JP2010016573A - Crosstalk canceling stereo speaker system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain a stereo speaker which reduces crosstalk without causing secondary crosstalk to reproduce a surround sound field and to obtain the same effect by general speakers. <P>SOLUTION: Main units which reproduce right and left channels by right and left speakers and auxiliary units which reproduce opposite channels in opposite phases are provided in parallel with both ears of a listener so that the right and left auxiliary units are on the inside of the interval between both ears with respect to the right and left main units, and crosstalk is limited to middle and low ranges wherein secondary crosstalk is satisfactorily canceled regardless of the occurrence of arrival difference from crosstalk when being canceled by this configuration, whereby crosstalk is canceled. In a crosstalk canceling method, each of right and left inputs is divided into two, and right and left filter block outputs are subtracted from opposite delay block outputs to obtain right and left outputs, and crosstalk is limited to middle and low ranges, whereby crosstalk is canceled. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a stereo speaker and a crosstalk canceling method for reducing the crosstalk of a stereo speaker without secondary crosstalk and reproducing a surround sound field.

  Conventionally, the ideal stereo playback is to record the sound of the recording site with L and R volume difference information and phase difference information by pair microphones installed close to the distance between human ears, and transmit and amplify the recorded sound with low distortion. Then, playback is performed with a signal converter equivalent to the pair microphone, and the pair microphone and the signal converter are offset in terms of audibility. I came.

There is binaural in the above method. When the recorded sound from the left and right microphones attached to the dummy head is played back with headphones, and the microphone and headphones are offset as equivalent signal converters, the listener can listen to the sound at the recording site with both ears. equal.
Note that biphonic is a method of reproducing stereo-recorded sound with headphones. Since the pair microphone and headphones are not equivalent, the reproduced sound field becomes narrower with in-head localization. In order to avoid this localization in the head, there is a technique for giving a reproduced sound a filter characteristic based on a head related transfer function (HRTF).

  On the other hand, stereophonic is a system in which two speakers are installed at the same angle and at the same angle in front of the listener to listen to stereo recorded sound with both ears. Stereo recorded sound usually includes the same components common to the L and R channels (hereinafter referred to as monaural components) and L and R components other than monaural components. Recognize the center localization between speakers by crosstalk of components, and recognize the three-dimensional sound field localization by vector synthesis of L, R volume difference information and phase difference information other than monaural components using the above-mentioned center localization as a reference vector To do.

However, when a stereo speaker is viewed from a small angle, crosstalk of L and R components other than monaural components increases, and the reproduction sound field is reduced between the left and right speakers.
In order to improve the above tendency, the stereo speaker is arranged to look at a large angle such as 60 ° recommended by ITU-R. As a result, when the crosstalk is reduced due to the head shielding effect, the stereo speaker also reproduces a certain surround sound field. However, the crosstalk remains in the lower sound range, and the reproduced sound field is distorted between the speakers, so that the size of a medium or large musical instrument cannot be reproduced. As the left and right speakers are viewed from a larger angle, the sound produced by each of the left and right speakers is heard by both ears with a time lag, and the transient characteristics deteriorate.
In addition, many persons skilled in the art think that the size of a medium-sized musical instrument is reproduced only by a large-diameter woofer.

Therefore, many crosstalk cancellation systems for canceling the crosstalk and reproducing the surround sound field originally included in the stereo source have been commercialized.
In the basic crosstalk cancellation circuit 5 shown in FIG. 2, the left audio input L is set to W and X, the right audio input R is set to Y and Z, and X and Y are delayed with respect to W and Z. After providing the filter characteristics, X is subtracted from Z to obtain the right audio output, and Y is subtracted from W to obtain the left audio output.
Note that the output method is not limited to the above two channels. When a crosstalk canceling speaker is used in addition to the left and right stereo speakers, the output of the crosstalk canceling circuit 5 is correspondingly three or more. However, in a speaker configuration with three or more channels, the relationship with the listener's ears is different, so in this specification, only the method of amplifying the left and right audio outputs of the crosstalk cancellation circuit 5 and reproducing them with stereo speakers will be mentioned. .

When the reproduced sound reaches the far ear from the left and right speakers at the same distance and the same angle with respect to the listener, the filter characteristic generated by the head shielding effect and the structure of the auricle is represented by a variable K (hereinafter the same in this specification). And a filter characteristic that is added to the crosstalk cancellation signal in accordance with this and that is different from normal K is a variable k (hereinafter also expressed in the same manner). In addition, the distance a measured by a straight line when the reproduced sound reaches the listener's one ear position from both the left and right speakers is defined as a variable a (hereinafter also expressed in the same manner), and one of the left and right speaker reproduced sounds. A distance difference added to a so as to wrap around the head is defined as a variable aa (hereinafter also expressed in the same manner).
If the variables K, k, a and aa are used, when the crosstalk cancellation signal -kR of the left speaker of the stereo speakers is delayed as (a + aa) behind the L playback position, The crosstalk KR and the listener's left ear are simultaneously reached and canceled. Similarly, when the crosstalk cancellation signal -kL of the right speaker is delayed as it is emitted (a + aa) backward from the R playback position, the crosstalk KL of the left speaker and the listener's right ear are simultaneously reached and canceled. The
The left and right audio outputs of the crosstalk cancellation circuit that delays the crosstalk cancellation signals −kR and −kL with respect to the L and R reproduced sounds shown in FIG. 2 are amplified and arranged at equal angles from the listener to the left and right. A device that reproduces with a stereo speaker is a current crosstalk canceling system (hereinafter also referred to in the present specification). In many cases, the current crosstalk cancellation system is widely sold in combination with a sound field correction technique for providing an input signal with a filter characteristic based on the head related transfer function (HRTF). When the talk cancellation system allows the left channel to be listened only to the left ear and the R channel to be listened only to the right ear “like a headphone”, this is described in patent documents and introductory texts to users.

In FIG. 2, if the delay corresponding to the distance difference (a + aa) is always satisfied according to the angle at which the listener looks at the left and right speakers, both the −kR playback position and the R playback position after (a + aa) from the L playback position When the left and right speakers are viewed at a certain angle or more with the left ear b as the center, the left circle is on the eccentric circle B (hereinafter also referred to as the same in this specification) that swells to the left. Similarly, the -kL playback position and the L playback position that are (a + aa) backward from the R playback position are both centered on the listener's right ear c, and when the left and right speakers are viewed at a certain angle or more, the eccentric circle C (current In the specification, the same applies to the following).
Note that the bulges of the eccentric circles B and C shown in the plan view are abbreviated and are different from actual bulges having a large individual difference. In addition, in the plan view, only the reproduced sounds of the four paths 1) to 4) are entered in a straight line with an arrow, and this is followed in the following figures for comparison.
1) L of the left speaker playback sound.
2) Crosstalk KL heard by the right ear with the filter characteristic K.
3) Crosstalk cancellation signal -kL of right speaker playback sound.
4) Crosstalk of a crosstalk cancellation signal (hereinafter referred to as “secondary crosstalk”) −kKL that is heard by the left ear with the filter characteristic K.
In FIG. 2, the above four paths are the crosstalk KL emitted from the left speaker and the crosstalk cancellation signal −kL emitted from the right speaker (a + aa) at the same time to reach the listener's right ear. It passes the left ear by (a + aa), and the secondary crosstalk -kKL requires a distance difference (a + aa), indicating that it has reached (a + aa) from the left ear.

  In the current crosstalk cancellation system, the filter characteristic k of the crosstalk cancellation signal is usually band-limited by the crosstalk filter characteristic K. This is better offset between crosstalk and crosstalk cancellation signals of opposite phase, even if there is a difference in arrival time for low-pitched sounds with longer wavelengths. This is because it is more practical to limit the high frequency range of the crosstalk cancellation signal than K in order not to limit it extremely.

On the other hand, although crosstalk occurs between the stereo speaker and the listener's ears, practical crosstalk cancellation technology that can be implemented by the speaker system itself without depending on external devices has not been generalized.
FIG. 3 shows the configuration and the relationship with the listener when the function of the current crosstalk cancellation system is replaced with a stereo speaker system. As shown in FIG. 3, when the function of the current crosstalk cancellation system is replaced with a stereo speaker system, the following configuration is required.
That is, in the left speaker system 3, the band is limited relative to the left main unit 1 by at least one speaker unit (hereinafter referred to as a left main unit) 1 for reproducing L and R by a filter unit. At least one speaker unit (hereinafter referred to as a left sub-unit in this specification) 1A that attenuates and reproduces in reverse phase is required. In the right speaker system 4, at least one speaker unit (hereinafter referred to as a right main unit) 2 that reproduces R and L is a band relative to the right main unit 2 by a filter unit. It requires at least one speaker unit (hereinafter referred to as a right sub-unit in this specification) 2A that restricts, attenuates and reproduces in reverse phase.
Ideally, the characteristics of the filter section are variable, but even if fixed, it has a certain effect. However, the left and right main and sub units require an appropriate front / rear distance difference with respect to the listener's ears according to the angle at which the listener looks at the left and right speaker systems 3 and 4. The main unit and the right sub unit should be located on the eccentric circle C, and the right main unit and the left sub unit should be located on the eccentric circle B.

  The left and right main and sub unit arrangements in FIG. 3 in which the functions of the current crosstalk cancellation system are replaced with four units are referred to as CCS arrangements (hereinafter also referred to as the same in this specification). In the CCS arrangement, the filter characteristic k of the sub unit can be easily satisfied by a general speaker manufacturing technique such as a network or an attenuator. However, it is difficult to appropriately retract the sub unit by (a + aa) with respect to the main unit for each angle at which the left and right speakers are viewed.

The prior art that always satisfies the left and right speakers when the left and right speakers are expected to be approximately 50 ° or less of the distance difference (a + aa) due to the unit arrangement of the left and right speaker systems has been disclosed.
As shown in FIG. 4, according to the above technique, a straight line parallel to a straight line connecting the listener's ears (hereinafter, the straight line is referred to as a reference line in this specification), that is, a reference line D1 from the listener. The left main unit 1 is disposed at the left p of the left and right equidistant and equal angles, and the right main unit 2 is disposed at the right q. Further, a left sub unit 1A is arranged at o on the left from the left main unit 1 by the listener binaural interval E, and a right sub unit 2A is arranged at r to the right from the right main unit 2 by the listener binaural interval E. Is done.
At this time, since the first parallelogram bcpo and the second parallelogram bcrq are congruent, ob = qb = pc = rc.
This relationship does not change even if the reference line D1 is moved up and down in the range where the eccentric circles B and C do not bulge, ob and qb become the radius of the perfect circle B centered on the listener's left ear b, and pc And rc are the radii of the perfect circle C centered on the listener's right ear c. Therefore, the crosstalk KL of the left main unit on the perfect circle C and the crosstalk cancel signal −kL of the right sub unit always reach the right ear c at the same time, and both cross the right main unit on the perfect circle B. The talk KR and the crosstalk cancel signal -kR of the left sub unit always reach the left ear b at the same time.
In the present specification, the above 4-unit arrangement is hereinafter referred to as an SDA arrangement.
In the SDA arrangement, the perfect circle B is replaced with the eccentric circle B, and the exact circle C is replaced with the eccentric circle C, and the left main unit and the right are at the intersection of the reference line D1 and the eccentric circle C on the left and right. If the sub unit is positioned and the right main unit and the left sub unit are positioned at the right and left intersections of the reference line D1 and the eccentric circle B, the distance difference (a + aa) is satisfied and an arbitrary angle is satisfied. Will cancel the crosstalk.

Although not shown in the figure, when the reference line D1 is moved away from the listener in the above SDA arrangement, the left and right main units overlap at the intersection of circles B and C, and at this time, the left of the binaural interval from the intersection to the left The sub unit is located, and the right sub unit is located to the right of the binaural distance from the intersection.
With an approximate unit arrangement, (L + R) is reproduced with two units stacked one above the other, left difference signal 2 (LR) is reproduced with the left sub unit, and right difference signal 2 ( It is known to those skilled in the art that there is a so-called “matrix speaker” that reproduces (LR), and that a “surround matrix” provides a certain surround effect.

On the other hand, current crosstalk cancellation systems involve secondary crosstalk, just as stereo speakers involve crosstalk.
In the current crosstalk cancellation system of FIG. 2, the secondary crosstalk is delayed by (a + aa) from the left speaker L. -kR requires a further distance difference (a + aa), and the head shielding effect further filters K. -KKR is heard by the right ear with characteristics, and is delayed by (a + aa) from R of the right speaker. -KL requires a further distance difference (a + aa), and the head shielding effect further increases K minutes. -KKL with a filter characteristic and heard by the left ear.
That is, in the current crosstalk cancellation system, the secondary crosstalk reaches the left and right ears with a delay of 2 (a + aa) from L and R by a distance difference.

Secondary crosstalk occurs as -kKR reaching the listener right ear from the left sub-unit and -kKL reaching from the right sub-unit to the listener left ear even in the SDA arrangement of FIG.
In the SDA arrangement, -kR and -kL are outside the binaural distance on the reference line from the L and R reproduction positions, and the amount of delay of the secondary crosstalk becomes larger than that in the CCS arrangement. This difference is assumed to be S (same notation in this specification), and even in the SDA arrangement, the left and right main and sub units are arranged at the intersection of the reference line and the eccentric circles B and C to satisfy the distance difference (a + aa). Then, in the SDA arrangement, the secondary crosstalk reaches the left and right ears with a distance difference of 2 (a + aa) + S from L and R.

However, the current crosstalk cancellation system in FIG. 2 and the SDA arrangement in FIG. 4 both have L on the left side intersection CL with the eccentric circle C on the straight line connecting the listener's ears b and c. -KR is reproduced at the left intersection BL, R is reproduced at the right intersection BR with the eccentric circle B, and -kL is reproduced at the right intersection CR with the eccentric circle C.
At this time, a becomes the interval between both ears E and S becomes 0, and the SDA arrangement is equal to the current crosstalk cancellation system.
At this time, if the radius not including the bulges of the eccentric circles B and C is set to the distance between both ears E, both the current crosstalk cancellation system and the SDA arrangement reproduce L at the left ear position b, and the L reproduced sound is Crosstalk that goes around the head and reaches the right ear c is reproduced at the right distance from the right ear c by -kL. Similarly, R is reproduced at the right ear position c, and the crosstalk in which the R reproduction sound goes around the head and reaches the left ear b is reproduced at the same distance left from the left ear b.
That is, both have the same structure as the sealed headphones. However, at this time, the delay time of the secondary crosstalk also becomes the maximum at a distance difference 2 (E + aa) from L and R.

That is, on the straight line bc, synonymous with the headphone structure, L and -kR reproduced by CL and BL, and R and -kL reproduced by BR and CR, the current crosstalk cancellation system and SDA arrangement are as follows: It can be regarded as playing in front of the listener.
That is, the current crosstalk cancellation system reproduces L on the straight line starting from the left ear b, reproduces -kR behind it, reproduces R on the straight line starting from the right ear c, and -KL is reproduced, and the left and right speakers can be expected to be 180 ° or less by the structure in which the two straight lines are rotated around both ears b and c.
Further, the SDA arrangement allows the left and right speakers to be expected to be approximately 50 ° or less so that the distance difference aa does not increase by the structure in which the reference line D1 is translated forward.
S is caused by the difference in the structure.

In crosstalk, the smaller the left and right speakers are, the smaller the playback sound field is on a straight line connecting the left and right speakers. On the other hand, in secondary crosstalk, in the current crosstalk cancellation system, the reproduction sound field is unrealistically expanded only on the rear side of the left and right speakers as the left and right speakers are viewed at a larger angle.
In the above SDA arrangement, the sound field expands unrealistically only to the left and right on the straight line connecting the left and right speakers. The above difference in sound field deformation can be easily confirmed by arranging four small speakers according to FIGS. 2 and 4 and performing stereo reproduction.

