JP4625756B2 - Loudspeaker array system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a loudspeaker including a plurality of transducers of at least two different sizes wherein the transducers are symmetrically arranged with respect to a first axis and with respect to a second axis perpendicular to the first axis, and the centers of some of the transducers are located along the loudspeaker at positions not on the first and second axes. <P>SOLUTION: The loudspeaker includes a plurality of transducers of at least two different sizes wherein the transducers are symmetrically arranged with respect to a first axis and with respect to a second axis perpendicular to the first axis, and the loudspeaker includes some of the transducers which are located along the loudspeaker at positions not on the first and second axes. <P>COPYRIGHT: (C)2007,JPO&amp;INPIT

Description

(Technical field)
The present invention generally relates to multi-way loudspeaker systems. Specifically, it consists of loudspeaker drivers arranged symmetrically in a two-dimensional plane, achieving high quality sound when used in combination with stereo loudspeaker systems, multichannel home entertainment systems, and public information broadcasting systems A multi-way loudspeaker system.

  Loudspeaker designers have a wide range of frequency bands, despite constraints on the size and number of transducers (eg, drivers) as well as the number of amplifiers (eg, ways) required in the system. Efforts are constantly being made to design loudspeaker systems with controlled directivity that achieve high quality sound over time. Achieving such high quality sound over a wide frequency range is due to the changing dimensions of the transducers for each dedicated portion of the audible frequency band and the constraints on the spacing between the transducers. Therefore, it has been regarded as a difficult task.

  High quality loudspeakers for the audible frequency range typically employ multiple drivers, dedicated to each dedicated portion of the audible frequency band, which include a tweeter (typically 2 kHz to 20 kHz), a midrange driver (typically 200 Hz to 5 kHz). ), And a woofer (generally 20 Hz to 1 kHz). Generally, higher frequency drivers are smaller in size than lower frequency drivers.

  In order to achieve high sound quality, it is desirable to place the drivers as close together as possible in the loudspeaker. However, due to the physical dimensions of dedicated drivers, the possibilities for placing drivers in close proximity to each other are limited. The further away the drivers are from each other, the more acoustic problems arise.

  Because of the driver's own physical dimensions due to the driver's own physical dimensions, the driver's acoustic output is transmitted to a loudspeaker, usually called an acoustic center, because the distance is comparable to the wavelength of the emitted sound. Add only on one vertical line to obtain a flat response independent of the intended frequency. At points off the axis, the frequency response is more or less distorted due to interference due to different acoustic path lengths from each driver to the point of space consideration. Thus, there have historically been many attempts to construct a loudspeaker having a controlled sound space so that a smooth response is obtained even in off-axis positions in large spaces.

  The current technology for controlling sound space in large spaces, such as public spaces, is to use horns to enhance the sound evenly arranged. However, the use of uniformly arranged horns has drawbacks. This is because uniformly arranged horns have a constrained frequency range, a fixed, non-redirectable polar pattern, and a relatively high distortion.

  A current two-dimensional array for home entertainment surround sound, the so-called sound projector, is an array of identical, small, wide-band drivers that are linearly spaced. This type of array can produce multiple sound beams, which radiate into the room and produce the desired sound effect during the bounce from the wall to the listener. However, the drivers in a two-dimensional, linearly spaced array are identical and the maximum sound pressure and sound quality of a sound projector is limited by the transducer's performance, which is generally exclusive. This is inferior to a drive unit optimized for a certain frequency band. In addition, sound projectors use a very large number of drivers that all need to be driven individually, which results in tremendous complexity and high cost.

  In this way, a minimum number of converters are used, as well as amplifiers, which are optimized for high performance by using dedicated drivers that span the audible frequency band, such as tweeters, mid-range drivers or woofers. There remains a need for a quality, low distortion, two dimensional loudspeaker configuration. Unlike fixed devices that use an array of horns, there is still a need for a two-dimensional loudspeaker configuration that electronically changes the beam width and direction angle on-site.

(Summary of the Invention)
(Item 1)
A loudspeaker having a plurality of transducers of at least two different dimensions, the transducers being arranged symmetrically with respect to a first axis and with respect to a second axis perpendicular to the first axis The loudspeaker is
A loudspeaker comprising a transducer whose center is located along the loudspeaker at a position that is not on either the first axis or the second axis.

(Item 2)
The transducer of the loudspeaker, arranged symmetrically with respect to both the first and second axes, converts the filtered digital output signal through at least one digital FIR filter into at least one power. The loudspeaker according to item 1, which is received from the D / A converter.

(Item 3)
The loudspeaker of item 1, wherein the plurality of transducers of the at least two different dimensions are tweeters and mid-range drivers.

(Item 4)
The loudspeaker of item 1, wherein the plurality of transducers of the at least two different dimensions are a tweeter and a woofer.

(Item 5)
The loudspeaker of item 1, wherein the plurality of transducers of the at least two different dimensions are midrange drivers and tweeters.

(Item 6)
The loudspeaker according to item 1, wherein the plurality of drivers include a tweeter, a mid-range driver, and a woofer.

(Item 7)
The loudspeaker according to item 1, wherein one of the plurality of transducers is centered at an intersection of the first and second axes.

(Item 8)
Item 8. The loudspeaker of item 7, wherein the central converter receives a digital output signal filtered through at least one digital FIR filter from at least one power D / A converter.

(Item 9)
Linear phase filter coefficients for each FIR filter determine an initial driver position, determine an initial directivity objective function for the system, and apply a cost minimization function based on the initial directivity objective function The loudspeaker of claim 2, wherein the loudspeaker is determined by: calculating a linear phase filter coefficient for each filter of the system.

(Item 10)
Item 10. The loudspeaker according to item 9, wherein the initial driver position is a coordinate related to an origin that is the center of the loudspeaker.

(Item 11)
Item 10. The loudspeaker according to item 9, wherein a frequency point is determined on a logarithmic scale of a predetermined frequency range based on the determined initial directivity objective function.

(Item 12)
Item 10. The loudspeaker of item 9, wherein the cost minimization is a function that starts at a minimum frequency and increases stepwise and is applied at the frequency point.

(Item 13)
Item 10. The loudspeaker according to item 9, wherein a Fourier approximation method is used to establish the linear phase filter coefficients.

(Item 14)
To change the directivity of the loudspeaker array along the axis perpendicular to the direction in which the array is arranged, while maintaining the directivity of the array along the axis in which the array is arranged The method comprising:
At least one transducer having a center on the axis on which the linear array is arranged, the transducer being substantially identical to the transducer having a center on the axis on which the array is arranged. Including replacing by at least one pair;
When the pair is replaced, one transducer of the pair is placed on one side of the axis where the array is arranged, and another transducer of the pair is placed on the axis where the array is arranged. On the other side of the array so that the array is arranged at the same distance from the axis where the array is arranged, and the opposing pairs are arranged at the same distance from the axis perpendicular to the axis where the array is arranged To be replaced.

(Item 15)
15. The loudspeaker of item 14, wherein the pair of converters receives a digital output signal filtered through at least one digital FIR filter from at least one power D / A converter.

(Item 16)
To change the directivity of the loudspeaker array along the axis perpendicular to the direction in which the array is arranged, while maintaining the directivity of the array along the axis in which the array is arranged The method comprising:
Replacing at least one transducer centered on the axis in which the linear array is arranged with at least one pair of transducers substantially identical to the transducer;
The pair of transducers are arranged at mirror points with respect to the axis at points along the axis in which the array is arranged, so that the directivity of the axis in which the array is arranged Is maintained, the way.

