Improved MIMO control methods, related control systems, and computer programs
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-16
- Publication Date
- 2026-08-14
Smart Images
Figure CN122580897A_ABST
Abstract
Description
[0001] The present invention relates to a method for controlling the sound field generated by a sound reproduction system.
[0002] Such a system includes several loudspeakers.
[0003] A loudspeaker (even if it includes more than one transducer) is powered by a single electrical excitation signal, which is usually derived from an amplifier.
[0004] It is known that some loudspeakers can reproduce all frequencies (bass, midrange, and treble), while others can only reproduce midrange or treble frequencies, and still others can only reproduce bass or sub-bass frequencies.
[0005] Such sound reproduction systems are implemented in rooms such as mixing rooms, cinemas, or concert halls, or in the living room of a house or apartment for home theater applications.
[0006] Typically, the sound field is reproduced in the listening area within the room.
[0007] However, the walls of the room (and any obstacles within the room) will cause reflections that can amplify, attenuate, or cancel out the sound field at certain locations in the listening area.
[0008] These changes to the sound field depend on the frequency and on the listener's position in the listening area.
[0009] In order to achieve a satisfactory listening experience from any location within the listening area, it is necessary to control the sound waves emitted by the amplifier of the sound reproduction system.
[0010] This control depends on a great many parameters, such as the number, location, and orientation of the loudspeakers, the geometry of the listening room, the material of the room walls and their frequency-dependent absorption or reflection characteristics, the location of any obstacles, etc.
[0011] One known process is the Multiple-Input Multiple-Output (MIMO) control process, which is described, for example, in the following articles: Kirkeby, O. and Nelson, PA, “Digital filter design for inversion problems insound reproduction”, Audio Engineering Society, Vol. 47 (No. 7 / 8), pp. 583-595, 1999; and Miyoshi, M. and Kaneda, Y., “Inverse filtering of room acoustics”, IEEE Transactions on acoustics, speech, and signal processing, Vol. 36 (No. 2), pp. 145-152, 1988, DOI: 10.1109 / 29.1509.
[0012] The MIMO control process for a sound field involves using several loudspeakers (forming multiple inputs) to optimally control the sound field at multiple measurement points (forming multiple outputs).
[0013] The implementation of the MIMO process requires measuring the cross acoustic response between all possible paths between each of the multiple loudspeakers and each of the multiple microphones during the calibration phase.
[0014] The MIMO process requires defining the target sound field. The target sound field consists of each desired acoustic response at each measurement point.
[0015] The MIMO process involves synthesizing digital filters based on the measured cross-response and target sound field during the optimization phase, and applying these filters to the signal sent to the amplifier of the sound reproduction system during the application phase.
[0016] The synthesis of MIMO filters is performed through an optimization process that involves determining a filter for each loudspeaker such that the combined effect of the filtered loudspeakers produces a modified sound field that approximates the target sound field as closely as possible. .
[0017] In particular, the MIMO control process differs from the single-input multiple-output (SIMO) process, in which the sound field is controlled by a single amplifier (the single input of the process), and the optimization phase determines a single filter to be applied to that single amplifier.
[0018] In the case of a system with multiple loudspeakers, the SIMO process is applied sequentially to each loudspeaker.
[0019] Unlike the MIMO process, the SIMO process does not take into account the combined effect of different loudspeakers used together, because the optimization is calculated separately and independently for each loudspeaker.
[0020] More specifically, the MIMO control process includes adjusting the response based on the target response. and response matrix To calculate the optimal filter or multiple filters .
[0021] matrix elements Let $\frac{s}{p}$ represent the impulse response (or transfer function) between the $s$-th loudspeaker of the sound reproduction system and the $p$-th microphone in a group of microphones arranged in the listening area. Element It originates from the measurement performed when the s-th loudspeaker is excited by the test signal while the other loudspeakers do not emit.
[0022] The filter sought satisfy:
[0023]
[0024] Therefore, we get:
[0025]
[0026] However, directly calculating the filter has several drawbacks:
[0027] First, the matrix It is typically not well-conditioned, and its inversion can lead to instabilities in the calculated filter, which translates into audible artifacts. This drawback can be mitigated by adjusting the cost function used for optimization. Regularization terms are introduced to address this issue. However, this leads to a loss of filter accuracy, the significance of which may depend on the weights of these regularization terms.
[0028] If the number of loudspeakers and microphones is large, that is, if the matrix elements in If the number of elements is very large, the numerical computation cost of inverting them becomes too high for commercial audio equipment calculators to handle, necessitating the use of matrix-based methods. This involves an efficient inversion technique for the structure. However, the number of measurement points and the number of loudspeakers limit the MIMO control process.
[0029] If the goal If the obtained filter does not conform to certain physical constraints related to the propagation of sound in a room, then... It may exhibit audible artifacts and / or strong noncausal components (“prering”). This problem is particularly pronounced when setting overly ideal synthesis targets.
[0030] Several methods have recently been proposed to improve MIMO control processes, thereby at least partially overcoming some or all of these drawbacks.
[0031] Specifically, in patent application FR n°23 00786, the applicant proposed a method comprising: arranging loudspeakers into two networks (“arrays”), the first network including emitting loudspeakers and the second network including non-emitting loudspeakers; and defining a target sound field obtained from the emitting loudspeakers, wherein the non-emitting loudspeakers are used to cancel reflections of waves emitted by the emitting loudspeakers. The matrix obtained by this method... and target This format greatly facilitates the solution of the problem and allows for the acquisition of filters with minimal artifacts and very high control over the sound field within the listening area for all amplifiers in a sound reproduction system. .
[0032] The transmitters are advantageously selected and / or positioned to minimize excitation of the room's lateral modes. They are placed, for example, on a wall referred to as the "front wall" and possibly on a wall opposite the room referred to as the "rear wall." This allows for the creation of a nearly constant wavefront across the entire listening area, thereby minimizing spatial variations in the sound field.
[0033] During the calibration phase, the process also relies on time windowing to preserve only the wavefront emitted by the transmitting loudspeaker.
[0034] This allows for a more compact digital equalization filter with fewer artifacts without excessive regularization.
[0035] The calculated filters help generate wavefronts similar to those corresponding to the target. Therefore, they allow for a reduction in reflections between the front and back walls.
[0036] However, the existence of a wavefront requires the wave to propagate within the room.
[0037] However, this is only possible when the wavelength is approximately twice the characteristic size of the room. This wavelength is associated with a cutoff frequency, which depends on the geometry of the room. More precisely, the cutoff frequency... Corresponding to the first longitudinal pattern of the room, ,in, m / s is the speed of sound, and This is the length of the room. For example, for a room that is 8.5 m long, the first longitudinal mode has a wavelength of 17 m, corresponding to a cutoff frequency of 20 Hz.