Therefore, in order to prevent the sound field from being distorted, a technology for reducing the influence of secondary crosstalk is used in combination with the current crosstalk cancellation system.
For example, in the crosstalk cancellation system, the secondary crosstalk (−kKL and −kKR) is further canceled by the secondary crosstalk cancellation signals (kkL and kkR) of the opposite speaker, in which case the secondary crosstalk cancellation signal is Further, the third-order crosstalk (kkKL and kkKR) heard by the opposite ear is canceled by the third-order crosstalk cancellation signals (-kkkl and -kkR) from the opposite speaker. There is a technique in which a cancel signal is calculated and generated, and is played back by left and right speakers one after another for a certain amount of time.
Currently, a sound field correction technique is generally used in which an input signal is given a filter characteristic based on a head related transfer function (HRTF). The above technique is similar to the technique of making the sound field that becomes narrow with the localization in the head and close to binaural, and obtains a sense of sound field surrounding the listener, which is lost by secondary crosstalk.
A technique in which the sound field correction technique based on the head related transfer function (HRTF) is used in combination with the crosstalk cancellation technique is generally called a virtual surround technique. The system using the virtual surround technology reproduces a sound field including a sense of siege with two front speakers. However, the vast majority of manufacturers who sell the system have explained to users that a more realistic surround sound field can be obtained with a multi-channel surround stereo system using rear speakers.
In addition, there are many users who feel that the playback sound of a properly installed 2-channel stereo system is more realistic than the virtual surround system described above. Is used.

Patent Document 1 is an example in which crosstalk cancellation is described as “providing an effect as if listening with headphones”.
Patent Document 2 is a prior art in the 1980s in which a difference signal component is reproduced by left and right sub-units in a speaker arranged in SDA.
Patent Literature 3 is a technology that was patented in 2007, which uses the technology of Patent Literature 2 together with the sound field correction technology based on the head related transfer function (HRTF).
Patent Document 4 is a technique in which high-order crosstalk that occurs at the time of crosstalk cancellation is canceled one after another for a certain period of time.
Non-Patent Document 1 is a speaker system using a technology based on Patent Document 2 that was sold in the 1980s. In this product description, the L and R playback sounds are expressed as “headphones (like headphone) L channel is heard in the left ear and R channel is heard only in the right ear” due to the crosstalk cancellation effect. .
Non-Patent Document 2 is an example of a so-called “matrix speaker” produced by the late audio critic Tetsuo Nagaoka.
Japanese Laid-Open Patent Publication No. 2007-202139 “Sound Image Localization Device” (JP.A) Description 0003 US Pat. No. 4,489,432 “METHOD AND APPARATUS FOR REPRODUCING SOUND HAVING A REALISTIC AMBIENT FIELD AND ACOUSTIC IMAGE” Special Table 2007-510334 (P2007-510334A) Public Relations (JP.A) "Multi-channel audio surround sound system from loudspeakers placed in front" Japanese Patent Application Laid-Open No. 11-187497 (JP.A) "Sound Image Sound Field Control Device" Claim 1 Catalog of SDA1B and SDA-SRS in the US Polk Audio homepage http: // www. polkaaudio. com / search / older. php Ongaku no Tomosha “Tetsuo Nagaoka's original loudspeaker design 4 Such a speaker has never seen Drawing Collection 3” 2002 P176-P177

The current crosstalk cancellation system causes secondary crosstalk to distort the sound field, and when the listener moves to the left or right, the sound field is greatly swung to the opposite side of the moving direction with a sense of reverse phase. Secondary crosstalk also occurs in a crosstalk cancellation system using three or more speakers, although the delay times for the L and R reproduced sounds are different. Therefore, the crosstalk cancellation technique is usually used in combination with a sound field correction technique that gives the input signal a filter characteristic based on the head related transfer function (HRTF). Technology is used together.
However, as most manufacturers recognize, when compared to the sound field obtained by the so-called virtual surround technology, which uses the above-mentioned sound field correction technology together with the crosstalk cancellation technology, the surround sound field uses multi-channel surround that uses rear speakers. The system is more realistic, and in stereo reproduction, a good quality stereo speaker system is more realistic.
In other words, the crosstalk cancellation technology originally intended to reduce the crosstalk of stereo speakers can be used as a general stereo speaker by reducing crosstalk, either as a stand-alone technology or in combination with other technologies. It is hard to say that it has a clear advantage.
This is because of secondary crosstalk caused by crosstalk cancellation, and the sound field distorted by secondary crosstalk is not sufficiently corrected by the sound field correction technology currently used in crosstalk cancellation technology. is there.
However, crosstalk originally occurs between stereo speakers and listeners' ears. Therefore, it is clear that the stereo speaker itself has an effect of reducing the crosstalk, and at that time, if the secondary crosstalk is not accompanied, the speaker has an advantage over the speaker accompanied by the crosstalk.
Note that if the crosstalk cancellation technology is established without secondary crosstalk in stereo speakers, a crosstalk cancellation effect that does not cause any contradiction can be obtained even when implemented by other methods.

(Causes of secondary crosstalk 1)
The cause of the occurrence of secondary crosstalk can be explained as follows in the speaker system of the SDA arrangement shown in FIG.
The perfect circles B and C are hereinafter simply referred to as circles B and C.
In the SDA arrangement shown in FIG. 4, the right main unit and the left sub unit are at the right intersection q and the left intersection o of the reference line D1 and the circle B, and the left main unit and the right sub unit are the reference The line D1 and the circle C are at the intersection point p on the left side and the intersection point r on the right side.
Changing the angle at which the listener looks at the left and right speaker systems 3 and 4 moves the reference line back and forth, and keeps the distance between the left and right main and sub-units at the distance between both ears E and 2 with the circles B and C, respectively. Equivalent to changing the angle at which two intersections are expected. Therefore, the reference line D1 is a reference line D2 that is in contact with the circle B in front of the listener's left ear and in contact with the circle C in front of the listener's right ear.
At this time, the left main unit and the right sub-unit both represent the virtual left unit VsL in front of the right ear while maintaining the condition on the circle C (the virtual left unit of the SDA arrangement is represented in the same manner in the present specification). ), And the right main unit and the left sub unit are both on the circle B while maintaining the condition of the virtual right unit VsR in front of the left ear (the virtual right unit of the SDA arrangement is referred to in this specification). Overlapping as notation).

That is, the left main unit (L) and the right sub unit (−kL) on an arbitrary reference line D1 reproduce L in front of the listener's right ear, and VsL reversed left and right with respect to both ears is used for crosstalk cancellation. Thus, it can be considered that the left unit for reproducing L and the right unit for reproducing -kL are equally divided on the left and right on the circle C.
Accordingly, when VsL itself reproduces L, secondary crosstalk does not occur, but the left main unit and the right sub unit equally divided on the left and right with respect to the right ear on the circle C, the right sub unit for the left ear After the unit is 2a + S further away, the left main unit L reaches the left ear, and then the right sub-unit −kL is delayed by the distance difference 2a + S and reaches the left ear, ie, second order crosstalk, −kKL.
Similarly, the right main unit (R) and the left sub unit (−kR) on an arbitrary reference line D1 use VsR that is reversed left and right with respect to both ears that reproduce R in front of the listener's left ear, and perform crosstalk cancellation. For the purpose, it can be considered that the right unit for reproducing R and the left unit for reproducing -kR are equally divided on the left and right on the circle B.
Therefore, secondary crosstalk does not occur when VsR itself reproduces R, but the left sub-unit and the right main unit equally divided on the left and right with respect to the left ear on the circle B are the left sub-unit with respect to the right ear. After the unit is 2a + S further away and R of the right main unit reaches the right ear, -kR of the left sub unit is delayed by the distance difference 2a + S, and becomes the second order crosstalk, -kKR, which reaches the right ear.

(Causes of secondary crosstalk 2)
On the other hand, the cause of secondary crosstalk in the current crosstalk cancellation system can be explained as follows.
The speaker system having the CCS arrangement shown in FIG. 3 in which the reproduction sound of the current crosstalk cancellation system is replaced with four units will be described.
As the left main unit with the CCS arrangement is arranged on the eccentric circle C and the right main unit on the eccentric circle B by equal angles, the angle expected by the listener is reduced, and the two overlap at the intersection of the eccentric circles B and C. At this time, the rear left sub unit on the straight line passing through the left ear b and the left main unit, and the rear right sub unit on the straight line passing through the right ear c and the right main unit also overlap at the intersection of the eccentric circles B and C. .
This overlaps as a virtual left unit VcL (a virtual left unit of CCS arrangement is expressed in the following description in the present specification) while maintaining the condition that both the left main unit and the right sub unit are on the eccentric circle C, and While maintaining the condition that the right main unit and the left sub unit are both on the eccentric circle B, it can be regarded as overlapping as a virtual right unit VcR (a virtual right unit of CCS arrangement is expressed in the same manner hereinafter). .
That is, the CCS arrangement is divided into a virtual left unit VcL that reproduces L at the intersection of the eccentric circles B and C in front of the listener, a left unit that reproduces L for the purpose of crosstalk cancellation, and a right unit that reproduces -kL. The virtual right unit VcR that divides on the circle C and reproduces R is divided on the eccentric circle B into a right unit that reproduces R and a left unit that reproduces -kR for the purpose of crosstalk cancellation. Become.

The VcL and VcR on the front of the listener are also reversed left and right with respect to both ears during actual use. That is, in order to divide VcL and VcR into left and right units, the left main unit is expanded on the eccentric circle C and the right main unit is expanded on the eccentric circle B at equal left and right angles. At this time, since the centers of the eccentric circles B and C are deviated from both ears, VcL and VcR are shifted as follows.
1) On the straight line connecting the right ear c and the right main unit on the eccentric circle B, the right sub-unit position at the intersection with the eccentric circle C moves larger on the eccentric circle C than the left and right main unit positions. Therefore, as the left and right main units are distributed to the left and right, VcL, which is the midpoint between the left main unit and the right sub unit, on the eccentric circle C shifts to the right.
2) On the straight line connecting the left ear b and the left main unit on the eccentric circle C, the left sub-unit position at the intersection with the eccentric circle B moves on the eccentric circle B larger than the left and right main unit positions. Therefore, as the left and right main units are distributed to the left and right, VcR, which is the midpoint between the right main unit and the left sub unit, on the eccentric circle B shifts to the left.
VcL itself that is reversed left and right with both ears reproducing L on the right side of the listener does not cause secondary crosstalk. However, for the left main unit and the right sub unit obtained by equally dividing VcL on the eccentric circle C, that is, right and left with respect to the right ear c, the right sub unit for the left ear b is approximately 2 (a + aa) than the left main unit. ), And after reaching L, -kL is delayed by a distance difference of 2 (a + aa) to become the second order crosstalk reaching the left ear, -kKL.
Also, VcR itself that is reversed left and right with both ears playing R on the left side of the listener does not cause secondary crosstalk. However, for the right main unit and the left sub unit obtained by equally dividing VcR on the eccentric circle B, ie, the left ear b, the left sub unit for the right ear c is approximately 2 (a + aa) than the right main unit. ), And after reaching R, -kR is delayed by a distance difference of 2 (a + aa) to become the second order crosstalk reaching the right ear, -kKR.
The left and right sub-units that are the -kR and -kL playback positions of the CCS arrangement are actual positions, whereas the -kR and -kL playback positions ii and jj of the current crosstalk cancellation system are virtual positions behind the speakers. For this reason, in the current crosstalk cancellation system, the delay time of the secondary crosstalk for L and R is 2 (a + aa).

In both the SDA arrangement of FIG. 4 and the CCS arrangement of FIG. 3, the virtual left unit that reproduces L in front of the listener is the left main unit that reproduces L and the right that reproduces −kL on the eccentric circle C. Divided into sub-units. Further, the virtual right unit that reproduces R in front of the listener is divided on the eccentric circle B into a right main unit that reproduces R and a left sub-unit that reproduces -kR.
At this time, -kL and -kR are vector-synthesized with L and R to create surround localization, so the stereo sound field is not reversed left and right. However, the phase difference information is confused to cause secondary crosstalk, and when the listener moves to the left or right, a phenomenon occurs in which the sound field is greatly shaken in the direction opposite to the moving direction.
As will be appreciated by those skilled in the art, the above phenomenon is the largest at the position where the listener has moved to the left and right by about half of the distance between both ears. It is the front of the unit XcL or the virtual right unit XcR. That is, when the listener moves to one side, the audible volume of the near-side main unit increases, but on the front side of the XcL and XcR, the listener is equivalent to the front side of the opposite unit, and the left and right volume levels are equal. The difference information contradicts the left and right phase difference information, and the sound image and the sound field are swung in opposite directions.
The above contradiction similarly occurs in current crosstalk cancellation systems that are approximately equal to SDA, CCS, and CCS deployments.

(Contradiction of conventional technology)
The contradiction in which the virtual unit is reversed left and right with respect to both ears, resulting in secondary crosstalk, is that the current crosstalk cancellation system technology "like headphones", L channel only to the listener's left ear, and R channel This is caused by the principle of listening to the listener's right ear only.
The expression “like headphones” has been used by many manufacturers since the 1980s at the latest. However, as it is expressed, the crosstalk cancellation system is a stereo recording source other than binaural recording, which is equivalent to biphonic recording and can only obtain a narrow sound field with in-head localization. However, current crosstalk cancellation systems do not reproduce a “headphone-like” sound field.

When a cone-type speaker unit having a large diameter is reproduced without using a baffle, a mid to low frequency range of about 500 to 1 kHz or less (hereinafter referred to as a frequency range of about 500 to 1 kHz or less in this specification) according to directivity. In the middle / low range, the frequency range of about 500 to 1 kHz or higher is referred to as the middle range or the middle / high range), and the sound pressure level partially decreases, but greatly decreases as a whole.
The above phenomenon is the same even if the speakers A and B having the enclosure are brought close to each other and reproduced in reverse phase. Further, when the normal phase sound or the reverse phase sound is reproduced by the speaker C having an enclosure at a certain distance from the speakers A and B, the speaker C influences the speakers A and B regardless of the listening point. Play the mid-low range well without receiving it.
That is, the reverse phase sound cancels out in a self-contained manner between the adjacent speakers regardless of the difference in the arrival time in the low frequency range having a longer wavelength, and either the normal phase sound or the reverse phase sound is heard between the speakers A and B. If it does not become excessive, it does not affect the distant speaker C playback sound.
Therefore, in both the CCS arrangement of FIG. 3 and the SDA arrangement of FIG. 4, the monaural components common to the L and R in the mid-low range are canceled between the main and sub units that are close to the left and right, respectively, and the distance between the listener's ears is reduced. Regardless, the crosstalk of the mid-low range monaural component of the opposite speaker is not canceled.

As an example, a description will be given of the mid-low range when the mid-low range level of the left and right sub units is set to -3 dB with respect to the left and right main units of the CCS arrangement or the SDA arrangement.
In this case, the monaural component common to L and -kR is canceled between the left main unit and the sub unit, -kR of the left sub unit loses the monaural component, and the monaural component of the left main unit L also becomes -3 dB. Also, the monaural component common to R and −kL is canceled between the right main unit and the sub unit, −kL of the right sub unit loses the monaural component, and the monaural component of the right main unit R also becomes −3 dB.
However, the monophonic component of the right main unit rises from -3 dB to 0 dB due to crosstalk and is centered, and R other than the monaural component is canceled by crosstalk, and the audible level remains at 0 dB. The monaural component of the left main unit rises from -3 dB to 0 dB due to crosstalk and is centered, and L other than the monaural component is crosstalk canceled and the audibility level remains 0 dB.
If the middle / low range level of the left and right sub-units is set to -3 dB to 0 dB with respect to the main unit, the monaural component is excessively offset between the adjacent main and sub-units, and the audibility level decreases. The unit monaural component is lost, and the monaural component of the main unit is centered between the speakers. Even if the mid-low range level of the left and right sub units is -3 dB or less, the left and right sub unit monaural components are still lost, and the monaural components of the left and right main units are centered between the speakers.
In other words, the current crosstalk cancellation system produces crosstalk regardless of whether a general stereo speaker is a monaural component or other components in the mid-low range, and increases the audibility level by about 3 dB, while lowering the mid-low range level. The L and R other than the monaural component can be canceled by crosstalk, but the monaural component is centered between the speakers with crosstalk, that is, speaker position information.