(Item 17)
A loudspeaker system comprising at least five transducers of at least two different dimensions,
The at least five transducers are arranged symmetrically with respect to both a first axis and a second axis perpendicular to the first axis, and at least one substantially identical transducer pair Are centered parallel to each other with respect to the first axis, one transducer of the pair is placed on one side of the second axis, and the other transducer in the pair is A loudspeaker system placed on the opposite side of the second axis so that the distance from the axis is the same distance as the other transducers of the pair.

(wrap up)
The present invention is a multi-way array loudspeaker capable of producing high quality sound in a high fidelity (high fidelity) stereo system, a multi-channel home entertainment system or a public information broadcasting system.

  In one embodiment, the array includes a plurality of tweeters, midrange drivers and woofers arranged in a single housing or assembled into a single unit, the housing or unit being a driver coupling. In order to avoid this, it has a sealed compartment separating several drivers from each other. The array can be single channel with various signal paths from the input to individual loudspeaker drivers or to multiple drivers. Each signal path comprises a digital input and includes a digital FIR filter, a D / A converter and a power amplifier connected to a single driver or multiple drivers, or a so-called power D / A converter.

  The performance, position and placement of the loudspeakers in the array can be determined by a filter design algorithm that establishes coefficients for each FIR filter in each signal flow path of the loudspeaker. A cost minimizing function is applied to a given frequency point using the initial driver position and the initial directivity objective function defined at a frequency point on a logarithmic scale within that frequency range. If the results obtained by applying the cost minimizing function do not meet the system performance requirements, the driver position can be modified and the cost minimizing function can be reinstated until the resulting results meet the system requirements. Can be applied. When the results obtained meet the system requirements, the filter coefficients for each of the linear phase FIR filters in one path are calculated using Fourier approximation or other frequency sampling methods.

  The multi-way loudspeaker of the present invention may include built-in DSP processing, D / A converters and amplifiers, and may be connected to a digital network (eg, IEEE 1394 standard). Furthermore, the multi-way loudspeaker system of the present invention can also be designed as a wall-mounted surround system because of its affordable dimensions.

  Multi-way loudspeaker systems can employ drivers with different dimensions, less distortion produced, high power handling, because dedicated drivers are different from their wide band identical driver arrays, This is because it can operate optimally in the frequency band. The multi-way speaker design of the present invention also provides better control of room response because it has a smooth response in an out-of-axis position. The system can also control the frequency response of the reflected sound, as well as the total sound power, and suppress floor and ceiling reflections.

  Other systems, methods, features and advantages of the present invention will become apparent to those skilled in the art upon examination of the following drawings and detailed description. It is intended that all these additional systems, methods, features and advantages be included in this description, be within the scope of the invention, and be protected by the accompanying claims.

  The invention can be better understood with reference to the following drawings. The components in the drawings are not necessarily drawn to scale, but are rather emphasized to represent the principles of the invention. Moreover, the same reference numbers in the drawings designate corresponding parts in the different drawings.

  FIG. 1 shows an embodiment of a one-dimensional (1D) multi-way loudspeaker 100 on which the present invention is based, and a block diagram of the signal flow to each loudspeaker driver of the system 100. As shown in FIG. 1, the multi-way loudspeaker 100 is connected to (1) a central tweeter 102 connected to the first power D / A converter 103, and (2) connected to the second power D / A converter 105. Two additional tweeters 104 and 106, (3) two mid-range drivers 108 and 110 connected to the third power D / A converter 107, and (5) a fourth D / A converter 109. Can be designed as a 4-way loudspeaker having two woofers 112 and 114 connected to each other. The connection between each loudspeaker and the amplifier represents a different way in the multi-way loudspeaker.

  In FIG. 1, a driver, also referred to as a transducer, may be mounted in a housing 116 consisting of compartments 120, 122, and 124 separated and sealed by separators 132 and 134 as shown. By installing the driver in a separate and sealed compartment, the coupling between adjacent drivers is minimized. Although various compartments can be seen in FIG. 1, the loudspeaker system can be designed so that the compartments are not visible to the consumer when embodied as a finished product. The compartment 124 containing the woofer 112 can be separated by a separator 132 from the compartment 120 containing the mid-range drivers 108 and 110 and the tweeters 102, 104 and 106. Similarly, the compartment 122 containing the woofer 114 may be separated by a separator 134 from the compartment 120 containing the mid-range drivers 108 and 110 and the tweeters 102, 104 and 106. All the tweeters 102, 104, 106 can be included in the same compartment 120 as the mid-range drivers 108 and 110 because the tweeters 102, 104, and 106 are generally hermetically sealed and remove the tweeters 102, 104, and 106 from the mid-range drivers. This is because there is no need to separate them.

  FIG. 1 shows a central tweeter 102, tweeters 104 and 106, midband drivers 108, 110 and low frequency woofers 112 and 114, which are linear along the y-axis and symmetrical with respect to the central tweeter 102. It has been incorporated. Typical arrangements include tweeters 102, 104 and 106 having an outer diameter of about 40 mm to 50 mm, mid-range drivers 108 and 110 having an outer diameter of about 80 mm to 110 mm, and woofers 112 and 114 having an outer diameter of about 120 mm to 250 mm. May be included. In general, the dimensions of the transducer cone can vary based on the desired intended use and the desired array dimensions. Further, although the transducer may use neodymium magnets, the use of that particular type of magnet is not necessary for the application described.

  When using a 50 mm diameter tweeter, a 110 mm midrange driver and a 160 mm woofer, an example of the system is a central tweeter built into the center point 0 which is the intersection of the x-axis and y-axis on the y-axis. 102 may be included. The tweeters 104 and 106 may be incorporated with their centers approximately +/− 60 mm from the center point. The midrange drivers 110 and 108 can then be incorporated at their centers about 0 / + 150 mm from the center point 0. Low frequency woofers 112 and 114 may then be incorporated at their centers approximately +/− 300 mm from the center point.

  FIG. 1 also shows a signal flow block diagram 140 of a multi-way loudspeaker system. Although FIG. 1 shows four ways of signal flows 142, 144, 146 and 148, the channel may be divided into two or more ways. The signal flow comprises a digital input 150 that can be implemented using a standard connection format, such as SPDIF or IEEE 1394 and their derivatives, via various paths or ways as shown in FIG. Can be connected to the driver. Each path or way 142, 144, 146 and 148 includes a digital FIR filter 152 and a power D / A converter 103, 105, 107 and 109, each connected to one or more loudspeaker drivers. obtain. Power D / A converters 103, 105, 107 and 109 are class D power amplifiers with a cascade of conventional audio D / A converters (not shown) and power amplifiers (not shown) or with direct digital inputs (Not shown). FIR filter 152 may be performed by a digital signal processor (DSP) (not shown). The loudspeaker driver can be a tweeter, midrange driver or woofer as shown.

  In operation, each output of the plurality of FIR filters 152 is connected to a plurality of power D / A converters 103, 105, 107 and 109, and then a plurality of loudspeakers attached to a baffle of the housing 116. Supplied to drivers 102, 104, 106, 108, 110, 112, and 114. More than one driver, such as 104 and 106, may be connected in parallel to a path or way 142, 144, 146 and 148 that includes power D / A converters 103, 105, 107 and 109.

  FIG. 2 shows a two-dimensional multi-way loudspeaker 200 obtained by separating the tweeters 104 and 106 and the mid-range drivers 108 and 110 of FIG. As described further below, the paired drivers can be electrically coupled together and supplied by the same filter as the one-dimensional (1D) multiway loudspeaker 100 of FIG. Therefore, the directivity along the y-axis is not affected, and the same directivity as originally defined is maintained in the long-distance space. However, new directivity characteristics can be introduced along the x-axis as desired.