[0038] Therefore, the MIMO control process in FR n°23 00786 is optimal only for the propagation regime.
[0039] It is not applicable to waves with wavelengths greater than the wavelength associated with the cutoff frequency.
[0040] However, sound reproduction systems can emit waves over a wide frequency range, including frequencies below the cutoff frequency. This is especially true when it includes a subwoofer or a speaker capable of reproducing low frequencies. These low-frequency sound waves contribute to the user's listening experience.
[0041] Below this cutoff frequency, and in a closed room with completely rigid and sealed walls, the sound field consists primarily of a single mode, where the sound field is quasi-constant throughout the listening area. This is known as the pressurization regime.
[0042] Under ideal conditions in the pressure zone, there is no need to control the spatial changes of the sound field.
[0043] However, a real listening room is not a pressure chamber with perfectly rigid and sealed walls. On the one hand, the walls themselves act as resonant membranes ("drum skin" effect), creating an uneven sound field within the room. On the other hand, a real room is never completely sealed, this is determined solely by ensuring ventilation, not to mention many semi-open rooms. A real room is not a pressure chamber that spontaneously generates a uniform sound field.
[0044] Therefore, it appears that the optimization strategy (i.e., the selection of the target to be achieved) must be adapted to every frequency band of the operating frequency range of the sound reproduction system, including not only the bands above the cutoff frequency but also the bands below the cutoff frequency.
[0045] Therefore, it is desirable to improve the process of the prior art so that it can also allow control of the sound field in the listening area by frequency band, which includes not only frequency bands above the cutoff frequency but also frequency bands below the cutoff frequency.
[0046] The purpose of this invention is to solve this problem.
[0047] Therefore, the present invention relates to a MIMO control method for a sound field in a listening area within a room, the sound field being generated by a sound reproduction system comprising a plurality of loudspeakers and a first computer, the control method being implemented by a control system connected to the plurality of loudspeakers and comprising a second computer and a plurality of microphones positioned in the listening area, the control method comprising: in a calibration phase performed by the second computer: defining at least a first frequency band and a second frequency band within the operating frequency range of the sound reproduction system; defining a first optimization strategy for filters in the first frequency band and a second optimization strategy for filters in the second frequency band; obtaining a response matrix of the loudspeakers and the room by sequentially applying a test signal to each loudspeaker and measuring the basic response of the activated loudspeaker at each microphone location; and determining a global filter based on the response matrix, the first strategy, and the second strategy.
[0048] According to a particular embodiment, the control method includes one or more of the following features (alone or in all technically possible combinations):
[0049] - Defining a first optimization strategy for the filter in the first frequency band and a second optimization strategy for the filter in the second frequency band includes defining a first objective for the first frequency band and a second objective for the second frequency band, and determining a global filter includes first determining a first filter for the first frequency band based on the response matrix and the first objective, and determining a second filter for the second frequency band based on the response matrix and the second objective, and then combining the first filter and the second filter to obtain the global filter.
[0050] - Defining a first optimization strategy for the filter in the first frequency band and a second optimization strategy for the filter in the second frequency band includes defining a first objective for the first frequency band and a second objective for the second frequency band, and determining a global filter includes first combining the first objective and the second objective to obtain a global objective, and then determining the global filter based on the response matrix and the global objective through optimization.
[0051] - Defining a first optimization strategy for the filter in the first frequency band and a second optimization strategy for the filter in the second frequency band includes defining a first target for the first frequency band and a second target for the second frequency band, and determining a global filter includes: filtering the response matrix in the first frequency band and the second frequency band respectively to obtain a first response matrix and a second response matrix respectively; through optimization, determining a first filter for the first frequency band based on the first response matrix and the first target, and determining a second filter for the second frequency band based on the second response matrix and the second target; and then combining the first filter and the second filter to obtain the global filter.
[0052] - Defining a first optimization strategy for the filter in the first frequency band and a second optimization strategy for the filter in the second frequency band includes defining a first target for a specific frequency in the first frequency band and a second target for a specific frequency in the second frequency band, and determining a global target that changes continuously with frequency by interpolating the first target and the second target, and then determining the global filter based on the response matrix and the global target through optimization.
[0053] - The first target at the position p of the synthesized microphone In the first form: Where A_0 is the amplitude of the sound field, and t_0 is the reference emission time of the test signal; in the second form: in, At a single reference point The response of the s-th loudspeaker measured at point ; in the third form: in, It includes A subset of the measurement points of each point, and It is at point The response of the s-th loudspeaker measured at point ; or, in the fourth form: .
[0054] - Location of the synthesized microphone The second objective In the first form: ,in It is the Dirac function. Measurement point The amplitude of the wave at that point, and It depends on the delay relative to the measured position of each loudspeaker emitting the test signal; in the second form: ,in, It is applied to the measured response The time window, and It is a subset of the loudspeakers involved in creating the target; in the third form: Or, in the fourth form: ,in, It is the speaker and the measuring point The distance between them.
[0055] - The room defines a cutoff frequency between the pressure zone state and the propagation zone state, the first frequency band is defined as being below the cutoff frequency, and the second frequency band is defined as being above the cutoff frequency.
[0056] The present invention also relates to a control system adapted to implement the above-described control method when associated with a sound playback system.
[0057] Preferably, the control system includes a first computer and a plurality of microphones, and the sound playback system includes a second computer and a plurality of loudspeakers, wherein the first computer and the second computer are either the same computer that is properly programmed or two different computers connected together to implement the control method.
[0058] The present invention also relates to a computer program comprising software instructions that, when executed by a computer, implement the above-described control method.
[0059] The invention and its advantages will be better understood by reading the detailed description following specific embodiments given only as non-limiting examples, which is made with reference to the accompanying drawings, in which:
[0060] Figure 1 This is a schematic representation of an embodiment of a control system for a sound field according to the present invention;
[0061] Figure 2 yes Figure 1 A representation of the first programming mode of the computer (and a first embodiment of the control method according to the invention) in the form of functional modules;
[0062] Figure 3 yes Figure 1 The second programming mode of the computer (and a second embodiment of the control method according to the invention) is represented in the form of functional modules;
[0063] Figure 4 yes Figure 1 The computer's third programming mode (and the third embodiment of the control method according to the invention) is represented in the form of functional modules;
[0064] Figure 5 yes Figure 1The fourth programming mode of the computer (and the fourth embodiment of the control method according to the present invention) is represented in the form of functional modules;
[0065] Figure 6 Several graphs are shown, each illustrating the amplitude of the processed signal as a function of frequency for different measurement points within the listening area, to allow for comparison between embodiments of the process according to the invention and embodiments of the process according to the prior art; and,
[0066] Figure 7 Different configurations are shown for defining a subset of transmitting loudspeakers for each frequency band after the operating frequency range of the sound reproduction system is subdivided.