It is the same with the current crosstalk cancellation circuit that the monaural component in the mid-low range is centered between the speakers.
In the middle / low range, when kR is subtracted from the left audio signal L and kL is subtracted from the right audio signal R, the waveform becomes almost reverse even if one is delayed by about 1 millisecond (ms), and the monaural component is It cancels out, the monaural component of the crosstalk cancellation signal is lost, and the monaural component in the L, R mid-low range is also attenuated. Accordingly, the monaural component of the left and right speaker reproduction sounds L and R decreases in level according to k, but is localized between the left and right speakers.
Note that in a circuit using an L, R difference signal as a crosstalk cancellation signal, the monaural component is excluded according to the L, R distribution of the difference signal, so that the monaural component is also centered between the speakers.

On the other hand, in a general stereo speaker, the crosstalk is reduced in the mid-high range according to the angle at which the listener looks at the left and right speakers, and surround playback is also possible. That is, crosstalk can be reduced to such an extent that it is difficult to accompany speaker position information.
However, this is not related to whether the monaural component is common to L and R or not.
Therefore, in general stereo speakers, the central localization according to the distance between the listener and the speaker is mainly obtained by crosstalk in the L, R mid-low range, which is difficult to reduce even when looking at the left and right speakers at a large angle. The stereo localization that can be accompanied by the surround sound field is obtained by the L, R middle and high sounds with little crosstalk, based on the central localization in the middle and low frequencies.

In other words, there is only a relative difference between the current crosstalk cancellation system and a general stereo system except for the crosstalk amount in the L and R middle and low frequencies other than the monaural component. Actually, the central vocals recorded on-mic both are localized between the left and right speakers, and the current crosstalk cancellation system is also based on the principle of stereophonic as in a general stereo system.
Therefore, the current crosstalk cancellation system has the same structure as that of the left and right virtual units that are reversed to the left and right, which has the same effect as headphones, on the straight line connecting both ears as described above. Contrary to the stereophonic principle of listening to the left and right speakers with both ears, phase confusion occurs, and a realistic reproduction sound cannot be obtained with the crosstalk cancellation technique alone.
In addition, since the current crosstalk cancellation system conforms to stereophonic and the central vocal recorded on-mic does not localize in the head, there is no necessity to replace the sound field equivalent to biphonic with the sound field equivalent to binaural. Therefore, the sound field correction technique that gives the input signal a filter characteristic based on the head related transfer function (HRTF) is a technique that further deforms the reproduced sound field deformed by the second-order crosstalk distortion so as to surround the listener. Therefore, the above contradiction of the current crosstalk cancellation system is not resolved.

(Task)
Considering the ideal of stereo playback, the stereo speaker becomes a signal converter equivalent to a pair microphone (this is hereinafter referred to as “a binaural pair microphone”) installed in the listener's binaural interval, Since the microphone and speaker are offset in terms of hearing, the listener can listen to the recorded sound with his / her ears, thus making the sound image and sound field of the recording site realistic including the surround information. it is conceivable that.
However, in stereophonic, as described above, at least the monaural component in the mid-low range is centered between the left and right speakers. For this reason, if the microphone and speaker are offset in the sense of hearing, the sound image and sound field of the recording site recognized by the listener will move away from the listener according to the distance to the speaker, and if the listener does not approach the speaker, the central vocal will be in front of you. There is no localization.

However, the current stereo speakers have little crosstalk in the mid-high range when properly installed, and include a certain surround sound field, but in the mid-low range, the playback sound field is on a straight line connecting the left and right speakers due to crosstalk. It is distorted, the transient characteristics are reduced, the size of the instrument and the realistic sound field are not reproduced, and it cannot be said that it is equivalent to a pair microphone installed at the listener's binaural interval.
On the other hand, the current crosstalk cancellation system also causes secondary crosstalk, and the sound image and sound field are unrealistically distorted, contradicting the binaural pair microphone. Also, the sound field correction technology currently used in conjunction with the crosstalk cancellation technology does not resolve the above contradiction.
Therefore, the present invention is a stereo speaker system using a general speaker unit, which is close to a signal converter equivalent to the paired microphone with the above-mentioned binaural spacing, and centered on a mid-low range where the crosstalk level is high in the stereo speaker. By reducing crosstalk other than monaural components common to L and R without secondary crosstalk, the stereo sound field including the surround sound field can be reproduced well, and driven by a general stereo amplifier. The challenge is to realize a crosstalk-cancelling stereo speaker system that can be widely commercialized by allowing various speaker unit arrangements according to the installation angle of the left and right speakers, and the mid-low range of general stereo speakers. Reduces crosstalk at the center and does not cause secondary crosstalk. It was a challenge to provide a click Cancel method.

In order to solve the above problems, the first invention is a stereo speaker system,
The left speaker system includes at least one speaker unit (left main unit) to which a left audio output L is input, at least one speaker unit (left sub unit) to which a right audio output R is input, and the left sub unit. The left main unit reproduces the left audio output L, the left filter unit restricts the reproduction band and volume, and the left sub unit outputs the right audio output. R is reverse phase regenerated,
In addition, the right speaker system includes at least one speaker unit (right main unit) to which the right audio output R is input, at least one speaker unit (right sub-unit) to which the left audio output L is input, and the right A right filter unit connected to the sub unit, the right main unit reproduces the right audio output R, the right filter unit restricts the reproduction band and volume, and the right sub unit reproduces the left audio output R. Play audio output L in reverse phase,
When the left and right speaker systems are equidistant to the left and right and at the same angle, the left main unit, the left sub unit, the right sub unit, and the right main unit are arranged on a straight line parallel to the straight line connecting the listener's ears. However, if the distance from the listener's both ears is kept approximately when placed on the straight line, the height difference between the left main unit and the left sub unit, and between the right main unit and the right sub unit Allow,
In the left speaker system, the left sub-unit is provided on the right side at a distance substantially equal to the listener's binaural interval with respect to the left main unit,
In the right speaker system, the right sub unit is provided on the left side at a distance approximately equal to the listener's binaural interval with respect to the right main unit.
When secondary crosstalk is canceled with the above configuration, even if the crosstalk cancellation signal of the right sub-unit reaches the listener's right ear before the crosstalk of the left main unit playback sound, no reverse phase sensation will occur and it will be good The right sub-unit playback sound is limited by the right filter unit within the canceled playback band and volume, and the left sub-unit crosstalk cancellation signal precedes the listener's left ear than the right main unit playback sound crosstalk. A crosstalk cancellation stereo speaker system characterized in that the left sub-unit reproduction sound is limited by the left filter unit within a reproduction band and volume that do not cause a reverse phase feeling and can be satisfactorily canceled even if they reach the sound. It is.

  The second invention is the crosstalk cancellation stereo speaker system according to item 0035, wherein the left sub unit is provided to the right with respect to the left main unit at an arbitrary distance g of 0 to 150% of the listener binaural interval. The crosstalk canceling stereo speaker system is characterized in that the right sub unit is provided on the left side at an arbitrary distance g of 0 to 150% of the listener binaural interval with respect to the right main unit.

The third invention is a crosstalk canceling method, which reproduces the left audio output L with a left speaker system and is arranged at an equal angle with respect to a listener in an arbitrary stereo speaker system, and a right speaker. Play the right audio output R on the system,
In the left sub-speaker comprising at least one speaker unit and a left filter unit connected to the speaker unit, the right audio output R is reproduced in reverse phase by limiting the reproduction band and volume by the left filter unit. In the right sub-speaker including at least one speaker unit and a right filter unit connected to the speaker unit, the left audio output L is limited in phase by limiting the reproduction band and volume by the right filter unit. Play and
The unit of the left speaker system that mainly reproduces the playback band of the left sub-speaker is the left speaker position (hereinafter referred to in the same manner), and the unit of the right speaker system that mainly plays the playback band of the right sub-speaker Is the right speaker position (same in this section),
Place the left speaker position, left sub-speaker, right sub-speaker, and right speaker position on a straight line. Allow height difference between speakers and between right speaker position and right sub speaker,
The left sub-speaker is arranged to the right at an arbitrary distance g from 0 to 150% of the listener binaural distance with respect to the left speaker position,
In addition, the right sub-speaker is arranged at an arbitrary distance g, 0 to 150% of the listener binaural distance with respect to the right speaker position,
When secondary crosstalk is canceled with the above configuration, even if the crosstalk cancellation signal of the right sub-speaker reaches the listener's right ear before the crosstalk of the left speaker system playback sound, no reverse phase sensation will occur and it will be good The right sub-speaker playback sound is limited by the right filter unit within the canceled playback band and volume, and the cross-talk cancellation signal of the left sub-speaker precedes the listener's left ear from the cross-talk of the right speaker system playback sound. Even if it reaches, the left sub-speaker playback sound is limited by the left filter unit within the playback band and volume that do not cause reverse-phase feeling and cancel well, and crosstalk cancellation without secondary crosstalk This is a crosstalk canceling method characterized by:

  The fourth invention is the crosstalk canceling method according to item 0037, wherein the left sub-speaker is arranged on the right with a distance substantially equal to a listener binaural distance from the left speaker position, and the right sub-speaker is The crosstalk canceling method is characterized in that it is arranged on the left side at a distance substantially equal to the listener binaural interval with respect to the right speaker position.

A fifth invention is a crosstalk canceling method, wherein a distance difference measured by a straight line when a reproduced sound arrives at one ear position of a listener from left and right speakers equidistant to and equal to the listener is a variable. When the listener looks at the left and right speakers at a large angle, the distance difference added to a when one of the left and right speaker playback sounds wraps around the listener's head when the listener looks at the left and right speakers is represented by the variable aa. If
The left audio input signal L is set to W and X, the right audio input signal R is set to Y and Z, W and Z are respectively input to the delay block, and a distance difference (a + aa) with respect to X and Y is input. Delay by an arbitrary distance difference of 10% to 150%, and input X and Y to the filter block to limit the frequency band and output level relative to W and Z,
Y output from the filter block is subtracted from W output from the delay block to obtain a left audio output, and X output from the filter block is subtracted from Z output from the delay block. Right audio output,
When the left and right audio outputs are amplified and reproduced by the left and right speakers arranged at equal angles to the left and right from the listener to cancel the secondary crosstalk, the crosstalk cancellation signal X is transmitted from the W crosstalk to the right of the listener. Even if the crosstalk cancellation signal Y reaches the ear earlier than the crosstalk of the Z and reaches the listener's left ear, it does not cause a sense of reverse phase and falls within the frequency band and output level that are well offset. This is a crosstalk canceling method characterized in that the crosstalk is canceled without secondary crosstalk by limiting X and Y by the filter block.

  A sixth invention is the crosstalk cancellation method according to item 0039, wherein W and Z are input to a delay block, and an arbitrary distance of 80 to 120% corresponding to a distance difference (a + aa) with respect to X and Y This is a crosstalk canceling method characterized by delaying by the difference h.

(Crosstalk cancellation stereo speaker system)
According to the crosstalk canceling stereo speaker system of the first invention configured as described above, the left and right speaker system can be expected to be within approximately 50 ° so that the distance difference aa where one of the reproduced sounds goes around the head does not increase. Even in the angle, the first parallelogram formed by the listener left ear position, right ear position, left sub unit position and left main unit position, and listener left ear position, right ear position, right main unit position and right sub unit position The second parallelogram made in is congruent. Therefore, since the left main unit and the right sub unit are both on a circle B centered on the listener's left ear, the secondary crosstalk -kKL of the right sub unit reaches the left main unit L and the left ear simultaneously. Offset. Since the right main unit and the left sub unit are both on the circle C centered on the listener's right ear, the secondary crosstalk -kKR of the left sub unit reaches the right main unit R and the right ear simultaneously. Offset.
However, in this case, secondary crosstalk does not occur, but for the right ear, the crosstalk cancellation signal -kL of the right sub unit reaches 2a + S ahead of the crosstalk KL of the left main unit and reaches the left ear. Thus, the crosstalk cancellation signal -kR of the left sub unit reaches the 2a + S ahead of the crosstalk KR of the right main unit. Therefore, by limiting the crosstalk cancellation signal to a reproduction band and a volume that can be satisfactorily canceled without causing a reverse phase feeling even at least with the arrival error, the crosstalk cancellation stereo speaker system according to the first aspect of the present invention, Crosstalk cancellation in the mid-low range is also established without secondary crosstalk.
As described above, since the monaural components common to L and R among L and -kR and R and -kL are offset between adjacent left main and sub units and between right main and sub units, the left main Of the L playback sound of the unit and the R playback sound of the right main unit, the monaural component in the middle / low range is accompanied by crosstalk. Also, crosstalk is a problem with stereo speakers, especially in the mid-low range. Crosstalk attenuates to such an extent that it can be accompanied by a sufficient sound field.
Therefore, due to the general characteristics of the stereo speaker and the listener's both ears, and the effect of the first invention that cancels L and R crosstalk other than the mid-low range monaural component, the monaural sound image is localized between the speakers by the mid-low range crosstalk. However, the crosstalk level is generally kept low in the entire reproduction band, and secondary crosstalk does not occur.

The SDA arrangement has a configuration in which the left and right main and sub units are reversed, and the unit arrangement of the first invention inevitably involves a virtual left unit XsL that reproduces L in front of the listener's left ear. XsL centered on the left ear for the purpose of canceling the crosstalk without secondary crosstalk in the mid-low range). Equally equivalent to reproducing L with the left main unit and reproducing -kL with the right sub unit.
In addition, a virtual right unit XsR that reproduces R in front of the listener's right ear (a virtual right unit that is reversed right and left with the virtual right unit VsR arranged in the SDA, is represented in the same manner in the present specification), For the purpose of canceling crosstalk without secondary crosstalk, the right main unit reproduces R on the right main unit, and R is divided equally on the circle C with the radius from the right ear to XsR. -Equivalent to playing kR.

XsL that reproduces L in front of the listener's left ear and XsR that reproduces R in front of the listener's right ear can be said to be signal converters equivalent to the above-mentioned binaural pair microphones. This is equivalent to listening to the sound image and sound field with both ears having the same characteristics as a stereo microphone. However, according to the stereophonic principle in which the center localization between speakers is determined by crosstalk in the mid-low range, each sound source is moved away from the microphone in terms of audibility according to the distance between the listener and the speaker.
Therefore, the crosstalk cancellation stereo speaker system according to the first aspect of the invention reproduces a sound image and a sound field with a surround sound field almost in the entire reproduction band according to the recording source and the speaker arrangement, and has excellent transient characteristics in the mid-low range. Since the volume difference information and the phase difference information do not cause confusion, the surround sound field naturally surrounds the listener and is realistic, and appropriately reproduces the size of medium and large instruments.
In addition, a pair of microphones that are similar to the paired microphones between the binaural intervals, a multi-track recorded sound that is stereo-positioned mainly by L and R volume differences, and a stereo source that is analog-encoded for various multi-channel surround stereos are also three-dimensional. To play.

  The first invention is a stereo speaker system that uses a general speaker unit and can be used in a stereo amplifier. Therefore, it is excellent in versatility and can produce a product with excellent cost performance due to its simple signal path, and can pursue high sound quality and sound field reproducibility. In addition, the three-dimensional and spacious sound field originally included in the recording source in the early stage of stereo is realistically reproduced with rich nuances by improving transient characteristics.