  Specifically, FIG. 2 shows a one-channel two-dimensional four-way loudspeaker 200 having a central tweeter 202 circled by four additional tweeters 204, 206, 208 and 210. In addition, the loudspeaker 200 includes four midrange drivers 212, 214, 216 and 218, and two woofers 220 and 222.

  The tweeters 204, 206, 208 and 210, the midrange drivers 212, 214, 216 and 218, and the two woofers 220 and 222 are all symmetrically aligned with respect to the central tweeter 202 and linearly aligned along the y-axis. Has been. The pair of tweeters 204 and 206 and the pair of tweeters 208 and 210 are placed on one side of the central tweeter 202, respectively, above and below the centerline defined by the x-axis. Similarly, the pair of midrange drivers 212 and 214 are located above the tweeters 202, 204, 206, 208 and 210, and the other pair of midrange drivers 216 and 218 are tweeters 202, 204, 206, 208 and 210. Is placed symmetrically with respect to the centerline defined by the x-axis.

  Similar to the loudspeaker system 100 of FIG. 1, the loudspeaker system of FIG. 214, 216 and 218, and woofers 220 and 222 having an outer diameter of about 120 mm to 250 mm. As previously mentioned, the dimensions of the transducer cone can vary based on the desired application and the desired array dimensions.

In general, the design of a system with n ways is the optimal position coordinates y _0, +/− (y _1, y _2, y _3, ..., Y _n−1 ) for a particular directivity objective function. , And filter coefficients for the filter FIR (0, 1, 2, 3,..., N−1). In the example where n is given equal to 4, when generating a two-dimensional array, drivers with subscripts (1,..., m), m ≦ n are pairs (here m = 1 and m = 2). ). The corresponding x-axis coordinate is +/− (x ₁ , x ₂ ,..., X _m ), while the y-coordinate remains as it is without changing the one-dimensional design value.

  The y-coordinate in the two-dimensional loudspeaker system 200 can be designed smaller than the physical dimensions of the driver, as shown in FIG. 2, because space is gained by separating the driver and moving in the x-axis direction. Because it is. Thus, an additional degree of freedom is gained from the two-dimensional design, which generally results in further performance improvements.

The directivity along the x-axis can be adapted by optimization of the position parameters x ₁ ,..., x _m and by the value of m itself. The driver with the suffix (m + 1),..., N−1 is not separated and remains in its original position. This means that the array on the x-axis is a truncated version of the original standard array designed on the y-axis. Thus, the directivity function exhibits a higher corner frequency.

The coefficients x ₁ ,..., X _m can be optimized so that a smooth, frequency-independent directivity function is obtained along the x-axis. In the case of x ₁ <y ₁ , x ₂ <y ₂ ,..., the directivity in the x-axis direction of the array is small. In the case of x ₁ = y ₁ , x ₂ = y ₂ ,..., both are equal at high frequencies.

  In the embodiment provided in FIG. 2, the central tweeter 202 can be incorporated at a center point 0 on the y-axis, which is the intersection of the x-axis and the y-axis, as shown in FIG. The tweeters 204, 206, 208 and 210 incorporate their centers at positions about +/− 30 mm along the x-axis and +/− 42 mm along the y-axis (+/− 30 mm, +/− 42 mm). It is.

  The midrange drivers 212, 214, 216 and 218 then center at about +/− 80 mm from the center point 0 along the x-axis and about +/− 120 mm along the y-axis (+/− 80 mm). , +/− 120 mm). The woofers 220 and 222 are then incorporated at their centers at a location approximately +/− 300 mm (+/− 0 mm, +/− 300 mm) from the center point.

  Similar to the loudspeaker system 100 of FIG. 1, the transducer may be incorporated into a housing 230 comprising compartments 232, 234 and 236 separated and sealed by separators 242 and 244 as shown. The compartment 232 containing the woofer 220 may be separated from the compartment 236 containing the mid-range drivers 212, 214, 216 and 218 and the tweeters 202, 204, 206, 208 and 210 by the separator 242. Similarly, compartment 234 containing woofer 222 may be separated by separator 244 from compartment 236 containing midrange drivers 212, 214, 216 and 218 and tweeters 202, 204, 206, 208 and 210.

  FIG. 2 also shows a signal flow block diagram 250 of the multi-way loudspeaker system 200. FIG. 2 shows the four ways 252, 254, 256 and 258 of the signal flow. The signal flow comprises a digital input 264 that can be implemented using a standard connection format and is connected to the driver via various paths or ways, such as the four ways shown in FIG. . Each path or way 252, 254, 256 and 258 may include a digital FIR filter 260 and a power D / A converter 262 connected to one or more loudspeaker drivers.

  FIG. 3 is a flowchart of a filter design algorithm 300 used to design the loudspeaker system of the present invention. The purpose of the filter design algorithm 300 is to determine the coefficients for each FIR filter that correspond to each signal flow path of the loudspeaker. As shown in more detail below, an initial driver position and an initial directivity objective function are first determined (310). Initial speaker and driver locations or design arrangements are designed according to a number of different variables depending on the application, such as desired speaker dimensions, intended use, manufacturing constraints, aesthetic taste or other product design aspects. Can be done. Driver coordinates are then defined for each driver along the principal axis. An initial estimate for the directional objective function is then set, which involves setting a frequency point within the interval of interest on a logarithmic scale. The cost function is then minimized (312) at the defined frequency points. In step 314, if the result does not match the system performance requirements, the driver position is modified and the cost minimization function is again applied (316). This cycle can be repeated until the results meet the requirements. If the result meets the requirement, linear phase filter coefficients are calculated (318). Additional calculations may also be performed 320 to equalize the drivers and to compensate for phase shifts and to allow adjustment of beam orientation.

  In an initial step 310, an initial driver position and an initial directivity objective function are set. As previously mentioned, the number, position, dimensions and orientation of the drivers are primarily determined by the product design aspects. Once the direction has been determined, initial coordinate values can then be defined for the N drivers on the main axis as initial driver coordinates p (n), n = 1,. For example, in the one-dimensional (1D) array shown in FIG. 1, N = 7: p (n) = [− 0.30, −0.15, −0.06, 0, 0.06, 0.15. 0.30] m (meters). In the two-dimensional (2D) array shown in FIG. 2, N = 7: p (n) = [− 0.30, −0.12, −0.042,0, 0.042, 0.12,0 .30] m.

  As depicted in FIG. 2, if the two-dimensional arrangement is symmetric about both the x and y axes, the design process for the two-dimensional arrangement is for one dimension, eg, for the principal axis as described above. Can be implemented. The same directivity characteristic due to symmetry is obtained for the other axes, except for higher corner frequencies.

  In order to determine the initial directivity objective function, an initial estimate for the directivity objective function T (f, q), which is determined based on the desired performance of the driver at a specific angle q, may be defined. is necessary. FIG. 4 shows a set of example objective functions for angle dependent attenuation at five specific angles q. The directional objective function defines the expected sound level attenuation in dB (y-axis), which is at the origin (center tweeter) at a sufficiently large distance (larger than the speaker dimensions) from the speaker in an anechoic environment. Can be measured at various frequencies at angle q from the vertical to. The frequency vector f defines a set of frequency points, for example, 100 on a logarithmic scale within the frequency interval of interest, for example, 100 Hz,.