[0067] General principles
[0068] The MIMO control process for the sound field according to the present invention involves subdividing the operating frequency range of the sound reproduction system into at least two frequency bands and defining an optimization strategy for each frequency band separately in order to take into account filter optimization parameters that vary with frequency.
[0069] Specifically, the frequency range is subdivided into a first low-frequency band below the room’s characteristic cutoff frequency and a second complementary frequency band or mid / high-frequency band above the cutoff frequency.
[0070] In this way, the optimization strategy specifically for the pressure region state can achieve filter optimization specifically in the first frequency band, and the optimization strategy specifically for the propagation region state can achieve filter optimization specifically in the second frequency band.
[0071] Therefore, a filter is introduced to restore a uniform pressure field for the pressure zone.
[0072] Implementing the MIMO process is meaningful because it utilizes the combined effect of different amplifiers through specially optimized filters, thereby restoring an optimal and uniform sound field in a real, non-ideal room.
[0073] Furthermore, even under ideal conditions, the use of MIMO processes remains meaningful because loudspeakers do not necessarily have the same characteristics, or the ability to generate sound pressure levels in the listening area may differ. Therefore, even if the target sound field is homogeneous, the optimal filter is not a simple bypass filter, because the MIMO process adapts to the specific characteristics of each loudspeaker and, more generally, to the matrix... H The details of the measured response contained therein.
[0074] Finally, although the sound field does not naturally exhibit spatial variation in ideal pressure zone conditions, the MIMO process according to the invention allows for the introduction of controlled variations to attenuate or enhance certain frequency bands in certain areas of the room. By selecting a non-uniform target sound field and using a combined MIMO process utilizing different amplifiers and different measurement points, a sound field with controlled spatial variation can be obtained. This technique allows for the introduction of personalized equalization, for example, based on the preferences of different listeners in the room.
[0075] Advantageously, the complementary frequency band itself is subdivided into several frequency bands, such as the intermediate frequency band and the high frequency band, with specific optimization strategies implemented for each frequency band.
[0076] Optimization parameters that vary with frequency are, for example:
[0077] - Physical parameters that affect sound propagation, such as the geometry of the walls and acoustic absorption. These parameters determine the shape and complexity of the sound field to be optimized;
[0078] - Subjective parameters related to sound perception, which may make it desirable to enhance certain frequency ranges even though the response is flat;
[0079] - Hardware parameters, such as the number of loudspeakers and microphones;
[0080] - Numerical parameters, such as the number of interpolation points per filter, the duration of each filter, the compactness of each filter, and the dynamic range of each filter.
[0081] Figure 1
[0082] Figure 1 This generally represents a control system for the sound field generated by a sound reproduction system.
[0083] The control system 10 is configured to control the sound field in the listening area 19 (preferably three-dimensional).
[0084] The listening area 19 is entirely located within room 20, such as a movie theater, concert hall, or equivalent. Alternatively, the listening area 19 has the same volume as room 20.
[0085] In this embodiment, the control system 10 is associated with the sound playback system 11.
[0086] The audio playback system 11 includes a set Amplifier 12 and computer 18.
[0087] Calculator 18, for example, is integrated into an audio amplifier.
[0088] Each speaker 12 is connected to a dedicated output of the calculator 18 via a dedicated link (wired or wireless).
[0089] In this embodiment, the loudspeaker group includes a plurality of loudspeakers 12, each of which operates across the entire frequency range. However, it is sufficient to have at least one loudspeaker operating in at least a first frequency band and at least one loudspeaker operating in at least a second frequency band.
[0090] For example, the geometry of the listening room 20 allows the cutoff frequency to be determined. This cutoff frequency can then be advantageously considered to define a transition frequency near the cutoff frequency. The transition frequency defines a first frequency band and a second frequency band within the operating frequency range of the sound playback system 11. The first frequency band corresponds to the pressure zone state, and the second frequency band corresponds to the propagation zone state.
[0091] Sound source 9 is connected to the input terminal of calculator 18. This can be, for example, a CD player, turntable, multimedia server, preamplifier, or equivalent.
[0092] The calculator 18 is properly programmed to apply a set of filters during use of the sound playback system 11. Applied to the raw electrical signal received from sound source 9 This group of filters This includes filters for each speaker in the sound reproduction system. Calculator 18 outputs a set of filtered signals. Each filtered signal is transmitted to the associated loudspeaker 12, causing it to generate sound waves.
[0093] The sound field is defined by the superposition of sound waves generated by different speakers in listening area 19.
[0094] The control system 10 is adapted to control the sound playback system 11. It includes... A sensor 16 and a control unit, in this embodiment, the control unit is advantageously a control unit 18.
[0095] Sensor 16 is a microphone, preferably identical to each other. Sensor 16 is preferably an omnidirectional microphone, the sensitivity of which is independent of or almost independent of the incident direction of the sound wave on the microphone.
[0096] Sensor 16 is pre-positioned at different points in the listening area 19 Each sensor defines the measurement location of the sound waves generated by the speaker 12.
[0097] Each sensor 16 is connected to the input of the controller 18 via a dedicated link (which may or may not be wired).
[0098] The processor 18 is appropriately programmed to calculate the filters that will be applied during the usage phase of the system 11 during the calibration phase of the sound playback system 11. .
[0099] More specifically, the processor 18 includes computing devices (such as processor 32) and storage devices (such as memory 30), as well as an input / output interface 34 for connecting the source 9, the speaker 12 and the sensor 16.
[0100] The memory of controller 18 specifically stores computer program instructions, including: program 38, which is executed during the use phase for filtering... Applied to the original signal Then the filtered signal Applied to each loudspeaker 12; and procedure 36, executed during the calibration phase, for implementing the MIMO control process according to the invention to optimally determine the filter. .
[0101] The execution of these programs provides several functions for controller 18, which are represented as functional modules in the accompanying drawings.
[0102] First Embodiment
[0103] exist Figure 2 In the first embodiment shown, the controller 18 is programmed to include a control module 102 and a filter module 104.