(Crosstalk cancellation method)
The above effect is also approximate when the output of the crosstalk cancellation circuit configured by the method of the fifth invention is amplified and reproduced by a general stereo speaker arranged at the same distance to the listener and at the same angle. It becomes.
The effect of the fifth invention will be described by taking as an example the case where the delay amount of the delay block is the distance difference (a + aa).
In the crosstalk cancellation circuit, L of the left audio output L-kR is delayed by (a + aa) from -kR, and R of the right audio output R-kL is delayed by (a + aa) from -kL. Therefore, when the left and right audio outputs are amplified and played by the left and right speakers, they are played back at the right speaker position. -KL and L played back (a + aa) from the left speaker position are equidistant from the listener's left ear. The secondary crosstalk −kKL that reaches the left ear with kL reaches the left ear simultaneously with L and cancels out. Also, -kR played at the left speaker position and R played back (a + aa) from the right speaker position are equidistant from the listener's right ear, and -kR reaches the right ear-secondary crosstalk -kKR is The right ear reaches R at the same time and cancels out.
However, in this case, secondary crosstalk does not occur, but for the right ear, the crosstalk cancellation signal -kL reproduced at the right speaker position is different from the crosstalk KL reproduced (a + aa) behind the left speaker. However, 2 (a + aa) is reached first. In addition, for the left ear, the crosstalk cancellation signal −kR reproduced at the left speaker position reaches by 2 (a + aa) ahead of the crosstalk KR reproduced by (a + aa) behind the right speaker. .
Therefore, by limiting the crosstalk cancellation signal to a reproduction band and volume that can be satisfactorily canceled without causing a reverse phase feeling at least with the arrival error, the crosstalk cancellation configured by the method of the fifth invention is performed. By amplifying the circuit output and reproducing it with the stereo speaker, crosstalk cancellation is also established in the mid-low range without secondary crosstalk.
In the crosstalk cancellation circuit, the middle and low frequencies of the monaural components common to L and R are canceled in the left audio output L-kR, and the monaural components common to L and R are canceled in the right audio output R-kL. Since the sound range is canceled out, the monaural component in the mid-low range is accompanied by crosstalk. Also, crosstalk is a problem with stereo speakers, especially in the mid-low range. Crosstalk attenuates to such an extent that it can be accompanied by a sufficient sound field.
Therefore, due to the general characteristics of the stereo speaker and the listener's both ears, and the effect of the fifth invention that cancels L and R crosstalk other than the mid-low range monaural component, the monaural sound image is localized between the speakers by the mid-low range crosstalk. However, the crosstalk level is generally kept low in the entire reproduction band, and secondary crosstalk does not occur.

The output of the crosstalk cancellation circuit configured by the method of the fifth invention is amplified and reproduced by speakers at equal distances on the left and right, and the delay amount corresponding to the distance difference (a + aa) according to the angle at which the listener looks at the left and right speakers Is always satisfied. In this case, the left speaker's L playback position and the right speaker's -kL playback position are both on the eccentric circle B centering on the listener's left ear, and the right speaker is on the eccentric circle C centering on the listener's right ear. Both the R playback position of the speaker and the -kR playback position of the left speaker are present.
However, since the centers of the eccentric circles B and C are deviated from both ears, as the angle at which the listener looks at the left and right speakers is widened, the deviation from the -kR of the left speaker position on the eccentric circle C by (a + aa) later. The L playback position on the circle B moves more greatly, and the R playback position on the eccentric circle C moves more (a + aa) behind the -kL of the right speaker position on the eccentric circle B. To do. Therefore, as the angle at which the left and right speakers are viewed is widened, the midpoint of the L reproduction position on the eccentric circle B and the midpoint of the -kL reproduction position shifts to the left from the intersection of the eccentric circles B and C, and the R reproduction position on the eccentric circle C The midpoint of the -kR playback position is shifted to the right from the intersection of the eccentric circles B and C.
Accordingly, in the stereo system, the L playback position and the −kL playback position on the eccentric circle B are the virtual left unit XcL that plays L at its midpoint (virtual left unit VcL that is placed in the CCS direction and that is reversed left and right. The left unit is indicated in the same manner in this specification), and it can be said that the left unit is equally divided on the eccentric circle B for the purpose of canceling the crosstalk without the secondary crosstalk in the middle / low range. The R playback position and the -kR playback position on the circle C are the virtual right unit XcR that plays R at its midpoint (the virtual right unit that is reversed left and right with the virtual right unit VcR in the CCS arrangement is referred to as the following in this specification) Similarly, it can be said that the left and right parts are equally divided on the eccentric circle C for the purpose of canceling the crosstalk without the secondary crosstalk in the mid-low range.
XcL for reproducing L is shifted to the left from the front of the listener, and XcR for reproducing R is shifted to the right from the front of the listener. Therefore, XcL and XcR can be said to be approximate to a binaural pair microphone.
Therefore, when both are canceled out as approximate signal converters, the listener is equivalent to listening to the sound image and sound field with his / her ears having the same characteristics as a stereo microphone at the recording site. However, in accordance with the stereophonic principle that determines the central location between speakers by mid-low range crosstalk, each sound source is moved away from the microphone in terms of audibility according to the distance between the listener and the speaker.

  The crosstalk cancellation circuit of the fifth invention can be easily implemented by simply changing the block configuration of a general crosstalk cancellation circuit and adjusting the characteristics of the filter circuit. The sound field is distorted by the secondary crosstalk. Therefore, it is not necessary to process the original sound with a filter based on the head related transfer function (HRTF), and it is possible to produce a product with excellent cost performance by simplifying the circuit, and to pursue high sound quality and sound field reproducibility.

1 and 5 show one embodiment of the first invention.
FIG. 1 is a perspective view showing one embodiment of the first invention, and FIG. 5 is a plan view showing a relationship between one left and right speaker system and a listener according to the embodiment of the first invention. .
The first invention is a stereo speaker system. In one embodiment, the left speaker system 31 includes a left main unit 1, a left sub unit 1A, and a filter f11 connected to the left sub unit 1A. The right speaker system 41 includes a right main unit 2 and a filter f21 connected to the right sub unit 2A and the right sub unit 2A.
FIGS. 1 and 5 show examples of the left and right speaker systems 31 and 41 that are separated into left and right enclosures. It can be built in various audio equipment. In addition, said example is an illustration that various audio equipments are various, and does not limit various audio equipments.

In one of the above embodiments, each of the left main unit 1, the left sub unit 1A, the right main unit 2, and the right sub unit 2A has a separate signal system. Therefore, the components for each signal system are distinguished from the left main unit 1 side, the left sub unit 1A side, the right main unit 2 side, and the right sub unit 2A side (both of which are similarly described below).
At this time, the left audio output L is input to the left main unit 1 side and directly connected to the left main unit 1, and is also input to the right sub unit 2A side, and the playback band and volume are limited through the filter unit f21. Reverse phase connection is made to the right sub-unit 2A. At this time, the right audio output R is input to the right main unit 2 side and directly connected to the right main unit 2, and also input to the left sub unit 1A side, and the playback band and volume are limited through the filter unit f11. Thus, reverse phase connection is made to the left sub unit 1A.

The filter f11 does not give a sense of reverse phase even if the crosstalk cancellation signal -kR of the left sub unit 1A reaches the left ear ahead of the crosstalk KR of the right main unit 2 by at least the distance difference 2a + S. The R that is reproduced in reverse phase by the left sub-unit 1A is limited according to the angle at which the left and right speaker systems 31, 41 are expected and the following production within the reproduction band and volume that are well offset, and the filter f21. Even if the crosstalk cancellation signal -kL of the right sub-unit 2A reaches the right ear at least a distance difference 2a + S ahead of the crosstalk KL of the left main unit 1, no negative phase sensation will be generated and offset well The right sub-unit 2A performs reverse-phase playback in accordance with the angle at which the left and right speaker systems 31 and 41 are viewed and the following production intention within the playback band and volume. Limit the that L, the same main as an example, among the sub-units can be exemplified by properties obtained by using a linear low pass filter and -6dB Attenuator cutoff 300Hz to the sub unit. The above characteristics are examples of characteristics that can be set in various ways for easy understanding by those skilled in the art, and are not limited.
Note that the production intention refers to, for example, an intention to limit the listening point to have a crosstalk cancellation effect up to a relatively high frequency band, or to further limit the reproduction band or volume so as not to limit the listening point.
The filters f11 and f21 are generally composed of a low-pass circuit and an attenuator, but both are not necessarily essential. For example, when the sub unit has a low-pass characteristic due to the unit characteristic or the structure of the enclosure, a low-pass circuit is not necessarily required. Further, if the sound pressure level of the main unit substantially exceeds the sub unit due to the difference in efficiency, impedance, or the number of units between the main and sub units, an attenuator is not necessarily required.
In short, the filter unit is intended to relatively limit and attenuate the sub unit reproduced sound relative to the overall L and R reproduced sound, and therefore the filter units f11 and f21 have main and sub units. As long as the relative filter characteristics are satisfied, free characteristics and circuits may be provided. Examples of free characteristics and circuits include the bandpass characteristics and irregular slope characteristics of the left and right sub-units, a circuit that corrects this when a filter causes phase rotation, and the listener adjusts the filter characteristics and attenuation. Refers to possible means. The above is an example of free characteristics and circuits, and is not limited.
If the relative filter characteristics between the main and sub units are satisfied, in addition to the filter parts f11 and f21, one or both of the left and right main units and the left and right sub units have a low boost characteristic. It does not matter if a unique sound is created according to the speaker.
The functions of the filters f11 and f21 may be replaced with a free external device such as a sound processor or an equalizer using a multi-amplifier, and the amplifier and the external device are incorporated in the left and right speaker systems 31 and 41. There is no problem.
With the above configuration, the left main unit 1 of the left speaker system 31 reproduces L, the left sub unit 1A reproduces -kR simultaneously, the right main unit 2 of the right speaker system 41 reproduces R, and the right sub unit 2A reproduces -kL. Are played simultaneously.

When the left and right speaker systems 31 and 41 having the above configuration are equidistant to the left and right and at the same angle, the left main unit 1, the left sub unit 1A, the right sub unit 2A, and the right main unit 2 are connected to the listener's ears b. , C on the straight line parallel to the straight line connecting to the straight line D, that is, on the reference line D, in the left speaker system 31, the left sub unit 1A is connected to the left main unit 1 by the listener. In the right speaker system 41, the right sub-unit 2A is installed on the left by a distance substantially equal to the listener binaural interval E. .
In this case, as shown in FIG. 5, the first parallelogram bcpo formed by the listener left ear b, right ear c, left sub unit 1A position p, and left main unit 1 position o is composed of the listener left ear b and the right This is congruent with the second parallelogram bcrq formed by the ear c, the right main unit 2 position r, and the right sub unit position q. Accordingly, the left and right speaker systems 31 and 41 have ob = qb = pc = rc established at any angle at which the distance aa at which one of the reproduced sounds goes around the head does not become large, and is approximately within 50 °, and o, q Is on a circle B centered on the left ear b, and p and r are on a circle C centered on the right ear c. For this reason, L of the left main unit 1 and -kL of the right sub unit 2A reach the left ear b at the same time, so that secondary crosstalk does not occur, and R of the right main unit 2 and the left sub unit 1A -KR reaches the right ear c at the same time, and the secondary crosstalk is canceled out.
However, at this time, -kL of the right sub-unit 2A reaches the right ear b ahead by 2a + S with respect to L of the left main unit 1, and also to R of the right main unit 2 as described above. Thus, -kR of the left sub unit 1A reaches the left ear b by 2a + S as described above.
However, due to the filters f11 and f21, the crosstalk cancellation signal of the left and right sub-units is limited to a reproduction band and volume that can be satisfactorily canceled without causing a sense of reverse phase, at least with the arrival error.
Therefore, in one embodiment of the first invention, crosstalk cancellation is established centering on the mid-low range, and secondary crosstalk does not occur.
At this time, the monaural mid-low range common to L and R is canceled between the left and right main and sub-units that are close to each other at approximately the distance between both ears, and the monaural mid-low range is not canceled by the crosstalk. It is centered in the middle, and crosstalk is a problem with stereo speakers, especially in the mid-low range. In the mid-high range, the listener expects left and right speakers as a general characteristic of stereo speakers and listeners' ears. Depending on the angle, the crosstalk is attenuated to such an extent that the surround sound field can be sufficiently accompanied.
Therefore, the monophonic sound image is a crossover of the mid-low range by one effect of the embodiment of the first invention that cancels the general characteristics of the stereo speaker and the listener's both ears and the L, R crosstalk other than the mid-low range mono component. The talk is localized between the speakers, and the crosstalk level is kept low in the entire reproduction band, so that secondary crosstalk does not occur and a good stereo sound field including a surround sound field can be obtained.

At this time, the distance between the listener's binaural ears refers to a distance obtained by connecting the listener left ear position and the listener right ear position with a straight line, which is usually used as 17 cm. However, it is desirable that the listener is used more appropriately in consideration of the age difference, sex difference, and individual difference. For example, in products for infants, the distance between the left and right main and sub units may be set to an appropriate distance smaller than 17 cm. In the drawings of this specification, for convenience, the speaker unit position is drawn as a diaphragm position, but the speaker unit position is preferably a voice coil position generally used by those skilled in the art.
The direction of the speaker unit may be arbitrary. For example, the main and sub unit positions on the reference line D that are substantially equal to the distance between the listener's ears are assumed to be virtual sound generation positions, and the omnidirectional unit or a plurality of units. Units that are close to omnidirectional by combining can be placed.
In addition, the shape of the left and right speaker systems 31 and 41 shown in FIG. 1 is an example of a shape appropriately taking the main and sub unit intervals, and the left and right speaker systems 31 and 41 are free shapes. One or both of the left and right main and sub units may be a plurality of units including a multi-way unit.

  The SDA arrangement is such that the left and right main and sub-units are reversed. The above unit arrangement inevitably has a virtual left unit XsL that reproduces L in front of the listener left ear b, with the left ear b as the center. This is equivalent to dividing left and right equally on a circle B having a radius up to XsL and reproducing L by the left main unit 1 and reproducing -kL by the right sub unit 2A. Further, the virtual right unit XsR that reproduces R in front of the listener's right ear c is equally divided into left and right on a circle C having a radius from the right ear c to XsR, and R is reproduced by the right main unit 2. This is equivalent to the reproduction of -kR in the left sub unit 1A, and can be regarded as a signal converter equivalent to the binaural pair microphone.

On the other hand, even if the left and right speaker systems 31 and 41 are arranged close to each other, the virtual left and right units XsL and XsR remain at the distance between both ears.
As the sub-units 1A and 2A approach the opposite main units 2 and 1, not only the monaural components common to the near-side main unit and L and R are cancelled, but also the other monaural components are canceled in the same manner. Low frequency crosstalk increases. However, since the arrival error 2a + S of the crosstalk and the crosstalk cancellation signal itself is reduced, the reproduction band of the sub unit is expanded to a higher sound range, and the band where the crosstalk cancellation is established is not narrowed. Therefore, if the characteristics of the filter parts f11 and f21 are adjusted according to the angle, the sound field reproduction effect due to the crosstalk cancellation is not lost.

In one of the embodiments of the first invention shown in FIG. 5, the left and right main and sub units 1 and 1A and 2 and 2A on the reference line are approximately equal to the distance between both ears. On the other hand, the purpose is to make the left main unit 1 and the right sub unit 2A equidistant, and to make the right main unit 2 and the left sub unit 1A equidistant from the listener's right ear c. Therefore, if the distance relationship with the listener's left ear b is substantially maintained when each unit is placed on the reference line, there is no problem even if there is a height difference between the left main unit 1 and the right sub unit 2A. If the distance relationship with the listener's right ear c is substantially maintained, there may be a difference in height between the right main unit 2 and the left sub unit 1A.
That is, if the distance between the listener's both ears is substantially maintained, a difference in height between the left main unit 1 and the left sub unit 1A and between the right main unit 2 and the right sub unit 2A is allowed.
For this reason, even when a large unit is used, the left and right main, sub units, 1 and 2 are installed by placing the sub units 1A, 2A obliquely above or below the main units 1, 2. It is easy to install the relative horizontal distance between 1A, 2 and 2A at a distance approximately equal to the distance between both ears.