The angle vectors q (i), i = 1,..., N _q define the set of angles at which optimization is performed. FIG. 4 shows the initial estimate for directivity at five angles,
(N _q = 5): q = [0, 10, 20, 30, 40] °,
In most cases it may be sufficient to define the directivity for only two angles, eg N _q = 2. In this case, the intended directivity can be defined at an outer angle, for example 40 degrees and 0 degrees, where a directivity of zero on the axis is defined. For example, q = [0, 40] °.

  Except for the objective function on the axis, the objective function at each angle is linear on the logarithmic scale from T = 0 dB at f = 0 to a value of T <0 dB at a specific frequency fc (eg, fc = 350 Hz). Decrease, and then keep a constant value. The on-axis objective function 402 maintains a constant value of 0 dB over the entire frequency range. The objective directivity functions for 10 (10) degree 404, 20 (20) degree 410, 30 (30) degree 412 and 40 (40) degree 414 all start at T = 0 dB and are represented by 350 Hz in FIG. The logarithmic scale descends until the function reaches fc, and then remains constant over the frequency range of interest.

  After the initial driver position and initial directivity objective function are determined, the next step 312 minimizes the cost function F (f) at a predetermined frequency vector point f. This minimization is done by starting with the smallest frequency increase step, eg 100HZ, using the resulting solution as the initial solution for the next step, respectively, and using the following equation:

および

and

ここで、Ｈ_ｍ（ｎ，ｆ，ｑ）は、対象のドライバｎ、周波数ｆ、および角度ｑに対する、軸上（角度０）で得られる応答に対して正規化された、測定された振幅周波数応答の組であり、その例は図５に示されている。図５は、ある取り付けられたツイータの、様々な垂直偏差角度において測定され、軸上の値に対して正規化された周波数応答５００を示す。図５において、線５０２は軸上の応答を表し、線５０４は１０度において測定された周波数応答であり、線５０６は２０度における応答であり、線５０８は３０度における応答であり、線５１０は４０度において測定された周波数応答であり、全て１ｋＨｚと２０ｋＨｚとの間の周波数範囲において測定される。

Where H _m (n, f, q) is the measured amplitude frequency normalized to the response obtained on-axis (angle 0) for the target driver n, frequency f and angle q. A set of responses, an example of which is shown in FIG. FIG. 5 shows a frequency response 500 of an attached tweeter measured at various vertical deviation angles and normalized to on-axis values. In FIG. 5, line 502 represents the on-axis response, line 504 is the frequency response measured at 10 degrees, line 506 is the response at 20 degrees, line 508 is the response at 30 degrees, and line 510 Are frequency responses measured at 40 degrees, all measured in the frequency range between 1 kHz and 20 kHz.

Further, minimization is performed by changing the real-valued frequency points of the channel filter C _opt (n, f) within the interval [0,1], where n is the sign of the driver and f is Is the frequency. Further, the following constraints C _opt (n, f) = 0, f> fo, f <fu
However, it must be satisfied according to the characteristics of the specific driver n. For example, in the woofer example, the upper operating limit is fo = 1 kHz, for the tweeter the lower operating limit is fu = 2 kHz, for the mid-range driver fu = 300 Hz, fo = 3 kHz. possible.

  The cost function minimization procedure described above is described in The MathWorks, Inc. Can be implemented by the function “fminsearch”, which is part of the Matlab® software package owned and distributed by. The “fminsearch” function in the Matlab software package uses the Nelder-Mead simplex algorithm or a derivative thereof. As an alternative, an exploratory exploration method can be applied on a predefined grid within constrained parameters. Other methodologies can also be used to minimize the cost function.

  If the difference between the obtained result and the target is small enough, or acceptable as determined by the practitioner for a particular design application, then each of the linear arrays FIR filter coefficients for the signal path are obtained.

  If the difference between the obtained result and the target is not acceptable for a particular design application, for example, or too large, the driver position or geometry and / or the parameter q (I) and the fc (see FIG. 4) of the objective function T (f, g) are then modified. After correction, the cost minimization function is applied again and the process is repeated until the results and targets obtained are sufficiently small or within acceptable limits for the application.

  As shown in FIG. 3, if the driver position and driver geometry are determined so that the algorithm gives results within the tolerance of the objective function, then each signal path n = 1,. The FIR filter coefficients for N need to be determined, as shown in step 318 of FIG. One way to determine FIR filter coefficients is to use a Fourier approximation method (frequency sampling method) to obtain a linear phase filter of a given degree. When applying the Fourier approximation method or other frequency sampling methods, it is necessary to select an order such that the approximation method is sufficiently accurate.

  The Fourier approximation method is described in The MathWorks, Inc. Can be implemented by the function “fires”, which is part of the Matlab® software package owned and distributed by. Similar methodologies can be used to minimize the cost function by being executed in other software systems.

  Additionally, modifications can be made to the FIR filter to equalize the measured frequency response of one or more drivers (especially tweeters, midrange drivers). The impulse responses of these filters can be obtained by well-known methods and need to be convolved with the impulse response of the linear phase channel filter when determining the coefficients of the FIR filter as described above. Furthermore, the voice coil (the driver's acoustic center) cannot be aligned. To compensate for this, an appropriate delay can be given to the filter by adding a leading zero to the FIR impulse response.

  Two-dimensional multi-way loudspeaker systems can be planned for use in conjunction with various applications, such as stereo loudspeaker systems, multi-channel home entertainment systems, and public information broadcast systems. The person skilled in the art can change the number, type and position of the drivers, the number of channels, the number of signal flow paths or circuits, as well as the axis to adapt the directivity for a particular application. The position parameters along can be modified.

  FIG. 6 is another two-dimensional multi-way loudspeaker, the loudspeaker system of FIG. 2 except that it includes four woofers 620, 622, 624 and 626 instead of two woofers. Same as loudspeaker. The arrangement described in FIG. 6 is a design that may be desirable for those skilled in the art to use for sound enhancement applications.

  In the example provided in FIG. 6, the central tweeter 602 may be attached to a center point 0 on the x-axis, shown as the intersection of the x-axis and y-axis in FIG. The tweeters 604, 606, 608 and 610 are centered about +/− 42 mm along the y-axis and about +/− 30 mm along the x-axis (+/− 30 mm, +/− 42 mm). It is attached.

  The mid-range drivers 612, 614, 616 and 618 are then centered at positions about +/− 110 mm from their center point 0 along the y-axis and about +/− 80 mm along the x-axis (+/− −80 mm, +/− 110 mm). Woofers 620, 622, 624 and 626 are then centered about +/− 300 mm along the y-axis and approximately +/− 180 mm along the x-axis (+/− 180 mm, +/− 300 mm). ).

  Similar to the respective loudspeaker systems 100 and 200 of FIGS. 1 and 2, the transducer includes a compartment 630, 632, and 634 separated and sealed by separators 636 and 642 as shown. Can be mounted inside.

  FIG. 7 shows a signal flow block diagram 700 of the multi-way loudspeaker system 600 of FIG. FIG. 7 shows the four ways 702, 704, 706 and 708 of the signal flow. The signal flow comprises a digital input 710 that can be implemented by using a standard connection format and is connected to the driver via various paths or ways such as the four ways shown in FIG. . Each path or way 702, 704, 706 and 708 is connected to one or more loudspeaker drivers, respectively, with digital FIR filters 712, 714, 716, 718 and power D / A converters 720, 722, 724, 726 may be included.

  As shown in FIG. 7, signal flow way 702 provides to woofers 620, 622, 624 and 626 of loudspeaker system 600 of FIG. Signal flow way 704 provides to mid-range drivers 612, 614, 616 and 618 of loudspeaker system 600 of FIG. The signal flow way 706 feeds the tweeters 604, 606, 608 and 610 of the loudspeaker system 600 of FIG. 6, and the signal flow way 708 feeds the central tweeter 602 of the loudspeaker system 600 of FIG. To do.