[0104] Control module 102 includes:
[0105] - Used to define the operating frequency range of a sound reproduction system Unit 112 of each frequency band;
[0106] - Used to obtain the response matrix Unit 114;
[0107] - Used for each of the multiple frequency bands Define the goal Unit 116;
[0108] - Used based on the response matrix and target To calculate for each frequency band Filter Multiple units 118 b ;as well as,
[0109] - Used for filtering Aggregation into a global filter Unit 120.
[0110] The filter module 104 of processor 18 is configured with a global filter after the calibration stage. Furthermore, during the usage phase, this filtering module is adapted to handle these global filters. Applied to the raw signal received from source 9 In order to obtain a set of filtered signals for driving the loudspeaker. The first in this group Filtered signal The first one applied to system 11 12 speakers.
[0111] The process implemented by the processor is as follows:
[0112] During calibration phase 101, the definition is... 16 sensors and the location of each of these sensors in the listening area.
[0113] Therefore, use One speaker and Each sensor is calibrated.
[0114] Then, module 102 is executed such that:
[0115] - In step 111, execution unit 112 defines the frequency band.
[0116] The operating frequency range of the sound playback system 11 is subdivided into Each frequency band uses a frequency between 1 and... Integers between Perform indexing.
[0117] This division is, for example, achieved by selecting and defining two consecutive frequency bands. and Transition frequency at the boundary between + and 1 This is how it is executed. As a variant, continuous frequency bands partially overlap in the interval near the transition frequency.
[0118] In a simplified embodiment, system 10 includes a human-machine interface that allows the operator to input the number of frequency bands. and transition frequency The operator can use a specialized device to measure the cutoff frequency of the room in which the sound playback system 11 is installed. He can also determine the characteristics of the loudspeaker 12 and its corresponding operating frequency range. This information allows him to define the frequency band configuration of the system 11.
[0119] - In step 113, execution unit 114 obtains the response matrix of sound playback system 11 and room 20. .
[0120] matrix For example, it is aimed at 12 speakers and A multi-channel convolution matrix for each sensor (16 channels). For example, its definition is as follows:
[0121]
[0122] element When the s-th speaker 12 is tested, the signal is... When excited (other speakers are silent), the basic convolution matrix associated with the measurement of the sound wave emitted by the s-th speaker by the p-th sensor 16.
[0123] More precisely, elements With during the time period Measurements taken during this period are associated with the sound waves emitted by the transmitting loudspeaker (wherein, It is the sampling time step, and (It is a predefined integer).
[0124] For example, element It is the Tollitz matrix:
[0125]
[0126] in, It is from 1 to The amplitude measured at sampling time t between the two sampling points.
[0127] - In step 115, execution unit 116 performs operations for each frequency band Define the goal .
[0128] For frequency band goal This corresponds to the control strategy selected for that frequency band.
[0129] Preferably, the control strategy is different for each frequency band.
[0130] Advantageously, as explained in French patent application number 23 00786, a subset of the entire loudspeaker set of system 11 is selected as the loudspeaker transmitter for creating the target.
[0131] To define different target values for each frequency band The set of loudspeakers used for transmitting varies depending on the frequency band being considered.
[0132] Target value For example, in the form of a column vector:
[0133]
[0134] Each component , or point The target at the location represents the point. The target response, of which It is between 1 and Integers between, and This is the number of measurement points where microphone 16 is placed.
[0135] Each component All are of size The tuple represents the frequency band. In the middle, at the point Targeting 1 to Each sampling time between t The impulse response of the sound field target.
[0136]
[0137] Below are examples of strategies suitable for the pressure zone state, and then examples of strategies suitable for the propagation zone state.
[0138] Determine the target value in the pressure zone. The effective range of these target values is near the first transition frequency between zero frequency and the frequency of the first longitudinal mode in room 20. Between these frequencies, starting from the first transition frequency, the sound field is no longer constant at every point in the room.
[0139] These target values It is synthesized into a uniform, constant sound field corresponding to the entire room:
[0140]
[0141] in, It is the amplitude of the desired or target sound field at the positioning point p of the measuring sensor, and It is a test signal Reference launch time.
[0142] It should be noted that although the target is constant, the filter obtained through the MIMO process is not an identity filter in the case of a non-ideal room and loudspeaker.
[0143] These targets can be advantageously filtered using bandpass or lowpass filters so that their spectrum matches the bandwidth of the loudspeaker 12.
[0144] Alternatively, the target for the pressure zone state is synthesized into a non-uniform sound field corresponding to the entire room:
[0145]
[0146] in, It is the amplitude of the desired or target sound field at the positioning point p of the measuring sensor, and It is a test signal Reference launch time.
[0147] "Sound field amplitude" refers to the amplitude of the sound wave at that point, rather than simply the power measurement at that point.
[0148] Alternatively, the target for the pressure zone state can be determined based on measurements of the system response.
[0149] Therefore, in a possible first implementation, the point The target is defined as a single point (called the reference point) emitted by each loudspeaker and within the listening area 19. The sum of the sound waves measured at the location:
[0150]
[0151] When the actual sound field in the listening area is very uniform below the transition frequency... The first implementation remains effective. In this case, the target has little dependence on the measurement point, and the measurement at a single point in the listening area is sufficient to represent the sound field of the entire listening area.
[0152] In a second possible implementation, the point The target at a given point is unique for all measurement points, but this time it is obtained as the average response measured at at least two distinct measurement points:
[0153]
[0154] in, It includes A subset of measurement points.
[0155] This spatial average allows for the definition of more robust targets to changes in the sound field that could occur in real-world situations due to large obstructions in the room or due to the use of amplifiers that differ significantly in sensitivity and / or frequency response.
[0156] In a third possible implementation, point The target at each measurement point They are different:
[0157]
[0158] These objectives allow for the consideration of larger sound field variations compared to a theoretically constant sound field.
[0159] Regardless of the implementation method chosen, the target can be... Additional processing, such as time windowing or time-domain or frequency-domain smoothing, can be applied to reduce the filtering workload and thus improve the efficiency of the inversion algorithm.
[0160] In the propagation zone state, in the first embodiment, the target It is configured to synthesize a plane wave propagating from the room wall selected as the front wall (and with the loudspeaker arranged on the wall as the transmitter and the other loudspeakers of system 11 as non-transmitters).
[0161] The amplitude of the plane wave is constant across the cross-section of the room for each measurement point. ,point Target response Defined as:
[0162]
[0163] ,in It is the Dirac function. Measurement point The amplitude of the wave at that point, and It depends on the delay relative to the measurement point of the transmitting loudspeaker.
[0164] This delay can be defined as: ,in, It is the speed of sound in the air of the room. This is the distance between the front wall and the measurement point. A reference time must be selected. This ensures that for all loudspeakers used as transmitters during the calibration phase, the target field obeys the causality condition ( .