When the listener looks at the left and right speaker systems 31 and 41 at a distance difference aa where one of the reproduced sounds goes around the head increases, approximately at 50 ° or more, the crosstalk is delayed from the crosstalk cancellation signal, and at the same time the secondary Crosstalk is delayed from the L and R playback sounds.
Of the above, the secondary crosstalk being delayed from the L and R reproduced sounds is equivalent to the left and right sub-units 1A and 2A moving backward from the left and right main units 1 and 2 by a distance difference aa. Accordingly, if the left and right main units are also moved backward by the distance difference aa accordingly, the secondary crosstalk is not delayed with respect to the L and R reproduced sounds.
In this case, however, the crosstalk is further delayed than the crosstalk cancellation signal. For this reason, even if there is an increase in the arrival error, it is only necessary to limit the reproduction sound of the left and right sub-units to a reproduction band and sound volume that are well offset without causing a sense of reverse phase.
Therefore, in the case where the left and right speaker systems 31 and 41 are viewed at an angle of increasing aa, the listener may change the filter characteristics and take a free method of moving the left and right main units 1 and 2 backward by aa. .

It is easy for the listener to change the filter characteristics as long as the inductance of the coil can be switched or the attenuator can be operated.
Further, as an example of a method for moving the left and right main units 1 and 2 backward by aa, there is a method in which the left and right speaker systems 31 and 41 are appropriately disposed outward with respect to the reference line D. Fine adjustment is possible for the increase.
As another example, since the left and right sub-units 1A and 2A may be brought closer to the listener with respect to the reference line D by an increase in aa, aa when the left and right speaker systems 31 and 41 are viewed at a specific angle. By preparing an inclusion such as an adapter ring or sub baffle with a thickness of a minute, the left and right speaker systems 31 and 41 can be expected at a free angle within approximately 50 ° if the unit is attached without the inclusion, and the intervention By attaching the unit to the baffle through an object, it is possible to adopt a method that enables installation at a specific angle of approximately 50 ° or more.
Conversely, when the listener's binaural distance is small, for example, when the listener is a child, the left and right speaker systems 31 and 41 with the main and sub-units as the distance between both ears are installed inward with respect to the reference line. Alternatively, the left and right main units 1 and 2 may be attached to the baffle and adjusted via the inclusions. The above example is merely an example, and the fine adjustment method according to the angle at which the left and right speaker systems 31 and 41 are viewed or the distance between the listener's ears is not limited.

A good stereo sound field including a surround sound field obtained in one of the embodiments of the first invention is within a range of approximately 50 ° within which the distance difference aa in which one of the right and left reproduced sounds wraps around the head does not increase. Even if the speaker is moved back and forth in the center of the speaker, it can be obtained in the same manner. However, in one of the above embodiments, the listening area is hardly limited to the left and right.
In the current crosstalk cancellation system, the left / right phase difference information is confused as described above, and when the listener moves to the left or right, the listener is in the reverse direction on the phase in front of the virtual left / right unit VcL or VcR. In contrast, the virtual left and right units XsL and XsR in one embodiment of the first invention have the phenomenon that the reproduced sound field is greatly shaken in the direction opposite to the actual moving direction. This is because it is equivalent to a binaural pair microphone, and the reproduction band of the left and right sub-units is limited to the mid-low range, so that the reverse phase feeling and the comb filter phenomenon hardly occur between the main and sub-units.
This is similar to the case where a general stereo speaker does not limit the listening area extremely, although the comb filter phenomenon occurs when the listener moves.

The secondary crosstalk reaches both ears simultaneously with the L and R playback sounds when the listener is in the center of the speaker as described above.
Therefore, when the listener moves to the left and right, the secondary crosstalk is delayed from the L and R reproduced sounds and the sound field tends to spread outward on the line connecting the left and right speakers. In the opposite ear, the secondary crosstalk arrives earlier than the L and R reproduced sounds, and the sound field tends to be drawn to the listener side as described in the next section.
This effect balances left and right, so that when the listener moves to the left or right, the sound image and sound field are not easily masked by the near-side speaker, and the entire sound field shifts in the direction of movement, but a certain surround sound field feeling is maintained. , Resolution is difficult to decrease.
In addition, since the crosstalk cancellation effect is established even if the left and right speaker systems 31 and 41 are installed on the rear side of the listener, one embodiment of the first invention can be used as a rear left and right speaker of multi-channel surround stereo.

As described above, when the secondary crosstalk is delayed or preceded by the L and R reproduced sounds, the reproduced sound field is distorted.
The tendency is to produce a surround sound field using a so-called effector, a device called Phaser or Phaseshifter, which is used when editing a multi-channel sound source and mixes the original sound and the sound whose phase is changed. However, the general tendency with continuity can be summarized as shown in 1-1) and 1-2) below.
1-1) When L and R are reproduced by the left and right speakers, R reverse phase sound is reproduced on the line connecting the left ear and the left speaker, and L reverse phase sound is reproduced on the line connecting the right ear and the right speaker. When L and R opposite phase sounds arrive late with respect to L and R, the reproduced sound field is distorted to the rear side of the straight line connecting the left and right speakers. On the other hand, when reaching in advance, the reproduced sound field is drawn to the listener side from the straight line connecting the left and right speakers and converges between both ears, and the sense of depth is reduced around the listener.
1-2) When L and R are played back by the left and right speakers, and R reverse phase sound and L reverse phase sound are played on the line connecting the left and right speakers, the L and R reverse phase sounds arrive at a delay relative to L and R. Then, the reproduced sound field is distorted to the left and right only on the extension of the straight line connecting the left and right speakers. On the other hand, when reaching in advance, the reproduced sound field is drawn to the listener side from the straight line connecting the left and right speakers, and the left and right widths of the sound field are maintained, but the sense of depth decreases before and after the listener.
The above tendency can be easily confirmed by arranging four small speakers as described above and performing stereo reproduction. Although this is not an effect of the present invention, it is described because it can be intentionally used in the following embodiments, and is temporarily described as “a sound field enhancement effect by secondary crosstalk” (hereinafter also denoted similarly). .

Among those skilled in the art, in stereo reproduction, it was often argued that the size of medium and large instruments was reproduced according to the caliber of the woofer. On the other hand, in one of the embodiments of the first invention, L, R volume difference information and phase difference information other than the monaural component of the mid-low range, which is likely to be lost due to crosstalk in a general stereo speaker, can be obtained. In addition, unlike the current crosstalk cancellation system, the above information is consistent with the binaural pair microphones, so in the case of a good stereo recording sound, it is recorded in the source as L, R volume difference information and phase difference information in the mid-low range. Regardless of the diameter of the speaker unit, the three-dimensional size of medium and large musical instruments is clearly reproduced according to the reproduction ability of the middle and low range.
Conversely, the reason why the size of each sound source is reproduced with a large-diameter woofer of a general stereo speaker is that the large-diameter woofer has a sharp directivity in the mid-low range around 100 to 500 Hz, and between the L and R channels. This is considered to be because the volume difference information and phase difference information can be easily heard. Although the description regarding the large-diameter woofer is different from the effect of the present invention, it is described because it has significant effectiveness in conveying the effect of the present invention to those skilled in the art.

In one of the above-described embodiments shown in FIGS. 1 and 5, L difference signal audio output L-nR (n> 0) can be input in reverse phase to the left sub-unit 1A without causing any contradiction, and the right sub-unit 2A. R phase difference signal audio output R-nL (n> 0) can be input in reverse phase.
In one of the above embodiments, the L difference signal audio output L-nR is input in reverse phase, passes through the filter f11 to become k (nR-L), and is reproduced in reverse phase in the left sub-unit 1A and k (L-nR). ) Further, the R difference signal audio output R-nL is input in reverse phase, passes through the filter f21, becomes k (nL-R), and is reproduced in reverse phase by the right sub unit 2A to become k (R-nL).
That is, the left main unit 1 simultaneously reproduces L, the left sub unit 1A simultaneously reproduces k (L-nR), the right main unit 2 simultaneously reproduces R, and the right sub unit 2A simultaneously reproduces k (R-nL). .

The crosstalk canceling effect when the above L and R difference signals are used with n = 1 will be described.
Between the reproduced sound L of the left main unit 1 and the difference signal −kL of the right sub-unit 2A, both reach the left ear b at the same time, and the secondary crosstalk is canceled. However, the difference signal −kL is a monaural component. However, the mono component of L is not attenuated, and an arrival error corresponding to the distance difference 2a + S occurs in the right ear c. However, L crosstalk other than the monaural component is canceled if the middle bass region is the center. However, the mono component of L is accompanied by crosstalk. Further, between the reproduced sound R of the right main unit 2 and the difference signal −kR of the left sub unit 1A, both reach the right ear c at the same time, and the secondary crosstalk is canceled. However, the difference signal −kR is Since the monaural component is not included, the monaural component of R is not attenuated, and an arrival error corresponding to the distance difference 2a + S occurs in the left ear b, but R crosstalk other than the monaural component also occurs if the center low frequency range is the center. However, the mono component of R is accompanied by crosstalk.
On the other hand, the left and right sub-units 1A and 2A are equidistant and equiangular from the listener. For this reason, the crosstalk kKL in which the left sub unit 1A difference signal kL reaches the right ear also reaches the right ear only with the right sub unit 2A difference signal -kL and the arrival error a, and the crosstalk is canceled. Further, the crosstalk kKR in which the right sub unit 2A difference signal kR reaches the left ear also reaches the left ear only with the left sub unit 1A difference signal -kR and the arrival error a, and the crosstalk is cancelled.
The left main unit 1 and the right sub unit 2A are obtained by equally dividing the virtual left unit XsL for reproducing L in front of the listener left ear, and the right main unit 2 and the left sub unit 1A are in front of the listener right ear. The relationship that is obtained by equally dividing the virtual right unit XsR for reproducing R in left and right is not changed.
On the other hand, between kL and -kL and between kR and -kR between the left and right sub-units 1A and 2A, a new virtual left and right unit that reproduces the difference signal of L and R in front of the listener is equally divided into left and right. The new virtual unit can be said to be equivalent to a one-point microphone, and does not cause a contradiction in which the left and right paired microphones are reversed.
However, the listener listens twice with a time difference except for the mono components of L and R, and the transient characteristics deteriorate.
The left and right sub-units 1A and 2A can also receive a fixed surround sound even if various multi-channel surround stereo rear left and right audio outputs, which are approximate to the L-nR and R-nL, are similarly input in reverse phase. The field is reproducible. This is because most of the rear channels are recorded with signal distribution according to L-nR and R-nL.

One of the embodiments of the first invention shown in FIG. 5 is to reduce the angle at which the left and right speaker systems 31 and 41 are viewed as described above. The function of the right sub-unit 2A may be replaced with at least one center unit, and the first embodiment including three units may be used.
Replacing the functions of the left sub-unit 1A and the right sub-unit 2A with the center unit becomes possible when the center unit reproduces -k (L + R) obtained by adding both reproduced sounds.
In addition, Example 1 can also be implemented by replacing with a configuration other than three units, as will be described later.

FIG. 6 shows a case where a method of inputting -kR and -kL to two voice coils of a so-called double voice coil speaker unit is used as one method of reproducing -k (L + R) by the center unit. Example 1 of the first invention constituted by the above three units is shown.
In FIG. 6, the left unit 11 replaces the left main unit 1 and reproduces L, and the right unit 21 replaces the right main unit 2 and reproduces R.
On the other hand, the center unit 8 of the double voice coil replaces the left sub-unit 1A and the right sub-unit 2A, and the R input is band-limited and attenuated by the filter unit 13 and connected in reverse phase, and the L input is connected. -K (L + R) is reproduced by band-limiting and attenuating and connecting in reverse phase with the filter unit 23.
When the center unit is arranged on the reference line D and in front of the listener on the reference line D, the left unit 11 is placed on the left of the center unit 8 at a distance substantially equal to the distance between both ears, and the right unit 21 is The center unit 8 is installed to the right by a distance substantially equal to the distance between both ears.

In the first embodiment with three units shown in FIG. 6, the mid-low range of the center unit 8 reproduced sound -k (L + R) is lost by canceling equally with the left and right units 11 and 21 regardless of the monaural component or other. . In addition, although the left and right units 11 and 21 have lower sound pressure in the middle and lower range by the amount offset with the center unit 8, in the L and R middle and lower ranges, both the monaural component and the monaural component both have crosstalk, and between the left and right units. Almost mono.
However, in the middle sound range, the monaural component of the center unit 8 reproduced sound -k (L + R) is canceled out equally with the left and right units 11 and 21, and thus has no crosstalk cancellation effect, and the interval between the left and right units 11 and 21 is small. The monaural component common to L and R is clearly centered between the left and right units 11 and 21 with a high level of crosstalk even in the midrange.
On the other hand, with the exception of the monaural component, crosstalk cancellation is established for the left ear b and the right ear c as follows.
For the left ear b, the center unit 8 reproduction sound −k (L + R) and the L reproduction sound of the left unit 11 reach simultaneously, and −kL corresponding to the secondary crosstalk is canceled by L. However, since the arrival error of -kR of the center unit 8 is small with respect to the crosstalk KR of the right unit 21, the limit of the R reproduction band by the filter f13 is small, and the crosstalk can be canceled widely in the mid range. For the right ear c, the center unit 8 reproduced sound -k (L + R) and the R reproduced sound of the right unit 21 arrive at the same time, and -kR corresponding to the secondary crosstalk is canceled by R. However, since the arrival error of -kL of the center unit 8 is small with respect to the crosstalk KL of the left unit 11, the limit of the L reproduction band by the filter f23 is small, and the crosstalk can be canceled widely in the mid range.
Further, in the high sound range, the interval between the left and right units 11 and 21 is narrow, but the crosstalk level is lower than the middle and low sound range due to the structure of the pinna.
Therefore, with the general characteristics between the stereo speaker and both ears, and the synergistic effect of the midrange crosstalk cancellation effect according to the first embodiment by the three units, the midrange monaural component is centered between the left and right units 11 and 21, During L and R other than the monaural component, crosstalk decreases in the high sound range.

In the first embodiment using the above three units, the virtual unit XsL for reproducing L in front of the listener's left ear is equally divided into a left unit 11 and a center unit 8 on a circle B by a half of the distance between both ears. In addition, the virtual unit XsR for reproducing R in front of the listener's right ear is divided into a right main unit 2 and a center unit 8 on a circle C by a half of the distance between both ears. For this reason, in the middle and high sound range, it becomes equivalent to a binaural interval pair microphone, and secondary crosstalk does not occur, and it is difficult to limit the listening area as in one embodiment of the first invention shown in FIG.
However, since the left and right volume difference information and the phase difference information cannot be obtained in the middle to low range in the first embodiment, the size of the middle and large musical instruments is not expressed.

Even when a general stereo speaker is viewed at a large angle, the crosstalk in the middle and high range is reduced, but the monaural component in the middle and high range is also less likely to be accompanied by crosstalk, and the central localization can be obtained only in the middle and low range. Absent. On the other hand, in the first embodiment with three units shown in FIG. 6, the monaural component has crosstalk in a wide band from the low range to the midrange, and crosstalk other than the monaural component in the wide range of the midrange. Therefore, a clear center position is obtained between the left and right speakers 11 and 21, and the listener uses the center position as a reference vector to generate a three-dimensional sound by vector synthesis of L and R volume difference information and phase difference information in the middle and high frequencies. Recognize typical sound field localization.
Therefore, the three-dimensional effect of the sound image and sound field in the middle sound range is clearer than when a general speaker is viewed from a large angle, and there is a surround sound field.

In one embodiment of the first invention, the center unit 8 of the double voice coil according to the first embodiment shown in FIG. 6 is also used in the same manner as the height difference between the main and sub units is allowed as described above. Even if the left and right units 11 and 21 have a vertical difference while maintaining a distance from the listener's both ears, the effect does not change.
Accordingly, the center unit 8 of the double voice coil is replaced with, for example, a normal unit 81 for reproducing -kR and a normal unit 82 for reproducing -kL, and the center unit 8 is located at a position when the center unit 8 is installed on the reference line D. On the other hand, it may be set up and down while maintaining a distance from the listener's ears to reproduce -k (L + R).
The above example is an example that the method of reproducing -k (L + R) in the center unit 8 is various, and is not limited. The method of reproducing -k (L + R) in the center unit 8 is free. It does not matter.