  FIG. 8 is a graph of acceptable results obtained for the frequency response of the four filters shown in FIG. 7 when applied to a loudspeaker system similar to that shown in FIG. 800 is shown. Specifically, line 802 represents the results for the frequency response of FIR filter 712. Line 804 shows the results for the frequency response of FIR filter 714, line 806 shows the results for the frequency response of FIR filter 716, and line 718 shows the results of the frequency response of FIR filter 718.

  FIG. 9 is a graph 900 illustrating the resulting horizontal (y-axis) frequency response at various angles. This graph shows the filter frequency response V (f, q) obtained after passing through step 314 of FIG. Passing means that the result met the requirement. Specifically, line 902 represents the resulting horizontal axis response V (f, q (1)), line 904 is the frequency response V (f, q (2)) at 5 degrees, Line 906 is the response V (f, q (3)) at 10 degrees, line 908 is the response V (f, q (4)) at 15 degrees, and line 910 is the response V (f, q (20) at 20 degrees. (5)), line 912 is the response V (f, q (6)) at 25 degrees, line 914 is the response V (f, q (7)) at 30 degrees, and line 916 is 35 The response in degrees V (f, q (8)), all shown in the frequency range between 100 Hz and 20 kHz.

  FIG. 10 is a graph 1000 illustrating the resulting vertical (x-axis) frequency response at various angles. Specifically, line 1002 represents the resulting vertical axis response V (f, q (1)), line 1004 is the frequency response V (f, q (2)) at 5 degrees, Line 1006 is the response V (f, q (3)) at 10 degrees, line 1008 is the response V (f, q (4)) at 15 degrees, and line 1010 is the response V (f, q (20) at 20 degrees. (5)), line 1012 is the response V (f, q (6)) at 25 degrees, line 1014 is the response V (f, q (7)) at 30 degrees, and line 1016 is 35 The response in degrees V (f, q (8)), all shown in the frequency range between 100 Hz and 20 kHz.

  FIGS. 11-22 show an embodiment of a multi-way loudspeaker for a loudspeaker system suitable for home entertainment.

  FIG. 11 shows an example of a one-dimensional (1D) 7-way loudspeaker system 1100 that is symmetrically incorporated along the x-axis, and a block diagram 1160 of the signal flow to each loudspeaker driver of the system. . This embodiment may serve as a basis for the two-dimensional (2D) multi-way loudspeaker system designs 1400 and 1700 shown in FIGS. 14 and 17, which system designs may be used for home entertainment applications or other known by those skilled in the art. Can be designed for use in appropriate applications.

  As shown in FIG. 11, a one-dimensional 7-way loudspeaker system 1100 includes (1) one central tweeter 1102 placed at the origin, and (2) both sides of the central tweeter 1102 along the x-axis. First pair of tweeters 1104 and 1106, placed at +/− 0.035 m, (3) on both sides of the first pair of tweeters, at +/− 0.07 m along the x-axis. A second pair of tweeters 1108 and 1110, (4) a first pair of midrange drivers 1112 and 1114, located at +/− 0.12 m along the x-axis, (5) x− A second pair of mid-range drivers 1116 and 1118, located at +/− 0.20 m along the axis, (6) located at +/− 0.34 m along the x-axis, Three pairs of midrange drivers 1120 and 112 , And (7) x- along the axis is placed at a position of +/- 0.54 m, may include a woofer 1124 and 1126, of the pair.

  As in the previously shown embodiments, the driver can be contained in a housing having various compartments. Tweeters 1102, 1104, 1106, 1108 and 1110, and midrange drivers 1112 and 1114 may be placed in a single partition 1130. A compartment 1136 placed adjacent to the compartment 1130 and separated one by the separator 1132 includes an intermediate zone driver 1116. A compartment 1138 that is placed on the opposite side of the compartment 1130 and separated by the separator 1134 includes an intermediate zone driver 1118. Compartment 1144 includes mid-range driver 1120, separated on one side from compartment 1136 by separator 1140, and on the opposite side from compartment 1152 containing woofer 1124 by separator 1148. Similarly, compartment 1146 includes mid-range driver 1122, separated on one side from compartment 1138 by separator 1142 and on the opposite side from compartment 1154 containing woofer 1126 by separator 1150.

  The loudspeaker system 1100 can receive a digital input 1180. The signal flow diagram 1160 shows that the central tweeter 1102 is supplied by a signal flow way 1174 that includes a FIR filter 1176 and a power D / A converter 1178. The first pair of tweeters 1104 and 1106 is provided by a signal flow way 1172 that includes an FIR filter 1178 and a power D / A converter 1178, and the second pair of tweeters 1108 and 1110 includes an FIR filter 1180 and a power D Supplied by a signal flow way 1170 including an A converter 1178. The first pair of midrange drivers 1112 and 1114 is provided by a signal flow way 1168 that includes a FIR filter 1182 and a power D / A converter 1178, while the second pair of midrange drivers 1116 and 1118 includes: Supplied by signal flow way 1166 including FIR filter 1184 and power D / A converter 1178. The third pair of midrange drivers 1120 and 1122 is provided by a signal flow way 1164 that includes a FIR filter 1186 and a power D / A converter 1178. Finally, the pair of woofers 1124 and 1126 is provided by a signal flow way 1162 that includes a FIR filter 1188 and a power D / A converter 1178.

  FIG. 12 is a graph 1200 showing the frequency characteristics of the seven filters of the loudspeaker system of FIG. 11, where a cost minimizing function is applied and the resulting result is small enough or acceptable for the desired application. It is already known to be in. The line represented by 1202 is the frequency response of the FIR filter 1176, the line 1204 is the frequency response of the FIR filter 1178, the line 1206 is the frequency response of the FIR filter 1180, and the line 1208 is the frequency response of the FIR filter 1182. , Line 1210 is the frequency response of FIR filter 1184, line 1212 is the frequency response of FIR filter 1186, and line 1214 is the frequency response of FIR filter 1188.

  13 is a graph 1300 illustrating the resulting horizontal (x-axis) frequency response of the loudspeaker system of FIG. 11, measured at various angles. This graph shows the filter frequency response V (f, q) obtained after meeting the requirements of step 314 of FIG. Specifically, line 1302 represents the resulting response V (f, q (1)) on the horizontal axis, line 1304 is the frequency response V (f, q (2)) at 10 degrees, Line 1306 is the response V (f, q (3)) at 15 degrees, line 1308 is the response V (f, q (4)) at 20 degrees, and line 1310 is the response V (f, q (30) at 30 degrees. (5)), all shown in the frequency range between 100 Hz and 20 kHz.

  FIG. 14 shows an embodiment of a two-dimensional (2D) multi-channel 7-way loudspeaker system 1400 that is symmetrically incorporated along the x-axis and y-axis. The loudspeaker system 1400 is obtained by separating the tweeters 1104, 1106, 1108 and 1110 and the mid-range drivers 1112 and 1114 of the loudspeaker system 1100 of FIG. 11 into pairs.

  The loudspeaker system 1400 controls the directivity in two directions, and includes a central tweeter 1402, four pairs of tweeters 1404 and 1406, 1408 and 1410, 1412 and 1414, and 1416 and 1418, and four pairs of intermediate regions. Drivers 1420 and 1422, 1424 and 1426, 1428 and 1430, and 1432 and 1434, and a pair of woofers 1436 and 1438 are included. The first two pairs of tweeters 1404, 1406, 1408, and 1410 are arranged in a square shape with respect to the central tweeter 1402. Third and fourth pairs of tweeters 1412, 1414, 1416 and 1418 are placed symmetrically along the x and y axes and on a more distant quadrant. The first and second pairs of midrange drivers 1420, 1422, 1424 and 1428 are placed symmetrically along the x and y axes and on a farther quadrant. As described further below, the inner quadrant is defined by an angle of 45 (45) degrees with respect to the x-axis.