[0165] amplitude You can choose to target all points. Is it a constant value (in the case of a lossless room), or based on a point? The location changes (in the case of a damaged room).
[0166] As mentioned earlier, the target can be bandpass-filtered or lowpass-filtered so that its spectrum matches the bandwidth of the transmitting loudspeaker.
[0167] Since the target consists only of the transverse planar patterns of the room, any contributions from higher-order modes are eliminated. This has the effect of minimizing spatial variations in the sound field.
[0168] The effective range of these targets is between the cutoff frequency. Second transition frequency Between. The latter is equal to the spatial aliasing frequency determined by the spacing between the measurement points. Typically, this is selected by arranging microphones 16. This ensures that the wavelength associated with that frequency is the distance between the measurement points. At least twice the spacing. This allows for the avoidance of ambiguity in sound field measurements due to spatial aliasing. Set as equal to One-third of the associated wavelength, then the second transition frequency is determined by Given this frequency, it must be high enough to control as many modes as possible, but low enough to properly sample the sound field without requiring too many measurement points. A good compromise is to take... = 1 m, which gives Hz.
[0169] The effective range of the target is also related to the spacing between the loudspeakers. For a two-dimensional loudspeaker array, the aliasing frequency is determined by... It is confirmed that, among them, and They are directions and The number of loudspeakers on the screen, and among them, and The rooms are in different directions. and The dimensions on top.
[0170] In the second embodiment, at the midpoint of the propagation region state The goal of the place It is determined based on the measured values of the system response:
[0171]
[0172] in, It is a subset of the loudspeakers, and It is a time window that only allows consideration of contributions made before the first wave.
[0173] These objectives ensure good uniformity of the sound field while implicitly adhering to the physical constraints of sound propagation in a room, such as the arrival time and amplitude attenuation of waves during their propagation.
[0174] - In step 117, different units 118 (b) They are executed independently of each other, and advantageously in parallel, in order to determine the frequency band for each band. Filter .
[0175] filter vector Include One optimized filter
[0176]
[0177] Each component Or the loudspeaker filter is a duration of tuples, Indicates from 1 to Each sampling time t between the amplifier The impulse response of the associated filter:
[0178]
[0179] The MIMO control method according to the present invention includes consideration of each frequency band independently. Find a filter that satisfies the following equation :
[0180]
[0181] Filter vector It is the result of optimization.
[0182] For example, filters It uses the least squares method to minimize the cost function Calculated:
[0183]
[0184] Among them, the function (r is 1 to Integers between [a certain number] The number of regularization functions (which depends on one or more regularization parameters) is determined by the number of regularization functions. The regularization function, and where, It is the regularization coefficient.
[0185] For the case where there is no regularization function ( ),filter The definition is as follows:
[0186]
[0187] In general, the following relationship can be written:
[0188]
[0189] in, It is the inverse function, when the cost function When it contains one or more regularization terms, the inverse function minimizes the cost function.
[0190] Finally, in step 119, execution unit 120 combines filters optimized for each frequency band. Determine the global optimization filter .
[0191] In a simplified embodiment, this combination is achieved through frequency transitions between adjacent optimized frequency bands. It is done by overlapping (“crossing”) in nearby frequency ranges.
[0192] The type of overlap can be freely chosen based on the desired response on both sides of the transition frequency. Adjusting the slope of the filter within the overlap interval yields a smoother transition between adjacent frequency bands.
[0193] In the usage phase 103 following the calibration phase 101, the execution unit 104 applies a global optimization filter. For signal Perform filtering. For each loudspeaker in system 11. Obtain the filtered signal .
[0194] Second Embodiment
[0195] exist Figure 3 In the second embodiment shown, the controller 18 is programmed to include a control module 202 and a filter module 104.
[0196] Control module 202 includes:
[0197] - Unit 112 for defining the frequency band within the operating frequency range of the sound playback system 11;
[0198] - Used to obtain the response matrix Unit 114;
[0199] - Used for each frequency band Define the goal Unit 116;
[0200] - Used for aggregation targets To define global goals Unit 220.
[0201] - Used for matrix and global goals To calculate the global filter Unit 218.
[0202] The filtering module 104 of computer 18 is equipped with a global filter after the calibration stage. Furthermore, during the usage phase, the filtering module is adapted to apply these global filters to the raw signal received from source 9. In order to obtain multiple filtered signals for driving each loudspeaker 12 .
[0203] The method implemented by a computer is as follows:
[0204] In calibration phase 201, the number and arrangement of loudspeaker sets 12 and the number and location of sensor sets 16 are determined.
[0205] Then, module 202 is executed such that:
[0206] - In step 211, which is the same as step 111, execution unit 112 defines the frequency band.
[0207] - In step 213, which is the same as step 113, execution unit 114 obtains the response matrix of sound playback system 11 and room 20. .
[0208] - In step 215, which is the same as step 115, execution unit 116 performs execution for each frequency band. Define goals Furthermore, it advantageously defines a subset of the transmitters used to create the target.
[0209] - In step 216, execution unit 220 will target each frequency band b. Aggregates into a global target for the entire working range of the sound playback system 11. .
[0210] - In step 217, execution unit 218 performs the operation according to the global objective. sum matrix Determine the global filter .
[0211] In step 217, the global filter is optimized by solving the following equation:
[0212]
[0213] Generally, the following relationship can be written:
[0214]
[0215] In the usage phase 203 following the calibration phase 201, unit 104 is configured with a global filter. And it is executed by applying a global optimization filter. To the signal Filtering is performed. Filtered signals are obtained for each speaker 12 in system 11. .
[0216] Third Embodiment
[0217] exist Figure 4 In the third embodiment shown, the controller 18 is programmed to include a control module 302 and a filter module 104.
[0218] Control module 302 includes:
[0219] - Unit 112 is used to define the frequency band within the frequency range of the sound playback system;
[0220] - Used to obtain the response matrix Unit 114;
[0221] - Used for each frequency band Define the goal Unit 116;
[0222] - Used for each frequency band Determine the response matrix Unit 322;
[0223] - For matrix-based and target To calculate the filter for each frequency band b Multiple units 318 b ;as well as,
[0224] - Used for filtering Aggregation into a global filter Unit 120.
[0225] The filter module 104 of processor 18 is configured with a filter after the calibration stage. Furthermore, during the usage phase, the filtering module is adapted to apply the global filter to the raw signal received from source 9. In order to obtain a filtered signal for driving each speaker. .