1 and 5 show one of the above-described embodiments of the first invention, and in Example 2 of the second invention, the left sub-unit 1A has a listener binaural interval E minutes from the left main unit 1. 1 to 5, the horizontal sub-unit 2 </ b> A is located in a region defined in the horizontal direction as 101 in FIGS. 1 and 5, which are 0 to 150% of a minute, indicates that they are installed in an area defined as 201 in the horizontal direction.
That is, in Example 2 of the second invention, in one of the embodiments of the first invention, the left sub unit 1A is arbitrarily set to 0 to 150% of the listener's binaural interval E with respect to the left main unit 1. This is implemented by providing the right sub-unit 2A to the left of an arbitrary distance g of 0 to 150% of the listener's binaural distance E with respect to the right main unit 2. .

  Therefore, Example 2 includes one of the embodiments of the first invention and Example 1 above, and further, the main and sub-unit intervals according to the angle at which the listener looks at the left and right speaker systems 31 and 41 are set to an arbitrary distance g. By changing to, it has multiple effects. Therefore, the left and right speaker systems 31 and 41 may be provided with a free means by which a listener can select two or more types of relative horizontal intervals between the main and sub units.

FIG. 7 schematically shows adapter rings AR1 and AR2 for interposing both the left main and sub units 1 and 1A when attaching to the enclosure, taking the left speaker system 31 as an example of the means.
The adapter rings AR1 and AR2 have a structure in which the unit mounting position is eccentric from the center of the adapter ring to one side, so that when the unit is attached to the adapter ring and the adapter rings AR1 and AR2 are further attached to the enclosure, Means for selecting a plurality of relative horizontal distances between the main and sub-units 1, 1A by selecting an attachment angle is shown.
Of course, the flange of the speaker unit may have the same form, and the adapter ring shown in FIG. 7 is not limited in shape, and may be, for example, a sub baffle. In short, the speaker unit and the enclosure are fixed. An example of means for selecting a plurality of relative horizontal distances between the main and sub-units due to the inclusions at the time is shown. In addition, when both the main and sub units are attached to one inclusion, and the inclusion is attached to the enclosure, a plurality of relative horizontal intervals between the main and sub units may be selected depending on the attachment angle.

In the left and right speaker systems 31 and 41, the left and right main and sub units 1, 1A and 2, 2A are installed on the baffle surface other than at an angle of 45 °, and the speaker is placed vertically or horizontally. There may be a structure in which a relative horizontal interval is selected, or the enclosure is a prism, cylinder, or polyhedron, and a plurality of relative horizontal intervals are selected depending on how the enclosure is placed.
FIG. 8 shows an example of an embodiment in which the enclosure has a substantially elliptic cylinder shape. In FIG. 8, the left and right main and sub units 1, 1A and 2, 2A are appropriately installed on the baffle surfaces of the left and right speaker systems 32 and 42, both of which are substantially elliptical, and the enclosure is set on a stand that does not limit the way of placement. Is an example of a shape in which a listener can freely replace and select a relative horizontal interval.
It should be noted that the adapter ring shown in FIG. 7 and the enclosure and pedestal shown in FIG. 8 are exemplifications that the above methods are various, and the method and shape are not limited, and there are a plurality of main and sub units. The method for selecting the horizontal distance is arbitrary.

In the second embodiment of the second invention, setting the main / sub unit interval to an arbitrary distance g out of 0 to 150% of the binaural interval E is the left main unit with respect to the listener left ear b. 1 is arranged rearward by an arbitrary distance difference h of 0 to 150% of the distance difference a from the left sub-unit 1A, and the right main unit 2 is connected to the right side of the listener right ear c This is equivalent to arranging an arbitrary distance difference h from 0 to 150% of the distance difference a from the sub unit 2A.
Therefore, when the positional relationship between the left and right speaker systems 31 and 41 and the listener is relatively fixed, the above-mentioned distance difference is satisfied by providing a front-back difference between the main and sub-units, or both the relative horizontal distance and the front-back difference are both satisfied. May be combined to satisfy the distance difference.
The main and sub units of the left and right speaker systems 31 and 41 are driven by multi-amplifiers with the main and sub unit intervals on the reference line set to an arbitrary distance gg substantially equal to the listener binaural interval E. The effects described below in the second embodiment of the second invention may be obtained by adding or subtracting the delay amount to either the main unit or the sub unit.

  In Example 2 of the second invention, the crosstalk canceling effect obtained in one of the above embodiments of the first invention is changed as shown in the following 2-1) to 2-4), and convenience in use is changed. Is added.

2-1) The first effect is that when the left and right speaker systems 31 and 41 are expected at a specific angle of approximately 50 ° to 180 °, the secondary crosstalk is L in one of the embodiments of the first invention. , R is an effect of correcting an error that is delayed with respect to the reproduced sound.
In one of the above-described embodiments of the first invention, the left and right main unit 1 corresponds to a distance difference aa caused when one of the left and right reproduced sounds goes around the listener's head when the angle at which the left and right speaker system is viewed increases. 2 can be finely adjusted by retreating the reference line D by aa and changing the characteristics of the filter portion as necessary.
This is because the delay of the secondary crosstalk is equal to the left and right sub-units 1A and 2A retracted by aa relative to the left and right main units 1 and 2, and therefore the left and right main units 1 and 2 are retracted by aa and L , R playback sound and secondary crosstalk reach simultaneously, in which case the crosstalk cancellation signal is limited to a playback band and volume that can be satisfactorily offset even if the arrival error between the crosstalk and the crosstalk cancellation signal becomes larger. Because it is good.
On the other hand, the delay of the secondary crosstalk is also equivalent to the fact that the left and right sub-units 1A and 2A approach the left and right main units 1 and 2 on the reference line D according to the distance difference aa.
Therefore, when the listener looks at the left and right speaker systems 31, 41 at a specific angle of approximately 50 ° or more, for example, 60 ° recommended by ITU-R, the left and right main, sub units 1 and 1A, and 2 and 2A, respectively. Is set at an appropriate interval of 100 to 150% of the binaural interval E, and in this case, the reproduction band and volume that are well offset even if the crosstalk is further delayed with respect to the crosstalk cancellation signal, It is sufficient to limit the crosstalk cancellation signal.
However, as the main / sub unit interval becomes larger than the binaural interval E, the relationship of ob = qb = pc = rc is lost when the listener moves back and forth, and the listening point becomes limited.

2-2) The second effect is that the “sound field emphasis effect by secondary crosstalk” described in 1-2) can be used according to the intention.
That is, when the left and right speaker systems 31 and 41 having a main and sub unit interval on a given reference line D and a specific interval equal to or greater than the binaural interval E are expected at a free angle within about 50 °, When looking at 50 ° or more, if the main and sub-unit intervals more than (a + aa) are taken, the secondary crosstalk reaches earlier than the L and R reproduced sounds. For this reason, the reproduced sound field is drawn to the listener side from the straight line connecting the left and right speakers, and the left and right widths of the sound field are maintained, and the sense of depth decreases in the front-rear direction.
This tendency can be used as an effect similar to the loudness control that emphasizes a surround component that is difficult to hear at the time of reproducing a small volume and makes it easy to hear. For example, it is reasonable to appropriately use the above-mentioned “sound field emphasis effect by secondary crosstalk” in a TV built-in speaker that is not used at a high volume as in a normal movie theater, or in a small stereo system that tends to be reproduced at a low volume. There is a case.

2-3) The third effect is an effect of making it difficult to limit the speaker shape when the left and right speaker systems 31, 41 are expected at a small angle.
In the second embodiment of the second invention, even when the left and right speaker systems 31 and 41 are viewed at a small angle, when the main and sub unit intervals are set to a distance substantially equal to the binaural interval E, it is most realistic and three-dimensional. Sound field can be obtained.
However, as the angle at which the left and right speaker systems 31 and 41 are viewed becomes smaller, a itself becomes smaller. For this reason, even if the distance between the left and right main and sub units 1 and 1A and 2 and 2A is equal to or less than the binaural distance E, the absolute delay amount of the secondary crosstalk with respect to L and R is small and the reproduced sound field is distorted. Hateful. Further, even when the left and right main and sub units 1 and 1A and 2 and 2A are vertically arranged and the horizontal distance difference becomes 0, the virtual units XsL and XsR only overlap at one point in front of the listener and do not reverse left and right. .
Therefore, when the left and right speaker systems 31 and 41 are viewed at a small angle, the left sub-unit 1A is moved to the right with respect to the left main unit 1 by an arbitrary distance g within a range of 0 to 100% of the binaural distance E. Even if the right sub-unit 2A is installed on the left with respect to the right main unit 2, the second embodiment of the second invention does not violate the stereophonic principle and is sufficiently realizable with a surround sound field. A stereo sound field can be reproduced, it is difficult to limit the shape of the speaker, and it can be mounted on a small acoustic device.

2-4) The fourth effect is an effect of stabilizing the reproduction sound field even when the listener moves to the left or right close to the speaker.
Since the left and right speaker systems 31 and 41 having the binaural interval E as the main and sub unit intervals have no confusion in the left and right phase difference, there is no contradiction in the reproduced sound field even if the listener moves left and right.
However, when the listener is close to the speaker, the arrival time of each unit reproduction sound to both ears b and c changes every moment, and even if the reproduction band and volume of the sub-units 1A and 2A are sufficiently limited, Sound field localization tends to be unstable.
On the other hand, when the left and right speaker systems 31 and 41 are expected at a small angle, the horizontal distance between the main and sub units 1 and 1A and 2 and 2A is 0% with 100% or less of the binaural distance E. Although the reproducibility of the sound field tends to decrease as it gets closer to, a stable sound field can be supplied to a wide listening area.

FIG. 9 shows a third embodiment of the crosstalk canceling method according to the third aspect of the present invention that cancels the crosstalk without generating the secondary crosstalk by limiting the crosstalk canceling band by using an arbitrary stereo speaker system and sub-speaker. It is a perspective view shown.
In any stereo speaker system, the left speaker system indicated by m1 in FIG. 9 reproduces L, and the right speaker system indicated by m2 reproduces R, and is arranged at an equal angle from the listener to the left and right. .
On the other hand, in the third embodiment, the left sub-speaker s1 including at least one speaker unit and the filter unit and the right sub-speaker s2 including at least one speaker unit and the filter unit are used. The right audio output R is connected to the speaker s1, the reproduction band and the volume are limited by the filter unit f14, the reverse phase reproduction is performed by at least one speaker unit 1AA, and the left audio output L is output to the right sub-speaker s2. Are connected, the playback band and volume are limited by the filter unit f24, and reverse-phase playback is performed by at least one speaker unit 2AA.

At this time, the filter parts f14 and f24 are substantially equal to the filter parts f11 and f21 in the embodiment of the first invention.
That is, the filter unit f14 does not generate a reverse phase feeling even if the crosstalk cancellation signal −kR of the left sub-speaker s1 reaches the left ear at least 2a + s ahead of the crosstalk KR of the right speaker system m2, and is good. The range is limited according to the angle at which the listener expects the left and right speaker systems m1 and m2 and the following production intention within the reproduction band and volume that are offset by.
In addition, the filter unit f24 does not cause a reverse phase feeling even when the crosstalk cancellation signal −kL of the right sub-speaker s2 reaches the right ear at least 2a + s ahead of the crosstalk KL of the left speaker system m1. The range is limited according to the angle at which the listener expects the left and right speaker systems m1 and m2 and the following production intention within the reproduction band and volume that are well offset.
The above production intention is, for example, that the listening area is limited to have a wide crosstalk cancellation effect up to a high frequency band, or that the listening area is limited to a low frequency band to have a crosstalk cancellation effect. Is intention.
The filter parts f14 and f24 are generally composed of a low-pass circuit and an attenuator, but both are not necessarily essential. For example, when the left and right sub-speakers s1 and s2 have a low-pass characteristic due to unit characteristics or an enclosure structure, a low-pass circuit is not necessarily required. Also, the sound pressure level of the left and right speaker systems m1 and m2 can substantially exceed the left and right subspeakers s1 and s2 due to the efficiency, impedance, and the number of units between the left and right speaker systems m1 and m2 and the left and right subspeakers s1 and s2. For example, an attenuator is not necessarily required.
In short, the filter sections f14 and f24 are intended to relatively limit and attenuate the reproduction sound of the left and right sub-speakers s1 and s2 relative to the reproduction sound of the left and right speaker systems m1 and m2.
The filter parts f14 and f24 may have free characteristics after satisfying the relative filter characteristics between the left and right speaker systems m1 and m2 and the left and right sub-speakers s1 and s2. The free characteristic is a special slope characteristic including, for example, a band pass characteristic or a combination of a plurality of band pass characteristics, and may include a free circuit or the like. Examples of the free circuit include a circuit that corrects phase rotation caused by a filter and a means that allows a listener to adjust filter characteristics and attenuation.
The functions of the filter units f14 and f24 can be replaced with a free external device such as a sound processor or an equalizer using a multi-amplifier, and the amplifier and the external device may be incorporated in the left and right sub-speakers s1 and s2.
The sub-speakers s1 and s2 have a filter characteristic k relative to the left and right speaker systems m1 and m2 by the filter units f14 and f24, so that the left speaker system m1 has L and the left sub-speaker s1 has -kR. , Right speaker system m2 reproduces R, and right sub-speaker s2 reproduces -kL simultaneously.

At this time, among the units constituting the left speaker system m1, the unit that mainly reproduces the reproduction band of the left sub-speaker s1 is the left speaker position 1V, and among the units constituting the right speaker system m2, the right sub-speaker s2 A unit that mainly reproduces the reproduction band is set to the right speaker position 2V.
FIG. 9 shows a case where the left and right speaker positions 1V and 2V are woofers of the three-way unit. FIG. 9 shows examples of the 1V and 2V positions defined as described above. 1V and 2V are not necessarily woofers. The left and right speaker systems m1 and m2 and the left and right sub-speakers s1 and s2 are also for indicating unit positions, and do not limit the form, number of units, and installation position of each speaker.

The left speaker position 1V, the left sub-speaker s1, the right sub-speaker s2, and the right speaker position 2V are arranged on a straight line, and the left and right speaker systems m1 and m2 are arranged equidistantly and at the same angle with respect to the listener. Therefore, the four positions are both on the reference line D.
With this as a precondition, the left speaker position 1V is approximately equal to the listener binaural distance E, the right position on the reference line D is ss1, and the right speaker position 2V is The left sub-speaker s1 is installed at ss1 and the right sub-speaker s2 is installed at ss2, with the position on the reference line D to the left at a distance substantially equal to the ear interval E, and the right sub-speaker s2 is installed at ss2. One form of crosstalk cancellation method is implemented.
However, the left and right sub-speakers s1 and s2 can be installed in the ss1 and ss2 with respect to the left and right speaker positions 1V and 2V of any stereo speaker system only when both are small.

On the other hand, it is equivalent to one of the embodiments of the first invention shown in FIG. 5, and the speaker arrangement of Example 3 of the third invention shown in FIG. 9 also has the left speaker position 1V and the right sub-speaker s2 with respect to the listener's left ear. Since the purpose is to make the right speaker position 2V and the left sub-speaker s1 equidistant with respect to the listener's right ear, the left speaker position 1V is maintained if the distance between the listener's ears b and c is substantially maintained. The same effect can be obtained even if there is a height difference between the left sub-speaker s1 and between the right speaker position 2V and the right sub-speaker s2.
Accordingly, when the straight line passing through the left and right speaker positions 1V and 2V is defined as the reference line D, the left sub speaker s1 is located on the circle 6 having a radius from between the listener's ears to the position ss1 on the reference line D. There may be a difference in height with respect to the position 1V, and the right sub-speaker s2 is located on a circle 7 having a radius from between the listener's ears to the position ss2 on the reference line D, and the right speaker position 2V. A difference in height may be generated.
However, the left and right sub-speakers s1 and s2 are generally preferably close to the left and right speaker positions 1V and 2V. The closer the two are, the more the monaural components common to L and R cancel each other up to a higher frequency, and the central location becomes clearer with crosstalk up to the higher frequency of the monaural component. This is because the sound field reproduced by vector synthesis of phase difference information becomes more realistic.