  Additionally, midrange drivers 1428, 1430, 1432 and 1434, and woofers 1436 and 1438 are linearly spaced on the x-axis. The (x, y) coordinates of the driver of the loudspeaker 1400 can be as follows:

Tweeter 1402: (0,0)
Tweeters 1404, 1406, 1408 and 1410: (+/− 35, +/− 35) mm
Tweeters 1412, 1414, 1416 and 1418: (+/− 70, +/− 70) mm
Intermediate areas 1420, 1422, 1424 and 1426: (+/− 120, +/− 120) mm
Midrange 1428 and 1430: (+/− 200,0) mm
Midrange 1432 and 1434: (+/− 340,0) mm
Woofers 1436 and 1438: (+/− 540,0) mm
As with the loudspeaker system shown in FIG. 11, the driver may be mounted in a baffle 1476 comprising separate and sealed compartments 1440, 1442, 1444, 1446, 1448, 1450 and 1452. . The tweeters 1402, 1404, 1406, 1408, 1410, 1412, 1414, 1416 and 1418, and the midrange drivers 1420, 1422, 1424 and 1426 can all be included in the partition 1440. On the right side, compartment 1440 may be separated from compartment 1444 by a separator represented by triangular line 1460. Compartment 1444 includes midrange driver 1430, and its right side can be separated from compartment 1448 that includes midrange driver 1434 by a separator indicated by line 1464. To the right of compartment 1448 is a compartment 1452 that includes a woofer 1438. Compartments 1448 and 1452 may be separated from each other by a separator represented by line 1468.

  Similarly, compartment 1440 can be separated from compartment 1442 on the left by a separator represented by triangular line 1462. Compartment 1442 includes midrange driver 1428, and its left side can be separated from compartment 1446 that includes midrange driver 1432 by a separator indicated by line 1466. To the left of the compartment 1446 is a compartment 1450 that includes a woofer 1436. Compartments 1446 and 1450 may be separated from each other by a separator represented by line 1470.

  As with the drivers of FIGS. 1 and 2, the tweeters 1402, 1404, 1406, 1408, 1410, 1412, 1414, 1416 and 1418 can be about 40 mm to 50 mm in outer diameter and the mid-range drivers 1420, 1422, 1424. , 1426, 1428, 1430, 1432 and 1434 may have an outer diameter of about 80 mm to 110 mm, and woofers 1436 and 1438 may have an outer diameter of about 120 mm to 160 mm.

  FIG. 15 is a block diagram 1500 of the signal flow to each loudspeaker driver in the loudspeaker system 1400 of FIG. As shown in FIG. 15, each previously set driver with the same set of position coordinates is fed by a different path or way, resulting in a seven way loudspeaker. The loudspeaker system 1400 receives a digital input 1502. The central tweeter 1402 is provided by a signal flow way 1504. The tweeters 1404, 1406, 1408 and 1410 are provided by a signal flow way 1506. The tweeters 1412, 1414, 1416 and 1418 are provided by a signal flow way 1508. The midrange drivers 1420, 1422, 1424 and 1426 are provided by the signal flow way 1510, while the midrange drivers 1428 and 1430 are provided by the signal flow way 1512 and the midrange drivers 1432 and 1434 are Supplied by signal flow way 1514. Pairs of woofers 1436 and 1438 are provided by a signal flow way 1516. Each signal flow way includes an FIR filter 1518 and a power D / A converter 1520.

  FIG. 16 is a graph 1600 representing the resulting vertical (y-axis) frequency response of the loudspeaker system 1400 of FIG. 14 measured at various angles. This graph shows the frequency response V (f, q) of the filter obtained after meeting the requirements of step 314 of FIG. Specifically, line 1602 represents the resulting response on the horizontal axis V (f, q (1)), line 1604 is the frequency response V (f, q (2)) at 10 degrees, Line 1406 is the frequency response V (f, q (3)) at 15 degrees, line 1608 is the frequency response V (f, q (4)) at 20 degrees, and line 1610 is the frequency response V (30 at 30 degrees. f, q (5)), all shown in the frequency range between 100 Hz and 20 kHz. As seen by FIG. 16, the vertical frequency response for the two-dimensional loudspeaker system 1400 of FIG. 14 is similar to the horizontal frequency response shown by FIG. 13 for the one-dimensional loudspeaker system 1100 of FIG. However, the lower corner frequency at which the directivity of the system appears above that frequency has a fairly high frequency value.

  FIG. 17 shows an example of a two-dimensional (2D) 5-channel multi-way loudspeaker system 1700 mounted symmetrically along the x-axis. The loudspeaker system 1700 is designed to have a pair of integrated two-way stereo speakers, mounted symmetrically along the x-axis, specifically designed for use for home theater applications. Yes. As further described below (FIGS. 18-20), the loudspeaker system 1700 includes five input channels L (left), R (right), C (center), LS (left surround), and RS. (Right surround).

  The loudspeaker system 1700 is the same as the system of FIG. 14 except that it includes two additional tweeters 1744 and 1746, and two additional woofers, with the difference that the outer woofer is y- With respect to the axis, it is separated into 1736 and 1738 pairs, and 1740 and 1742 pairs, with an additional pair of tweeters 1744 and 1746 located between each pair of woofers 1736 and 1738, 1740 and 1742. By having tweeters 1744 and 1746 designated for woofer pairs 1736 and 1738, and 1740 and 1742, respectively, loudspeaker system 1700 provides independent stereo speaker channels (eg, tweeters are provided by separate channels). Can be provided with an array of signals). The purpose of the independent stereo speaker channel is to provide an integrated surround sound system with directional sound beams generated by traditional stereo speakers and arrays, and indirectly behind the surroundings using listening room wall reflections (Ambient rear) Playing a channel.

  Similar to the loudspeaker system 1400 shown in FIG. 14, the loudspeaker system 1700 of FIG. 17 has two pairs arranged in a square shape with respect to (1) the central tweeter 1702 and (2) the central tweeter 1702. Tweeters 1704 and 1706, and 1708 and 1710, (3) two additional pairs of tweeters 1712 and 1714 placed symmetrically along the x- and y-axes and on a farther quadrant. , And 1716 and 1718, and (4) two pairs of mid-range drivers 1720 and 1722, and 1724, placed symmetrically along the x-axis and y-axis, on a further quadrant. And 1726. The quadrant is defined by an angle of 45 (45) degrees with respect to the x-axis.

  Additionally, the loudspeaker system 1700 includes midrange drivers 1728, 1730, 1732 and 1743, which are spaced linearly along the x-axis. The (x, y) coordinates of the driver of the loudspeaker system 1700 may be as follows:

Tweeter 1702: (0,0)
Tweeters 1704, 1706, 1708 and 1710: (+/− 35, +/− 35) mm
Tweeters 1712, 1714, 1716 and 1718: (+/− 70, +/− 70) mm
Intermediate areas 1720, 1722, 1724 and 1726: (+/− 120, +/− 120) mm
Midrange 1728 and 1730: (+/− 200,0) mm
Intermediate zone 1732 and 1734: (+/− 340,0) mm
Tweeters 1744 and 1746: (+/− 540,0) mm
Woofers 1736, 1738, 1740 and 1742 (+/− 540, +/− 90) mm
As with the loudspeaker system shown in FIGS. 1, 2, 6, 11 and 14, the loudspeaker system 1700 drivers are represented by lines 1766, 1768, 1770, 1772, 1774 and 1776, respectively. It can be mounted in a divider or housing 1750 comprising separated and sealed compartments 1752, 1754, 1756, 1768, 1760, 1762 and 1764 that are separated from each other by a separator.