[0226] The process implemented by the processor is as follows:
[0227] In calibration phase 301, the number and arrangement of speaker sets 12 and the number and location of sensor sets 16 are determined.
[0228] Then, module 302 is executed such that:
[0229] - In step 311, which is the same as step 111, execution unit 112 defines the frequency band.
[0230] - In step 313, which is the same as step 113, execution unit 114 obtains the response matrix of sound playback system 11 and room 20. .
[0231] - In step 315, which is the same as step 115, execution unit 116 performs execution for each frequency band. Define goals And advantageously for each frequency band Define a subset of loudspeaker emitters.
[0232] - In step 316, execution unit 322 adjusts according to the transition frequency. Lists and matrices Determine each frequency band response matrix For example, by applying bandpass, bandstop, or high-pass filter banks, the cutoff frequency of these filter banks is determined by the transition frequency. definition.
[0233] - In step 317, each unit 218 is executed. b According to the frequency band goal sum matrix Determine the frequency band Associated filters .
[0234] In step 317, the filter is optimized by solving the following equation. :
[0235]
[0236] Generally, the following relationship can be written:
[0237]
[0238] Finally, in step 319, the execution unit 120 combines each frequency band. Optimized filter on Determine the global filter .
[0239] In the usage phase 303 following the calibration phase 301, the execution unit 104 applies a global filter. For signal Filtering is performed for each speaker in system 11. Obtain the filtered signal .
[0240] Fourth embodiment
[0241] In the fourth embodiment, the target It is determined by continuously changing the optimization parameters according to the frequency, rather than using specific optimization parameters for each frequency band and therefore being constant in each frequency band.
[0242] Figure 5 The fourth embodiment is shown in the figure.
[0243] The controller 18 is programmed to include a control module 402 and a filter module 104.
[0244] Control module 402 includes:
[0245] - Unit 112 is used to define the frequency band within the frequency range of the sound playback system;
[0246] - Matrix response acquisition unit 114 ;
[0247] - Global Target Determination Unit 416 This includes the response matrix H used to determine the transmitter. E Block 452 is used to calculate the transformed response matrix H of the transmitter. ET Block 454, and the blocks used to synthesize each measurement point. The target block 456; and,
[0248] - Global filter calculation unit 218, used to calculate based on matrix and global goals To calculate the global filter .
[0249] The filter module 104 of processor 18 is configured with a filter after the calibration stage. Furthermore, during the operation phase, the filtering module is adapted to apply these filters to the raw signal received from source 9. In order to obtain a filtered signal for driving each speaker. .
[0250] The process implemented by the processor is as follows:
[0251] In calibration phase 401, the number and arrangement of speaker sets 12 and the number and location of sensor sets 16 are determined.
[0252] Then, module 402 is executed such that:
[0253] - In step 411, which is the same as step 111, execution unit 112 defines the frequency band.
[0254] - In step 413, which is the same as step 113, execution unit 114 obtains the response matrix of sound playback system 11 and room 20. .
[0255] - In step 415, execution unit 416 defines a global target. .
[0256] More specifically, in substep 451, block 452 is executed to define the transmitter for each frequency band. The filter is then determined. The filter represents an index. The contribution of the loudspeaker to target creation varies with frequency.
[0257] filter Applied to matrix H Each element .
[0258] For example, in the case of the second frequency band corresponding to the propagation zone state, only the loudspeaker on the front wall is the transmitter. In the middle of this second frequency band, there is a filter for the transmitter. It equals 1, and the filter for non-emitter speakers equals 0.
[0259] In the first frequency band corresponding to the pressure zone state, all loudspeakers are transmitters. In the middle of this first frequency band, a filter C is applied to all loudspeakers. s ( f ) equals 1.
[0260] filter The intermediate values are obtained by interpolating these fixed values based on the frequency. This applies to filters for non-emitter loudspeakers in propagation mode. The value of gradually changes from 0 to 1 as the frequency decreases, in order to switch from one state to another.
[0261] Then use a filter For matrix Perform filtering to obtain the filtered matrix. It is called the transmitter response matrix. For each measurement point... and each speaker We have:
[0262] = .
[0263] In substep 453, block 454 is executed to transform the transmitter speaker response by applying a time window.
[0264] Initially, for each frequency band Define specific time windows. Then, in the second step, interpolate these time windows according to the frequency to define the time-frequency transform. This time-frequency transformation is applied to the matrix Each element Thus, the transformed response matrix is obtained for the transmitter loudspeaker. For each measurement point and each speaker We have:
[0265] = ( )
[0266] The transformation selects the portion of the measured response used to construct the target. A narrow time window only allows consideration of the first wavefront emitted by the speaker, while a wider time window also allows consideration of reflections from the room walls.
[0267] For example, in the case of a second optimized frequency band specific to the propagation zone state, only the first wavefront is selected, excluding the room. In the middle of this second frequency band, a transformation is performed. Windowing is applied for relatively short time intervals.
[0268] In the case of the first optimization frequency band specific to the pressure zone state, the entire response must be considered to maximize the pressure. In the middle of this first frequency band, the transformation... This corresponds to a very wide time window.
[0269] The length of the time window interval is interpolated between these fixed values based on the frequency. This is equivalent to performing a window whose width depends on the frequency.
[0270] In substep 455, block 456 is executed to determine the target. 。
[0271] In the simplest embodiment, point The target at that location is point Combination of transformed responses at:
[0272]
[0273] In a more advanced implementation, the measurement points The goal is to use time-frequency weighting Applied to each loudspeaker To achieve
[0274]
[0275] This more advanced mode allows for compensation of transmitter loudspeaker positioning errors and / or optimization of wavefront synthesis in the propagation zone state.
[0276] Finally, in step 417, execution unit 418 determines the global filter. .
[0277]
[0278] During operation phase 403 following calibration phase 401, execution unit 104 applies a global optimization filter. For signal Filtering is performed for each speaker in system 11. Obtain the filtered signal .
[0279] In this fourth embodiment, a time / frequency window is used to define an optimization strategy that varies continuously with frequency. This offers advantages in terms of flexibility compared to other embodiments.
[0280] Example 1
[0281] Combination Figure 6 An application example of the method according to the present invention is presented.
[0282] Size is The parallelepiped-shaped component is equipped with There are 12 loudspeakers, including 4 transmitting loudspeakers on the front wall and 4 non-transmitting loudspeakers on the rear wall.
[0283] The loudspeakers are regularly placed on the front and back walls so that below a certain frequency, only planar modes are not excited in the room.