According to the speaker arrangement as described above, the left and right speaker systems m1 and m2 can be connected to the listener's left ear position at any angle at which the distance aa at which one of the reproduced sounds circulates around the head does not increase and is generally within 50 °. The first parallelogram created by the right ear position, ss1, and the left speaker position 1V is always congruent with the listener's left ear position, right ear position, right speaker position 2V, and the second parallelogram created by ss2. become.
Therefore, both the left speaker position 1V and the right sub-speaker s2 are on a circle B centered on the listener's left ear or are in a distance relationship equal to this, so that the secondary crosstalk of the right sub-speaker s2 − kKL reaches the left speaker position 1V and the left ear at the same time and cancels out. In addition, the right speaker position 2V and the left sub-speaker s1 are both on a circle C centered on the listener's right ear or are in a distance relationship equal to this, so that the secondary crosstalk of the left sub-speaker s1 − kKR reaches the right speaker and R of the unit at the right speaker position 2V at the same time and cancels out.
However, in this case, the crosstalk cancellation signal −kL of the right sub-speaker s2 reaches the right ear ahead of the crosstalk KL of the unit at the left speaker position 1V by 2a + S, and the left sub-speaker s1 The crosstalk cancellation signal −kR reaches the left ear ahead of the crosstalk KR of the unit at the right speaker position 2V by 2a + S.
However, the reproduced sound of the left and right sub-speakers s1 and s2 is limited to a reproduction band and a volume that can be satisfactorily offset by the filter units f14 and f24 without causing a reverse phase feeling even at least with the arrival error. According to one of the embodiments of the third invention, crosstalk cancellation is also established centering on the mid-low range without secondary crosstalk.
The monaural component common to L and R is canceled out between the unit at the adjacent left speaker position 1V and the left sub-speaker s1, and between the unit at the adjacent right speaker position 2V and the right sub-speaker s2. Crosstalk cancellation is not performed for L and R monaural components in the sound range. In the middle and high sound range, the crosstalk level is attenuated due to the general characteristics between the left and right speaker systems m1 and m2 and the listener's ears b and c.
Therefore, due to the general characteristics between the stereo speaker and the listener's both ears, and the effect of the third embodiment of the third invention that cancels crosstalk other than the L and R monaural components at the center of the mid-low range, the monaural sound image is the left and right speaker systems m1 and m2. In the middle, the crosstalk level is kept low in the entire reproduction band, and secondary crosstalk does not occur, and a good stereo sound field including a surround sound field can be obtained.

  In the third embodiment, the position ss1 is set to 0 to 150% of the listener binaural distance E with respect to the left speaker position 1V on the reference line D, that is, in the region 101 defined in the horizontal direction in FIG. The position ss2 is set to the right by an arbitrary distance g, and the position ss2 is 0 to 150% of the listener's binaural distance E with respect to the right speaker position 2V on the reference line D, that is, in the horizontal direction in FIG. It is also implemented by providing an arbitrary distance g in the defined area 201 on the left.

Accordingly, in the third embodiment, the left and right sub-speaker positions ss1 and ss2 are set to appropriate distances of 100 to 150% of the binaural distance E with respect to the left and right speaker positions 1V and 2V, and the filter characteristics are appropriately changed. By providing a margin, as described in 2-1) as one effect of the second embodiment of the second invention, even when the left and right speaker systems m1 and m2 are arranged at approximately 50 to 180 °, 2 The crosstalk can be canceled without causing the next crosstalk. Further, as described in 2-2), the “sound field enhancement effect by secondary crosstalk” can be used.
Further, as the left and right speaker systems m1 and m2 are viewed at a small angle, the left and right sub-speaker positions ss1 and ss2 are set to an arbitrary distance of 0 to 100% of the binaural distance E with respect to the left and right speaker positions 1V and 2V. By appropriately changing the k of the filter unit accordingly, as described in 2-3), the sound field reproducibility tends to decrease within the range that does not contradict the stereophonic principle, but a stable sound field can be obtained. , Can be supplied to a wide listening area.

In the third embodiment of the third invention shown in FIG. 9, the left sub-speaker s1 is also inverted from the L difference signal audio output having a free L and R signal ratio n, which can be expressed as L-nR (n> 0). It is possible to input the R difference signal audio output having a free L and R signal ratio n, which can be expressed as R-nL (n> 0), to the right sub-speaker s2 in reverse phase, as described above. It is consistent with the stereophonic principle. However, the listener listens twice with a time difference except for the mono components of L and R, and the transient characteristics deteriorate.
Even if various multi-channel surround stereos, which are approximate to the difference signal audio output, are input to the left sub-speaker s1 and the rear right-channel output is input to the right sub-speaker s2 in reverse phase, the surround sound is constant. The sound field can be reproduced. This is because most of the rear channels are recorded with signal distribution according to L-nR and R-nL.

  The left and right sub-speakers s1 and s2 shown in the third embodiment can be manufactured as dedicated speakers for a specific stereo speaker system, and are also manufactured as general-purpose sub-speakers that add the above-described crosstalk cancellation effect to a general stereo speaker system. It doesn't matter.

In FIG. 10, the fourth embodiment of the fifth invention is a crosstalk cancel circuit shown in a block diagram of 50. FIG.
That is, in the crosstalk cancellation circuit 50, the left audio input L is set to W and X, the right audio input R is set to Y and Z, and the above W and Z are input to the delay block, respectively. Is set or adjusted to a delay characteristic of an arbitrary distance difference h within a range of 10% to 150% of the distance difference (a + aa) corresponding to the angle at which the left and right speakers 33 and 43 are viewed.
On the other hand, the above X and Y are respectively input to the filter block, and have relative filter characteristics described in the next section with respect to W and Z.
From the W output from the delay block, Y output from the filter block is subtracted to obtain the left audio output, and from the Z output from the delay block, X output from the filter block is subtracted. By using the right audio output, the most suitable crosstalk cancellation effect and the effect according to the intention can be obtained in the fourth embodiment.

The filter block amplifies the left and right audio outputs of the crosstalk cancellation circuit 50 and reaches the right ear c from the left speaker 33 when reproduced by the left and right speakers 33 and 43 arranged at equal angles from the listener. Even if the crosstalk cancellation signal −kL of the right speaker 43 reaches the listener's right ear c at least by a distance difference of 2 (a + aa) with respect to the crosstalk KL, no reverse phase sensation occurs, and the crosstalk is canceled well. In addition, the crosstalk cancellation signal −kR of the left speaker 33 with respect to the crosstalk KR reaching the left ear b from the right speaker 43 is in reverse phase even if the listener reaches the listener's left ear b at least by the distance difference 2 (a + aa). Within the frequency band and output level that cancel well, without causing a feeling, the following producers or users It is intended to limit X, the Y in accordance with FIG.
The intention of the above producer or user is to limit the listening area to have a crosstalk cancellation effect up to a higher frequency band, or to limit the listening area to a low frequency band in order to widen the listening area. The intention is to have
The filter block is generally composed of a low-pass circuit and an attenuator or attenuation circuit. In addition, the filter block may satisfy the above filter characteristics and give X and Y free characteristics. Is an irregular slope characteristic including, for example, a combination of a bandpass characteristic and a plurality of bandpass characteristics, and the filter block may include a free circuit, etc. There are a circuit that corrects the phase rotation when the phase rotation occurs and a means that allows the user to adjust the filter characteristics and attenuation.

As shown in FIG. 10, the crosstalk cancel circuit 50 amplifies the circuit output by the L and R amplifiers and reproduces it by the left and right speakers 33 and 43 as described above. In addition to being combined into two output systems, three or more audio outputs may be used depending on the system configuration.
For example, in the case of a system configuration in which a multi-amplifier is used to play back with the left and right speakers 33A and 43A together with the left and right speakers 33 and 43, the left and right speakers 33 and 43 and the left and right speakers 33A and 43A for canceling the crosstalk are The W output from the delay block is used as the left audio output, the Z output from the delay block is used as the right audio output, and the output from the filter block. The left audio output for crosstalk cancellation can be obtained with Y as the reverse phase, and the right audio output for crosstalk cancellation can be obtained with X output from the filter block as the reverse phase.
Similarly, in the case of a system configuration in which crosstalk cancellation is performed by the center speaker 9 together with the left and right speakers 33 and 43 using a multi-amplifier, the left and right speakers 33 and 43 and the center speaker 9 are arranged at an equal distance from the listener. As a reference, W output from the delay block is set as L audio output, Z output from the delay block is set as R audio output, and Y and X output from the filter block are added. Thus, the phase can be reversed and the center channel output can be obtained.
When three or more speakers are used, there may be included means for comprehensively matching the arrival time of each channel to the listener while satisfying the delay characteristics according to the distance difference between the listener and each speaker.

The circuit of each of the filter block and the delay block constituting the crosstalk cancellation circuit 50 can be freely designed. Note that the band limit and attenuation amount of the filter block are usually much larger than the current general crosstalk cancellation circuit, but the delay amount itself corresponding to the distance difference (a + aa) of the delay block itself is the general crosstalk cancellation circuit. This is the same as the delay characteristic created in step 1, and the block configuration is completely different.
The crosstalk cancellation circuit 50 may have a free function and a circuit while satisfying the delay characteristic and the filter characteristic, and the free function and the circuit correspond to, for example, the angle at which the left and right speakers 33 and 43 are viewed. Functions that change the filter characteristics and delay characteristics with relevance, functions that specify the positional relationship between the listener and the left and right speakers 33 and 43, and a method for specifying the face shape of the listener, and the playback sound There is a circuit for adding a pseudo reverberation sound.
Further, in the fourth embodiment of the fifth invention, when the above-mentioned “sound field emphasis effect by secondary crosstalk” is used with an arbitrary value of 100 to 150% of (a + aa), the L and R reproduced sounds have 2 As the next crosstalk precedes, the reproduced sound field tends to converge between both ears. Therefore, a circuit for appropriately rotating the X, Y, or W, Z phase may be provided so that the sound field does not converge between both ears.
The examples of the above free functions and circuits are exemplifications that the crosstalk cancellation circuit 50 satisfies the above delay characteristics and filter characteristics, and that various functions and circuits can be provided without contradiction. The crosstalk cancel circuit 50 may have a free function or circuit.

In the fourth embodiment shown in FIG. 10 configured as described above, when the delay amount is the distance difference (a + aa), L of the left audio output L-kR of the crosstalk cancellation circuit 50 is equal to −kR. In the right audio output R-kL, R is delayed by the distance difference (a + aa) with respect to -kL.
Therefore, when the left audio output L-kR is amplified by the L amplifier and reproduced by the left speaker 33, and the right audio output R-kL is amplified by the R amplifier and reproduced by the right speaker 43, the following is performed. Thus, the secondary crosstalk is canceled.
That is, as shown in FIG. 10, -kR reproduced at the left speaker position ii and R reproduced at j behind (a + aa) from the right speaker position are the eccentric circles around the listener right ear c. Are both on C, so −kR secondary crosstalk −kKR reaches the right ear c simultaneously with R and cancels out. -KL played at the right speaker position jj and L played at i (a + aa) backward from the left speaker position are both on the eccentric circle B around the listener's left ear b, and therefore- kL secondary crosstalk -kKL is canceled when L reaches the left ear b simultaneously with L.
However, in this case, although the secondary crosstalk is canceled, the crosstalk cancellation reaching the right ear c from the right speaker position jj with respect to the crosstalk KL reaching the right ear c from i behind (a + aa) from the left speaker position. The signal −kL reaches in advance by the distance difference 2 (a + aa). In addition, with respect to the crosstalk KR that reaches the left ear b from j behind (a + aa) from the right speaker position, the crosstalk cancellation signal −kR that reaches the left ear b from the left speaker position ii is equal to the distance difference 2 (a + aa). Reach ahead.
However, in the above filter block, even if X and Y, which are crosstalk cancellation signals, reach at least 2 (a + aa) ahead of the crosstalk of the W and Z reproduced sound, no reverse phase sensation is produced and the frequency is canceled well. Since it is limited to the band and the output level, the crosstalk cancellation is also established centering on the mid-low range without secondary crosstalk by the effect of the crosstalk cancellation circuit of the fourth embodiment.
Note that the L-kR output from the crosstalk cancel circuit 50 has reverse characteristics in the middle and low frequencies of the monaural component common to L and R, almost independently of the delay amount of about 1 millisecond (ms), and is in the mono component of L. The bass range is attenuated and the mid-bass range of the monaural component of -kR is lost. Similarly, in R-kL, the monaural component bass range common to L and R has an inverse characteristic almost regardless of the delay amount of about 1 millisecond (ms), and the R mono component mid-bass range is attenuated. -KL mono component mid-low range is lost.
Therefore, the mono components in the L and R mid-low range are not canceled by the crosstalk, but are centered between the left and right speakers 33 and 43 along with the cross talk, and only the L and R monaural components in the mid and low range are cross-talk canceled. . Also, crosstalk is a problem with stereo speakers, especially in the mid-low range. Crosstalk attenuates to such an extent that it can be accompanied by a sufficient sound field.
Therefore, the left and right speakers are combined with the general characteristics between the stereo speakers and the listener's both ears, and the effect obtained in the fourth embodiment of the fifth aspect of the present invention that cancels crosstalk other than the L and R monaural components at the center of the mid-low range. In 33 and 43, the monaural sound image is localized between the speakers due to the crosstalk in the mid-low range, and the crosstalk is kept low in almost the entire reproduction band, so that secondary crosstalk does not occur and a good stereo sound including the surround sound field is obtained. A place is obtained.

In the fourth embodiment shown in FIG. 10, it is necessary to set the delay amount according to the angle at which the listener looks at the left and right speakers 33 and 43, that is, the distance between the listener and the speaker.
However, as described above, the L playback position i and the -kL playback position jj are obtained by equally dividing the virtual left unit XcL on the left front of the listener on the circle B, and the R playback position j and the -kR playback position ii are the listeners. The right front virtual right unit XcR is equally divided on the circle C, and is consistent with the stereophonic principle. Therefore, when the listener moves to the left or right, the uncomfortable feeling that tends to occur in a general crosstalk cancellation system does not occur.
When the listener is in the center of the speaker, the secondary crosstalk reaches both ears simultaneously with the L and R reproduced sounds. Therefore, when the listener moves to the left and right, the secondary crosstalk is delayed from the L and R reproduced sounds, and the moving sound field tends to spread to the rear of the left and right speakers. In the opposite ear, the secondary crosstalk reaches earlier than the L and R reproduced sounds, and the opposite sound field tends to be drawn to the listener side.
For this reason, even if the listener moves left and right, the sound field feeling is easily maintained, and the sound image and the sound field are not easily masked by the near-side speaker. Therefore, although the entire sound field is shifted in the moving direction, a certain surround sound field feeling is maintained and the resolution is not easily lowered.
Note that the above-described crosstalk cancellation effect is established without contradiction even behind the listener.

In Example 4 of the fifth invention, when W and Z are set to an arbitrary delay characteristic of 100% to 150% of the distance difference (a + aa) corresponding to the angle at which the left and right speakers 33 and 43 are viewed, Secondary crosstalk arrives at both ears b and c in advance with respect to the L and R reproduced sounds, and as described in 2-2), the “sound field enhancement effect by secondary crosstalk” is obtained, and the volume is small. It has the effect of making it easier to hear surround sound fields that are difficult to hear during playback.
In this case, in the fourth embodiment of the fifth invention, the reproduced sound field is attracted from the left and right speakers to the listener side and tends to converge between both ears. Therefore, the phase of X, Y or W, Z is used. The playback sound field may be appropriately deformed so that it does not converge between the listener's ears.

On the other hand, when the listener looks at the left and right speakers 33 and 43 at a small angle, that is, as aa becomes 0 in (a + aa) and a itself becomes smaller, the delay characteristic is 10 to 100% of the distance difference (a + aa). The effect of secondary crosstalk on the sound field is small.
During this time, the virtual unit XcL is on the left side of the listener front, and the XcR is on the right side of the listener front, which is consistent with the stereophonic principle.
In this case, as described in 2-4), although the sound field reproducibility tends to be lower than when the delay characteristic is a distance difference (a + aa), a stable sound field can be provided in a wide listening area. Can supply.