FIGS. 18-20 show block diagrams of signal flow for the five input signals of the loudspeaker system 1700 of FIG. 18 is a block diagram 1800 of the signal flow for the surround channel for the loudspeaker system 1700 of FIG. Since the signal flow for the right and left surround channels of the system 1700 is the same except for different delay values, the block diagram 1800 of FIG. It represents the signal flow path for both the right and right surround. Thus, both left and right surround input signals are transmitted via a signal path system similar to that shown in FIG. The sum of each output signal is then calculated and connected to the transducer, as described in FIG. The frequency response of the output of the FIR filter is shown in FIG. 12, and is connected to the delay line D ₀ and the pair of delay lines D _{+/− (1,..., 6)} , respectively.

The signal flow block diagram 1800 of FIG. 18 shows that a delay is added to each path according to the following equation:

Δt = p / c · sin α, (p = driver coordinates (meter), c = 345 m / sec (sound speed))
Here, the main sound beam is otherwise perpendicular to the main axis, but can be directed in the desired direction of the angle α. Typical values for α are-(40, ..., 60) degrees for the left surround and + (40, ..., 60) degrees for the right surround. Means that the sound beam is formed and directed toward the side walls in the directions of angles α and −α and bounces off the walls, thereby reaching the listener as a surround signal.

As shown in FIG. 18, signal flow path diagram 1800 shows the respective flow paths for digital inputs 1802 and 1804 for the right and left surround sound channels. The output of FIR filter 1822 for path 1806 is connected to delay line (D ₀ ) 1840, which is connected to central tweeter 1702. The output of FIR filter 1824 for path 1808 is connected in parallel to delay lines (D ₋₁ ) 1842 and (D ₊₁ ) 1844. Delay line 1842 is connected to the right pair of tweeters 1708 and 1710 and delay line 1844 is connected to the left pair of tweeters 1704 and 1706. Similarly, the output of FIR filter 1826 for path 1810 is connected in parallel to delay lines (D ₋₂ ) 1846 and (D ₊₂ ) 1848. Delay line 1846 is connected to the right pair of tweeters 1716 and 1718 and delay line 1848 is connected to the left pair of tweeters 1712 and 1714. Delay lines (D ₋₃ ) 1850 and (D ₊₃ ) 1852 are connected to mid-range drivers 1720 and 1722, and 1724 and 1726, respectively, which are connected in parallel to path 1812, which is the output path for FIR filter 1828 It is.

Midband drivers 1728 and 1730 are connected to delay lines (D ₊₄ ) 1856 and (D ₋₄ ) 1854, respectively, which is the output path 1814 to FIR filter 1830. Midrange drivers 1732 and 1734 are connected to delay lines (D ₊₅ ) 1862 and (D ₋₅ ) 1860, respectively, which is the output path 1816 to FIR filter 1832.

The right pair of woofers 1740 and 1742 are connected to delay line (D- ₆ ) 1864,
The left pair of woofers 1736 and 1738 is connected to delay line (D ₊₆ ) 1866. Delay line (D ₊₆ ) 1866 and delay line (D ₋₆ ) 1864 are connected in parallel to output path 1820 for FIR filter 1834.

  FIG. 19 is a block diagram of the signal flow for the right and left channels for the loudspeaker system of FIG. The left and right channels are integrated like a conventional 2-way speaker. The left channel consists of a tweeter 1744 that is not part of the beamforming array, and woofers 1736 and 1738. The right channel consists of a tweeter 1746 and woofers 1740 and 1742.

  As shown by FIG. 19, signal processing 1900 for the left and right channels includes a stereo widening circuit consisting of an HD filter 1910 and an HI filter 1912 to expand the stereo basis, and a low pass. A crossover circuit having a filter 1914 and an HP high pass filter 1916 is used.

  FIG. 20 is a block diagram of signal flow for the center channel for the loudspeaker system 1700 of FIG. The center channel includes tweeters 1702, 1704, 1706, 1708, 1710, 1712, 1714, 1716 and 1718, and midband drivers 1720, 1722, with FIR filters having coefficients determined as defined in FIG. Reproduced by the inner array of 1724 and 1726.

  The digital signal output 2010 for the central channel is divided into four signal paths 2002, 2004, 2006 and 2008, each path having a FIR filter 2012, 2014, 2016 and 2018, respectively, and a power D / A converter 2020. , 2022, 2024 and 2026, respectively. The path 2002 feeds the central tweeter 1702. Path 2004 feeds the innermost quadrant tweeters 1704, 1706, 1708 and 1710. Path 2006 feeds to the outermost quadrant tweeters 1712, 1714, 1716 and 1718, and path 2008 feeds the middle quadrant midrange driver 1720, 1722, 1724 and 1726.

  FIG. 21 is a graph 2100 showing the frequency response of the four FIR filters used for the center channel (FIG. 20) of the loudspeaker system of FIG. Line 2102 represents the frequency response of FIR filter 2012, line 2104 represents the frequency response of FIR filter 2014, line 2106 represents the frequency response of FIR filter 2016, and line 2108 represents the frequency response of FIR filter 2018.

  FIG. 22 shows the resulting horizontal (x-axis) and vertical (y-axis) of the center channel output (FIG. 20) of the loudspeaker system 1700 of FIG. 17, measured at various angles. ) Is the same graph 2200 representing the frequency response. This graph shows the frequency response V (f, q) of the filter obtained after meeting the requirements of step 314 of FIG. In particular, line 2202 represents the resulting horizontal axis response V (f, q (1)), line 2204 is the frequency response V (f, q (2)) at 5 degrees, Line 2206 is the frequency response V (f, q (3)) at 10 degrees, line 2208 is the frequency response V (f, q (4)) at 15 degrees, and line 2210 is the frequency response V (20 at 20 degrees. f, q (5)), line 2212 is the frequency response V (f, q (6)) at 25 degrees, line 2214 is the frequency response V (f, q (7)) at 30 degrees, Line 2216 is the frequency response V (f, q (8)) at 35 degrees, all shown in the frequency range between 100 HZ and 20 kHz.

  While various embodiments of the present invention have been described, it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the present invention. Accordingly, the invention should not be limited except in light of the attached claims and their equivalents.

  The present invention is a multi-way loudspeaker system that provides affordable loudspeaker configurations and filter design methodologies that operate in the area of digital signal processing. Furthermore, the loudspeaker system can be designed as a multi-way loudspeaker system with loudspeaker drivers arranged symmetrically in a two-dimensional plane, covering a wide range of high quality sound, ie both vertical and horizontal A certain directivity can be achieved and can be used in combination with a stereo loudspeaker system, a multi-channel home entertainment system and a public information broadcasting system.