[0284] Two transition frequencies are defined, which define two frequency bands:
[0285] - In the first frequency band, the optimization strategy involves maximizing the response in the lowest part of the spectrum to obtain a richer and stronger response (stress mode), and it is applied between zero frequency and frequencies close to the cutoff frequency, typically, or .frequency It can be estimated in advance based on the length of the room. .therefore, It is a frequency at which the sound field is constant at all points in the listening area when the frequency is below that.
[0286] - In the second frequency band: the sound field is strongly influenced by room resonance modes. Optimization strategies involve minimizing the spatial and frequency deviations in the response caused by the excitation of these modes. The target field is then defined as the sum of the responses of the transmitting amplifiers, time-windowed to consider only the first wavefront. This operating mode is applied between the first transition frequency and a second transition frequency related to the spacing between the microphone or amplifier. Typically, .
[0287] The sound field at the maximum spacing is of Measurements are taken at the set of points. The responses between all points and all loudspeakers are measured and assembled into a matrix. middle.
[0288] Then, two target values are defined for each frequency band:
[0289] - For the first frequency band: Since the goal is to maximize the response in this frequency band, the target response is defined as the sum of the responses of all loudspeakers.
[0290] - For the second frequency band: the target response is obtained as the sum of the initial responses from only the transmitting end. Since the objective is to eliminate the contributions of higher-order modes, a time window is applied to the initial response so that only the contribution of the first transmitted wavefront is considered.
[0291] Calculate an optimized filter for each frequency band based on the target response.
[0292] Combine these filters to obtain the overall filter.
[0293] Apply these filters to each loudspeaker to obtain each point. Frequency-filtered response at the point The calculation formula is as follows:
[0294]
[0295] in, and It is a filter and response Fourier transform.
[0296] Figure 6 The response for all measurement points is shown. The amplitude (in decibels (dB)) varies with frequency (Hz).
[0297] Graph A shows the processed response when the stress-mode strategy is applied across the entire frequency range. It is observed that below the frequency range... At that time, the desired behavior was obtained, namely, the response was amplified while maintaining good spatial homogeneity. However, for values higher than... The frequency was observed to exhibit strong spatial variations, with a strong peak in the response near 53 Hz, which is attributed to the excitation of one of the room's resonant modes.
[0298] Curve graph The processed response is shown when the propagation mode optimization strategy is applied across the entire frequency range. This strategy allows for... and The room behavior was controlled, particularly the resonant peak at 53 Hz was completely eliminated, and a flat and relatively uniform response was obtained for all measurement points. However, below... No amplification was observed at that frequency.
[0299] Graph C shows the processed response obtained by combining the two previously described optimization strategies using this MIMO control method. Below... At that frequency, a significant amplification of the response was observed, similar to the behavior obtained in curve A, while to Excellent modal control is achieved at frequencies between [specific frequencies].
[0300] It is noteworthy that the spatial variation produced using the method according to the invention is less than that obtained by using a single optimization strategy (specifically a propagation mode strategy) across the entire frequency range. This behavior is attributed to the fact that propagation modes are applied only to frequencies. and The second frequency band between these frequencies reduces the filtering workload and improves the performance of the inversion algorithm in this band.
[0301] Example 2
[0302] Combination Figure 7 A second example is presented, which divides the frequency range into four frequency bands.
[0303] Consider a room 20 in the shape of a parallelepiped, the front wall (plane yz) of which is equipped with a two-dimensional arrangement of 14 loudspeakers 12.
[0304] The purpose of this example is to demonstrate how the proposed method can be used to extend acoustic control to higher frequencies by using the loudspeaker 12 of system 11 more intelligently and effectively.
[0305] The desired target response in each of the four frequency bands within the subdivided frequency range is shown in each of the diagrams A through D.
[0306] In each diagram, the black circles represent a set of loudspeakers selected as the transmitting loudspeakers for the frequency band in question, thus contributing to the frequency band in question. Create a target The white circles represent non-emitting loudspeakers, which are not used in the frequency band. Create target response .
[0307] Figure 7 Only loudspeakers that may have different effects across different optimized frequency bands are shown. However, examples of this embodiment may also include other loudspeakers with purely non-emitting functions. These loudspeakers may be placed, for example, on a rear wall, as described in the previous examples. For simplicity, these non-emitting loudspeakers are not shown... Figure 7 As shown in the image.
[0308] The optimization strategy for the first two frequency bands is similar to the previous example.
[0309] In the first frequency band (Figure A), the target frequency is the frequency of the first mode, which is close to the room's frequency. The following achieves a uniform response with high amplitude. For this purpose, all 14 loudspeakers 12 in transmit mode are used simultaneously in the first frequency band. A smaller subset of loudspeakers can also be used depending on the desired amplitude for the target.
[0310] In the second frequency band, a second subset of transmitting loudspeakers 12 is used. The transmitting loudspeakers are also spaced apart to generate plane waves (Figure B). Unlike the previous case, we now consider a sufficiently fine grid of measurement points such that the maximum frequency in this second frequency band is now determined by… and The spacing is given in the middle. In fact, the positioning of the loudspeaker allows it to operate at a certain frequency. The following generates a plane wave, and beyond this frequency, the loudspeaker will excite higher-order modes. After that, it will no longer be possible to obtain a plane wave. This frequency is given by the following equation: , Number of loudspeakers in the direction Equals four. Number of loudspeakers in the direction It equals two, and and The front wall is direction and Dimensions in the direction.
[0311] If the same optimization strategy is expected to achieve good acoustic control, then it is necessary to... and Increasing the density of the loudspeaker network will result in an excessive number of loudspeakers.
[0312] To overcome this limitation, optimization strategies in the third frequency band involve using only rows of transmitting loudspeakers. Advantageously, these rows of transmitting loudspeakers are located near the edge of the front wall, which allows the room to be used as a waveguide to generate an interference-free cylindrical wavefront with a constant amplitude throughout the listening area (Figure C). The frequency constraints of this strategy are given by the following equation: ,in, It represents the number of loudspeakers in a column (oriented along the z-direction). Therefore, acoustic control of a room at higher frequencies can be achieved without using a very large number of loudspeakers.
[0313] To achieve acoustic control at even higher frequencies, the optimization strategy in the fourth frequency band involves using a single transmitting loudspeaker. Advantageously, this single transmitting loudspeaker is located in a corner of the room (Figure D). This positioning of the transmitting loudspeaker allows the room to be used as a waveguide again, so that this time spherical waves are generated that do not reflect off the walls.
[0314] Combining various optimization strategies to obtain optimized filters The precise definition and processing of the target can be performed according to one or more of the above embodiments.
