative algorithm has higher acoustic contrast and lower array output power than existing arrays with the same number of loudspeakers. Attached Figure Description

[0016] Figure 1a This is a schematic diagram of a circular array.

[0017] Figure 1b This is a schematic diagram of an arc-shaped array.

[0018] Figure 2 This invention compares the broadband acoustic contrast performance of an array optimized by the iterative algorithm proposed in this invention and an unoptimized arc array when the number of loudspeakers is the same.

[0019] Figure 3 compares the performance of the array optimized by the iterative algorithm proposed in this invention and the unoptimized arc array when 10 positions are selected from 60 candidate speaker positions for speaker placement. In the figure, (a) shows the sound contrast ratio and (b) shows the array power. Detailed Implementation

[0020] To enhance understanding of the present invention, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. These embodiments are only used to explain the invention and do not limit the scope of protection of the invention.

[0021] Example: An algorithm for speaker placement in a time-domain broadband optimized personal audio zone system. The specific algorithm flow is as follows:

[0022] Step 1: Assuming L0 loudspeakers are located at L0 candidate positions, mute the l-th loudspeaker (1≤l≤L0), and calculate the time-domain filter coefficients for the remaining L0-1 loudspeakers. The calculation method is as follows:

[0023] w o =Φ[(γR D +αC RTE +βC SUC +λI) -1 R B (1)

[0024] Among them, w o R represents the optimal filter coefficients for L0-1 loudspeakers. B =H B T H B / M B ,R D =H D T H D / M D H B and H D These are the room impact response matrices from the speaker array to the bright and dark areas, respectively, M B and M D denoted by the number of microphones in the bright and dark areas, I is the identity matrix, λ is the regularization coefficient, α, β and γ are weighting coefficients, and Φ[*] represents the eigenvector with the largest eigenvalue of the corresponding matrix.

[0025] Step 2: Based on the optimal filter coefficients of the loudspeakers obtained in Step 1, calculate the performance index obtained by the remaining L0-1 loudspeakers when the l-th loudspeaker is muted: broadband acoustic contrast ratio (BAC(l)). The calculation method is as follows:

[0026]

[0027] Step 3: Repeat Step 1 and Step 2 to calculate the performance index BAC(l) corresponding to l taking values ​​of 1, 2, ..., L0 in sequence, and find the speaker position corresponding to the minimum performance index, denoted as l0, i.e., l0 = arg min[BAC(l)]. Remove the l0th speaker from the candidate positions and take the remaining L0-1 speaker positions as new candidate positions.

[0028] Step 4: For the new L0-1 candidate speaker positions, repeat steps 1 to 3 for iterative optimization until the pre-defined L speaker positions (L...) are selected. <L0)。

[0029] This embodiment takes the selection of 10 speakers from 60 speakers to form a multi-zone sound field reproduction system as an example to describe its implementation process in detail.

[0030] 1. Simulation settings

[0031] Sixty candidate loudspeakers were initially evenly distributed on a circular ring with a radius of 1.5 meters. The distance between the bright and dark areas was 0.8 meters, and the radius of each area was 0.14 meters, as shown in Figure 1(a). The optimization objective was to select 10 loudspeakers from these 60 candidates as a personal audio system for sound reproduction, and to compare the reproduction results with an arc array with 10 loudspeakers as shown in Figure 1(b). The room impulse response matrix used in the simulation was obtained from experimental measurements.

[0032] 2. Simulation Results

[0033] During the optimization process, the broadband acoustic contrast performance index of the array optimized by the time-domain broadband algorithm proposed in this invention varies with the number of loudspeakers as follows: Figure 2 The circular interconnected lines are shown. For comparison, the result of the same number of arc arrays is as follows. Figure 2 The figure shows the series lines in the square. As can be seen from the figure, the array optimized by the proposed time-domain broadband algorithm has a higher broadband acoustic contrast ratio than the arc array with the same number of speakers. When 10 speakers are available, the acoustic contrast ratio and array power of the optimized array compared to the arc array as a function of frequency are compared in Figure 3. As can be seen from the figure, the optimized array improves the acoustic contrast ratio by 5-15 dB compared to the arc array, while reducing the array power by 12-28 dB, significantly improving system performance.

[0034] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. An algorithm for optimizing placement of loudspeakers in a time-domain broadband optimized personal audio zone system, characterized by, The method comprises the following steps: Step one: assuming that L0 loudspeakers are located at L0 candidate positions, muting the 1th loudspeaker, calculating the time-domain filter coefficients of the remaining L0-1 loudspeakers, and the calculation method is: w o = Φ[(γR D + αC RTE + βC SUC + λI) -1 R B ] where w o represents the optimal filter coefficients of L0-1 loudspeakers, R B = H B T H B / M B , R D = H D T H D / M D , H B and H D are the room impulse response matrices of the loudspeaker array to the bright zone and the dark zone, M B and M D are the number of microphones in the bright zone and the dark zone, I is the unit matrix, λ is the regularization coefficient, α, β and γ are the weight coefficients, and Φ[*] represents the eigenvector corresponding to the maximum eigenvalue of the corresponding matrix. Step two: according to the optimal filter coefficients of the loudspeakers obtained in step one, calculating the performance index obtained by the remaining L0-1 loudspeakers when the 1th loudspeaker is muted; Step three: repeating step one and step two, calculating the performance index corresponding to 1 taking 1, 2, …, L0 in turn, and finding out the loudspeaker position corresponding to the minimum performance index, denoted as l0, removing the 1th loudspeaker from the candidate positions, and taking the remaining L0-1 loudspeaker positions as new candidate positions; Step four: for the new L0-1 loudspeaker candidate positions, repeating steps one to three for iterative optimization until the pre-set L loudspeaker positions are selected by optimization; In step two, according to the optimal filter coefficients of the loudspeakers obtained in step one, the performance index obtained by the remaining L0-1 loudspeakers when the 1th loudspeaker is muted is calculated: the wideband sound contrast BAC(l), and the calculation method is:

2. The algorithm for speaker placement in a time domain wideband optimized personal audio zone system as claimed in claim 1, wherein, In step one, 1≤l≤L0.

3. The algorithm for speaker placement in a time domain wideband optimized personal audio zone system of claim 2, wherein, In step three, l0=argmin[BAC(l)].

Images