FIG. 3 shows an example of a one-dimensional four-way loudspeaker system, mounted symmetrically at the origin along the y-axis, and a block diagram of the signal flow to each loudspeaker driver in the system. FIG. 2 shows an example of a two-dimensional 4-way loudspeaker system, mounted symmetrically at the origin along the x-axis and y-axis, and a block diagram of the signal flow to each loudspeaker driver in the system. Fig. 3 is a flow chart of a filter design algorithm used to design a loudspeaker system. It is a graph which shows a directivity objective function with respect to attenuation depending on an angle. FIG. 6 is a graph showing the amplitude frequency response of an attached tweeter measured at various vertical off-axis deflection angles. Fig. 5 shows an example of another two-dimensional four-way loudspeaker system mounted symmetrically at the origin along the x-axis and y-axis. FIG. 7 is a block diagram of signal flow to each loudspeaker driver shown in FIG. 6. 7 represents the frequency response of four filters of the loudspeaker system of FIG. FIG. 7 shows the resulting horizontal (y-axis) frequency response of the loudspeaker system of FIG. 6 measured at various angles. FIG. 10 shows the resulting vertical (x-axis) frequency response of the loudspeaker system of FIG. 6 corresponding to the horizontal response shown in FIG. FIG. 4 shows an example of a one-dimensional (1D) 7-way loudspeaker system, mounted symmetrically at the origin along the y-axis, and a block diagram of the signal flow to each loudspeaker driver in the system. 12 shows the frequency response of seven filters of the loudspeaker system of FIG. FIG. 12 shows the resulting horizontal (x-axis) frequency response of the loudspeaker system of FIG. 11, measured at various angles. Fig. 4 illustrates an embodiment of a two-dimensional (2D) multi-channel 7-way loudspeaker system mounted symmetrically along the x-axis and y-axis. FIG. 15 is a block diagram of signal flow to each loudspeaker driver of the loudspeaker system of FIG. 14. FIG. 15 shows the resulting vertical (y-axis) frequency response of the loudspeaker system of FIG. 14 measured at various angles. Fig. 4 illustrates an example of a two-dimensional (2D) 5-channel multi-way loudspeaker system that is mounted symmetrically along the x-axis and y-axis and designed for home theater applications. FIG. 18 is a block diagram of the signal flow for the right and left surround channels for the loudspeaker system of FIG. FIG. 18 is a block diagram of signal flow for the right and left channels for the loudspeaker system of FIG. FIG. 18 is a block diagram of signal flow for the center channel for the loudspeaker system of FIG. 17. FIG. 18 shows the frequency response of four filters in the center channel of the loudspeaker system of FIG. FIG. 18 shows the resulting horizontal (x-axis) frequency response of the center channel of the loudspeaker system of FIG. 17 measured at various angles.

Claims (15)

A loudspeaker array,
The loudspeaker array
At least one digital FIR filter configured to receive a digital audio signal from an audio sound source;
At least one power D / A converter configured to receive a filtered signal from the at least one FIR filter, the at least one digital FIR filter relative to an optimized transducer position; At least one power D / A converter including the determined linear phase filter coefficients;
A plurality of transducers of at least two different dimensions, wherein the plurality of transducers are arranged symmetrically about the first axis and the second axis at the position of the optimized transducer; The second axis is perpendicular to the first axis , and the position of the optimized transducer in a two-dimensional loudspeaker array is the coordinates of the plurality of transducers along the first axis. Determined and optimized according to a cost minimization function, wherein the plurality of transducers is a first pair of identical transducers arranged symmetrically about the first axis, A first pair of identical transducers aligned parallel to the second axis, and a second pair of transducers identical to the first pair of transducers, The second pair is disposed symmetrically with the first pair about the second axis, and the second pair A plurality of converters coupled to the at least one power D / A converter; and a loudspeaker array.   The loudspeaker array of claim 1, wherein the plurality of transducers of the at least two different dimensions are tweeters and mid-range drivers.   The loudspeaker array of claim 1, wherein the plurality of transducers of the at least two different dimensions are a tweeter and a woofer.   The loudspeaker array according to claim 1, wherein the plurality of drivers includes a tweeter, a mid-range driver, and a woofer.   The loudspeaker array according to claim 1, further comprising a center transducer, wherein the center of the center transducer is at the intersection of the first axis and the second axis.   The loudspeaker array of claim 5, wherein the central converter receives a signal filtered through the at least one digital FIR filter from the at least one power D / A converter.   The linear phase filter coefficients for each FIR filter determine an initial driver position, determine an initial directivity objective function for the system, and determine a cost minimization function based on the initial directivity objective function. The loudspeaker array of claim 1, determined by applying and calculating a linear phase filter coefficient for each filter of the system.   The loudspeaker array according to claim 7, wherein the initial driver position is a coordinate with respect to an origin that is a center of the loudspeaker.   8. A loudspeaker array according to claim 7, wherein frequency points are determined on a logarithmic scale of a predetermined frequency range based on the determined initial directivity objective function.   8. The loudspeaker array of claim 7, wherein the cost minimizing function is applied at the frequency points, starting at a minimum frequency and increasing in steps.   The loudspeaker array of claim 7, wherein a Fourier approximation method is utilized to determine the linear phase filter coefficients. A method for changing the directivity of a two-dimensional loudspeaker array along a first axis perpendicular to a second axis, wherein each center of a plurality of transducers of the one-dimensional loudspeaker array is initially Is on the second axis;
The method
Forming the two-dimensional loudspeaker array by replacing at least one transducer of the one-dimensional loudspeaker array with at least one pair of transducers, the at least one pair of transducers comprising: Substantially identical to the replaced at least one transducer, the at least one pair of transducers being arranged symmetrically with respect to the first axis and the second axis;
The transformation of the two-dimensional loudspeaker array by changing the coordinates of at least one transducer along the first axis until the two-dimensional loudspeaker array meets a predetermined performance requirement according to a cost minimizing function. Optimizing the position of the vessel;
Calculating linear phase filter coefficients for at least one digital FIR filter connected to a transducer located at the location of the optimized transducer.   The method of claim 12, wherein the pair of converters receives a digital output signal filtered through the at least one digital FIR filter from at least one power D / A converter. A method for determining linear phase filter coefficients for a plurality of digital filters connected to a plurality of transducers arranged in a two-dimensional array comprising:
The method
Determining an initial position for the plurality of transducers, the plurality of transducers being symmetric about a first axis and a second axis, wherein the second axis is the first axis; Being perpendicular to the axis,
Determining an initial directivity objective function for the plurality of transducers at the initial position;
Minimizing the cost function at a specified frequency point;
Comparing the result of the step of minimizing the cost function with the initial directivity objective function;
If the difference between the result of the step of minimizing the cost function and the initial directivity objective function meets a predetermined performance requirement, calculating the linear phase filter coefficients for the plurality of filters;
If the difference between the result and initial directivity target function the step of minimizing the cost function it does not meet the predetermined performance requirements are the positions of the plurality of transducers along the axis of the first Changing and minimizing the cost function and comparing the result with the initial directivity objective function. A loudspeaker system,
The loudspeaker system includes:
At least one digital FIR filter configured to receive a digital audio signal from an audio sound source;
At least one power D / A converter configured to receive a filtered signal from the at least one FIR filter, the at least one digital FIR filter relative to an optimized transducer position; At least one power D / A converter including the determined linear phase filter coefficients;
At least five transducers of at least two different dimensions, wherein four of the at least five transducers are arranged symmetrically with respect to both the first axis and the second axis. The second axis is perpendicular to the first axis, and the center of one of the at least five transducers is at the intersection of the first axis and the second axis. And the symmetrically arranged transducer comprises at least five transducers arranged at the location of the optimized transducer, and
The position of the optimized transducer in a two-dimensional loudspeaker array is determined and optimized according to a cost minimizing function by changing the coordinates of the plurality of transducers along the first axis. And
The at least five transducers include at least one pair of identical transducers, and the at least one pair of identical transducers is symmetrically arranged about the first axis and is on the second axis A loudspeaker system, wherein the plurality of converters are configured to receive an audio signal output from the at least one power D / A converter.

Images