[0315] For example, the previously proposed formula can be used based on the measured values. Obtain the target response of cylindrical or spherical waves:
[0316]
[0317] For cylindrical waves And for spherical waves , It is a subset of amplifiers used to create these waves, and It is a time window that only allows consideration of the contribution of the direct sound field before the reflection caused by the relative wall arrives.
[0318] These objectives can also be defined by synthesizing analytical solutions from two-dimensional or three-dimensional acoustic wave equations.
[0319] For example, for 2D cylindrical waves:
[0320]
[0321] For example, for 3D spherical waves:
[0322]
[0323] In these equations, It is a reference range, and It is a loudspeaker. The distance between point p and point p.
[0324] Alternative embodiments
[0325] In an alternative embodiment, the transition frequency It is not user-defined, but based on the room size. and It is determined automatically, especially when using the formulas mentioned above.
[0326] The automatic determination of the response can also be based on calculating the performance of each optimization strategy as a function of frequency, so as to empirically determine the effective range of each strategy under specific conditions.
[0327] It can also be based on iterative computation, where the performance of each optimization strategy obtained at the end of an iteration allows for adjustment of the transition frequency for the next iteration. The value of the transition frequency is determined, and the iteration continues until the performance criteria are validated. The initialization of the transition frequency value at the start of this iteration process can, for example, be based on the room size and the number of loudspeakers used.
[0328] If, in the above embodiment, calculator 18 is programmed to perform both a calibration phase and a usage phase, then in a variant, two different calculators can be used, each dedicated to performing a specific phase. At the end of the calibration phase, the filter calculated by the calibration calculator of the control system is downloaded to the resident calculator of the sound playback system for use in the usage phase. It should be noted that during the calibration phase, the control system must be connected to the sound playback system in order to apply test signals to the amplifier of the sound playback system. In this case, the control system consists, for example, a microphone deployed at the site to be controlled and a laptop computer connected to the sound playback system.
[0329] The software implementation of the present invention has been presented in the above embodiments. In variants, all or part of the modules, units and / or blocks may be implemented in hardware form (e.g., in the form of an FPGA).
[0330] advantage
[0331] Although MIMO control methods have proven their effectiveness in controlling sound fields, this invention allows for the definition of different optimization strategies based on frequency.
[0332] Compared to existing technologies that only provide a single control strategy, this invention allows for the relaxation of constraints in the computation of optimization filters and avoids certain audible artifacts.
[0333] Furthermore, the use of various optimization strategies enables full and effective utilization of all amplifiers in the sound reproduction system.
Claims
1. A method for MIMO control of a sound field in a listening area (19) within a room (20), the sound field being generated by a sound playback system (11) comprising a plurality of loudspeakers (12) and a first computer, the method being implemented by a control system (10) connected to the plurality of loudspeakers and comprising a second computer (18) and a plurality of microphones (16) positioned in the listening area, the method being characterized in that the method includes the following steps in a calibration phase (101) implemented by the second computer: - The room (20) defines a cutoff frequency between the pressure zone state and the propagation zone state, thereby defining (111) at least one first frequency band below the cutoff frequency and a second frequency band above the cutoff frequency within the operating frequency range of the sound playback system; - By selecting a first objective to be achieved, define a first optimization strategy for the filter in the first frequency band that is specifically designed for the state of the pressure zone; - By selecting the second objective to be achieved, a second optimization strategy is defined for the filter in the second frequency band, specifically for the propagation region state; - The response matrix of the loudspeaker and the room is obtained by sequentially applying a test signal to each loudspeaker and measuring the basic response of the activated loudspeaker at each location point of the microphone (113). );as well as, - Based on the response matrix, the first strategy, and the second strategy, a global filter is determined by optimizing the sound field in the listening area relative to the first target and the second target.
2. The method for MIMO control according to claim 1, wherein, Determine the global filter ( The initial step involved optimization based on the response matrix ( The first filter for the first frequency band is determined by the first target and the first target. ), and according to the response matrix ( The second target determines the second filter for the second frequency band. Then, the first filter and the second filter are combined to obtain the global filter.
3. The method for MIMO control according to claim 1, wherein, Determine the global filter ( The initial goal includes combining (216) the first objective and the second objective to obtain the global objective. ), and then through optimization, based on the response matrix ( ) and the global target ( (217) is used to determine the global filter.
4. The method for MIMO control according to claim 1, wherein, Determine the global filter ( )include: - The response matrix is applied to the first frequency band and the second frequency band respectively. ) Perform filtering (316) to obtain the first response matrix respectively ( ), second response matrix ( ); - Through optimization, a first filter (117) for the first frequency band is determined based on the first response matrix and the first target. And based on the second response matrix and the second target, determine the second filter for the second frequency band. );Then, - Combine (319) the first filter and the second filter to obtain the global filter.
5. The method for MIMO control according to claim 1, wherein, The global target, which varies continuously with frequency, is determined by interpolating the first target and the second target. ), and then through optimization, based on the response matrix ( The global filter is determined by the global target and the global target. ).
6. The control method according to any one of claims 2 to 5, wherein, Synthetic microphone positioning point The first target at the location : - In the first form: in, It is the amplitude of the sound field, and It is the reference transmission time of the test signal; - In the second form: in, It is the amplitude of the sound field, and It is the reference transmission time of the test signal; - In the third form: in, At a single reference point The response of the s-th loudspeaker measured at point ; - In the fourth form: in, It includes A subset of the measurement points of each point, and It is at point The response of the s-th loudspeaker measured at point ; or - In the fifth form: .
7. The method for control according to any one of claims 2 to 6, wherein, Synthetic microphone positioning point The second target at the location : - In the first form: ,in It is the Dirac function. Measurement point The amplitude of the wave at that point, and It depends on the measured position relative to one or more speakers emitting the test signal; - In the second form: ,in, It is applied to the measured response The time window, and It is a subset of the speakers involved in creating the target. - In the third form: ;or - In the fourth form: ,in, Is the index as Speaker and measuring point The distance between them.
8. A control system (11), characterized in that The control system is adapted to implement the MIMO control method according to any one of the preceding claims when associated with a sound playback system.
9. The control system according to claim 8, wherein, The control system includes a first computer and multiple microphones, and the sound playback system includes a second computer and multiple speakers. The first computer and the second computer are either the same computer that is properly programmed or two different but connected computers for implementing the MIMO control method.
10. A computer program (36) comprising software instructions that, when executed by a computer of a control system associated with a sound playback system according to claim 8 or claim 9, implement the MIMO control method according to any one of claims 1 to 7.
Citation Information
Patent Citations
Nouveaux composes azoiques, leur preparation et leur utilisation comme colorants
FR2300786A1