  FIG. 10 shows a block diagram when the left audio input L is divided into W and X and the right audio input R is divided into Z and Y in the fourth embodiment of the fifth invention. In the cancel circuit 50, the left audio signal L is input to W, the right audio signal R is input to Z, and X can be expressed as R-nL (n> 0). An R difference signal having a free L and R signal ratio n, which can be expressed as L-nR (n> 0), may be input in reverse phase.

When the R difference signal R-nL is input to X in reverse phase, the filter block output becomes k (nL-R), and the output is subtracted from the Z delay block output R. The right audio output is R + k (R−nL), but the difference signal −knL does not subtract the monaural component common to R and R from R in R−knL. In addition, when the L difference signal L-nR is input to Y in reverse phase, the filter block output becomes k (nR-L), and the output is subtracted from the W delay block output L. The left audio output is L + k (L-nR), but the difference signal -knR does not subtract the monaural component common to L and R from L in L-knR.
Therefore, when the output is amplified by the L and R amplifiers and reproduced by the left and right speakers 33 and 43, the middle and low frequencies of the monaural component of the L and R reproduced sounds are not attenuated and do not attenuate, and the audibility level is 3 dB at maximum. It rises to a certain extent and is centered between the left and right speakers 33, 43.
On the other hand, when the delay characteristic is the distance difference (a + aa), the left audio output, L + k (L-nR), between L and the difference signal -nR, and the right audio output, R + k (R-nL), R Between the difference signal and nL, there is a crosstalk cancellation effect as in the case where L and R are input to X and Y except that the monaural component is not attenuated.
Further, the crosstalk kKR of the difference signal kR reproduced at the right speaker position jj reaches the left ear with a delay of the distance signal (a + aa) from the difference signal −knR reproduced at the left speaker position ii. Accordingly, both the frequency band and the output level that are satisfactorily canceled even with the arrival error of (2a + aa) are well crosstalk canceled. In addition, the crosstalk kKL of the difference signal kL reproduced at the left speaker position ii reaches the right ear after being delayed by a distance difference (a + aa) from the difference signal −knL reproduced at the right speaker position jj. Accordingly, both the frequency band and the output level that are satisfactorily canceled even with the arrival error of (2a + aa) are well crosstalk canceled.

Even when the above difference signal is used, L played back by i of (a + aa) behind the left speaker and -knL played back at the right speaker position jj is a virtual left that plays L at the front left of the listener. The unit XcL is equally divided into left and right, and R is played back at j (a + aa) behind the right speaker and -knR is played back at the left speaker position ii. The relationship that the target left unit XcR is equally divided into the left and right is the same as when the difference signal is not used.
On the other hand, between kL and −knL reproduced at the left and right speaker positions ii and jj, and between −knR and kR, a new virtual left and right unit that reproduces the difference signal of L and R at the intersection of the eccentric circles B and C is moved to the left and right. The relationship is equally divided. The new virtual unit can be said to be equivalent to a one-point microphone, and does not cause a contradiction that is reversed left and right with respect to both ears.
However, when a difference signal is used for X and Y, the listener listens twice with a time difference except for the monaural components of L and R, and the transient characteristics deteriorate.
In addition, various multi-channel surround stereo rear right audio signals that are approximate to R-nL are input in reverse phase to X, and Y is the rear left of various multi-channel surround stereo that is approximate to L-nR. It is also possible to obtain a constant surround sound field by inputting audio signals in reverse phase. This is because most of the rear channels are recorded with signal distribution according to L-nR and R-nL.

The fourth embodiment of the fifth invention is implemented as a crosstalk cancellation circuit indicated by 50 in FIG. 10, but its application is not limited to a stereo reproduction device.
For example, the crosstalk cancellation circuit indicated by 50 in the fourth embodiment also has a certain crosstalk cancellation effect on the recorded sound and the produced sound when producing various audio sources in applications where the positional relationship between the speaker and the listener can be specified to some extent. When reproducing various audio sources produced in this way, a crosstalk cancellation effect can be added to a part of the reproduced sound or the entire reproduced sound by an audio system using a general stereo speaker.
The various audio sources mentioned above refer to media from which sound can be obtained, such as CD and DVD sound, TV sound, movie sound, radio, and sound of portable small wireless telephones, Note that the above examples are examples of the variety of media from which sound can be obtained, and are not intended to limit the media. For example, the playback sound field is extremely limited in the sound of guidance broadcasting and car navigation (Satellite navigation system). For example, the sound image can be localized at the listener's ears to attract attention, or the direction can be indicated by sound.

In the fourth embodiment of the fifth invention, the microphone and the speaker are equivalent signal converters as long as they do not contradict the stereophonic principle as described above. Therefore, the crosstalk cancellation circuit indicated by 50 in FIG. 10 based on the above principle also sets the delay characteristic to a negative distance difference, and subtracts Y and X output from the filter block from W and Z output from the delay block. By adding them instead, it can be applied as a crosstalk generating circuit for headphones, and there is no contradiction in the stereophonic principle.
Note that delaying W and Z by an arbitrary distance difference − (a + aa) relative to X and Y means delaying X and Y from W and Z by an arbitrary distance difference (a + aa), respectively. be equivalent to.
Therefore, the crosstalk generating circuit may have the block configuration shown in the block diagram 55 of FIG. 11 as the fifth embodiment of the fifth invention.
The crosstalk generating circuit 55 sets the left audio input signal L to W and X, and the right audio input signal R to Y and Z.
Each of X and Y is input to a delay block, delayed by an arbitrary distance difference (a + aa) with respect to W and Z, and each of X and Y is also input to a filter block. Even if the arrival error occurs, it is arbitrarily limited as long as the comb filter phenomenon and the reverse-phase sensation do not increase, and X and Y may be input in either the delay block or the filter block,
The Y output that has passed through both the delay block and the filter block is added to W as the left audio output, and the X that is output through both the delay block and the filter block is added to Z as the right audio output. It is carried out.
In addition, the configuration of the fifth embodiment of the fifth invention allows free application while satisfying the above basic configuration.
As an example, a mono component common to L and R obtained by subtracting the difference signal 2LR and the difference signal 2R-L from L + R obtained by adding the L and R audio signals, for example, can be arbitrarily set to L and R. And an input example in which the signals are mixed and input to X and Y, and the monaural component common to L and R in the mid-low range and the crosstalk amount other than the monaural component are relatively changed. In addition, said example is an example and does not limit a free application.
Note that the above-described circuit for obtaining a three-dimensional headphone sound field by using a normal stereophonic recording sound is not limited to use as a headphone for listening to music, and is effective for use as a hearing aid. In this frequency band, three-dimensional sound field reproducibility is obtained, so that sufficient sound field reproducibility can be obtained even for a so-called bone conduction type headphone.

  It should be noted that the crosstalk cancellation circuit of the fourth embodiment of the fourth invention and examples of its use, and the application examples and examples of its use shown in the fifth embodiment are examples of the variety and are not limiting. Therefore, the fifth invention may use the crosstalk cancellation principle that does not cause secondary crosstalk by limiting the crosstalk cancellation band, which is common to the first and second inventions, in various embodiments.

It is a perspective view which shows the form of 1st invention, and the sub unit installation area | region of 2nd invention. It is a block diagram which shows the basic composition of the present crosstalk cancellation circuit. It is a block diagram at the time of replacing the "current crosstalk cancellation system" with a speaker. It is a top view which shows SDA arrangement | positioning and the relationship with a listener. It is a top view which shows the relationship between a 1st invention and a listener's both ears, and the sub unit installation area | region of 2nd invention. It is a block diagram which shows Example 1 by 3 units. It is a perspective view which shows an example of an adapter ring. It is a perspective view which shows an example of an elliptical columnar enclosure. It is a perspective view which shows Example 3 and the range in which a sub speaker is installed. FIG. 10 is a block diagram illustrating a crosstalk cancellation circuit according to a fourth embodiment. FIG. 10 is a block diagram illustrating a crosstalk generating circuit according to a fifth embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Left main unit 11 3 units left unit 101 Left sub unit installation area 1A Left sub unit 1AA Left sub speaker unit 1V Left speaker position 2 Right main unit 21 3 unit right unit 201 Right sub unit installation area 2A Right sub-unit 2AA Right sub-speaker unit 2V Right speaker position 3, 31-33 Left speaker system 4, 41-43 Right speaker system 5, 50 Crosstalk cancel circuit 55 Crosstalk generation circuit 6, 7 Between listener's ears Vertical circle 8 with radius up to 1 ss and 2 ss in the center 8 Center unit a Distance difference aa measured by a straight line when the left and right playback sound reaches one ear position aa One of the left and right playback sound wraps around the head Distance difference AR1, AR2 Adapter ring b Left ear position B b , And eccentric circle BL -kL reproduction position BR on line bc R reproduction position c on line bc Circle around right ear position Cc, and eccentric circle CL L reproduction position CR line bc on straight line bc Upper -kL playback position D, D1, D2 A straight line parallel to the line connecting both ears (reference line)
E Interval between both ears f11-14 Left filter part
f21-24 Right filter section i L playback position ii -kR playback position j R playback position jj -kL playback position m1, m2 Stereo left and right speaker system o The intersection of the left side of circle B, eccentric circle B and reference lines D, D1 p The intersection point of the circle C, the eccentric circle C and the reference line D, D1 on the left side q The circle B, the eccentric circle B and the reference line D, the right intersection point of the D1 r The circle C, the eccentric circle C and the reference line D, D1 on the right side Intersections s1, s2 Left and right sub-speakers ss1, ss2 Virtual left and right sub-speaker positions VcL, VcR Virtual left and right units VsL, VsR SDA arranged virtual left and right units WL Signal side X L XcL, XcR Virtual left-right unit XsL, XsR of the fourth invention Crosstalk cancellation signal side ZR signal side for processing the virtual left-right unit YR of the first invention

Claims (6)

A stereo speaker system,
The left speaker system includes at least one speaker unit to which the left audio output L is input (hereinafter referred to as a left main unit in claims 1 and 2) and at least one speaker to which the right audio output R is input. The left main unit reproduces the left audio output L by including a unit (hereinafter referred to as a left sub-unit in claims 1 and 2) and a left filter unit connected to the left sub-unit. The left filter unit limits the playback band and volume, and the left sub unit reproduces the right audio output R in reverse phase,
In addition, the right speaker system includes at least one speaker unit to which the right audio output R is input (hereinafter referred to as a right main unit in claims 1 and 2) and at least one speaker unit to which the left audio output L is input. And a right filter unit connected to the right sub-unit to reproduce the right audio output R in the right main unit. In addition, the right filter unit limits the playback band and volume, and the right sub unit reproduces the left audio output L in reverse phase.
When the left and right speaker systems are equidistant to the left and right and at the same angle, the left main unit, the left sub unit, the right sub unit, and the right main unit are arranged on a straight line parallel to the straight line connecting the listener's ears. However, if the distance from the listener's both ears is kept approximately when placed on the straight line, the height difference between the left main unit and the left sub unit, and between the right main unit and the right sub unit Allow,
In the left speaker system, the left sub-unit is provided on the right side at a distance substantially equal to the listener's binaural interval with respect to the left main unit,
In the right speaker system, the right sub unit is provided on the left side at a distance approximately equal to the listener's binaural interval with respect to the right main unit.
When secondary crosstalk is canceled with the above configuration, even if the crosstalk cancellation signal of the right sub-unit reaches the listener's right ear before the crosstalk of the left main unit playback sound, no reverse phase sensation will occur and it will be good The right sub-unit playback sound is limited by the right filter unit within the canceled playback band and volume, and the left sub-unit crosstalk cancellation signal precedes the listener's left ear than the right main unit playback sound crosstalk. A crosstalk cancellation stereo speaker system characterized in that the left sub-unit reproduction sound is limited by the left filter unit within a reproduction band and volume that do not cause a reverse phase feeling and can be satisfactorily canceled even if they reach the sound. .   2. The crosstalk canceling stereo speaker system according to claim 1, wherein the left sub unit is provided to the right at an arbitrary distance g of 0 to 150% of a listener binaural interval with respect to the left main unit, and the right A crosstalk canceling stereo speaker system, wherein the right sub-unit is provided on the left side at an arbitrary distance g of 0 to 150% of the listener binaural distance with respect to the main unit. A crosstalk canceling method, in which a left audio output L is reproduced by a left speaker system, which is arranged at an equal distance from the left and right with respect to a listener of an arbitrary stereo speaker system, and a right audio output is output by a right speaker system. Play R,
In the left sub-speaker comprising at least one speaker unit and a left filter unit connected to the speaker unit, the right audio output R is reproduced in reverse phase by limiting the reproduction band and volume by the left filter unit. In the right sub-speaker including at least one speaker unit and a right filter unit connected to the speaker unit, the left audio output L is limited in phase by limiting the reproduction band and volume by the right filter unit. Play and
The unit of the left speaker system that mainly reproduces the reproduction band of the left sub-speaker is set as the left speaker position (in the following, the same applies hereinafter), and the right speaker system that mainly reproduces the reproduction band of the right sub-speaker When the unit is set to the right speaker position (hereinafter the same applies in this claim)
Place the left speaker position, left sub-speaker, right sub-speaker, and right speaker position on a straight line, provided that the distance from the listener's ears is approximately the same when placed on the straight line. Allow height difference between speakers and between right speaker position and right sub speaker,
The left sub-speaker is arranged to the right at an arbitrary distance g from 0 to 150% of the listener binaural distance with respect to the left speaker position
In addition, the right sub-speaker is arranged at an arbitrary distance g of 0 to 150% of the listener binaural distance with respect to the right speaker position,
When secondary crosstalk is canceled with the above configuration, even if the crosstalk cancellation signal of the right sub-speaker reaches the listener's right ear before the crosstalk of the left speaker system playback sound, it does not cause a reverse phase feeling and is good The right sub-speaker playback sound is limited by the right filter unit within the canceled playback band and volume, and the cross-talk cancellation signal of the left sub-speaker precedes the listener's left ear from the cross-talk of the right speaker system playback sound. Even if it reaches, the left sub-speaker playback sound is limited by the left filter unit within the playback band and volume that do not cause reverse-phase feeling and cancel well, and crosstalk cancellation without secondary crosstalk A crosstalk canceling method characterized by:   4. The crosstalk canceling method according to claim 3, wherein the left sub-speaker is arranged on the right with respect to the left speaker position at a distance substantially equal to a listener binaural distance, and the right sub-speaker is placed on the right speaker position. On the other hand, the crosstalk canceling method is characterized in that it is arranged on the left side at a distance substantially equal to the listener's binaural interval. In the crosstalk canceling method, the distance difference measured by a straight line when the reproduced sound reaches the one ear position of the listener from the left and right speakers of the same distance and the same angle with respect to the listener is represented by a variable a (this claim) And the distance difference added to a when one of the left and right speaker playback sounds wraps around the listener's head when the listener looks at the left and right speakers at a large angle. )
The left audio input signal L is set to W and X, the right audio input signal R is set to Y and Z, W and Z are respectively input to the delay block, and a distance difference (a + aa) with respect to X and Y is input. Delay by an arbitrary distance difference of 10% to 150%, and input X and Y to the filter block to limit the frequency band and output level relative to W and Z,
Y output from the filter block is subtracted from W output from the delay block to obtain a left audio output, and X output from the filter block is subtracted from Z output from the delay block. Right audio output,
When the left and right audio outputs are amplified and reproduced by the left and right speakers arranged at equal angles to the left and right from the listener to cancel the secondary crosstalk, the crosstalk cancellation signal X is transmitted from the W crosstalk to the right of the listener. Even if the crosstalk cancellation signal Y reaches the ear earlier than the crosstalk of the Z and reaches the listener's left ear, it does not cause a sense of reverse phase and falls within the frequency band and output level that are well offset. A crosstalk canceling method characterized by limiting X and Y by a filter block to cancel crosstalk without accompanying secondary crosstalk.   6. The crosstalk canceling method according to claim 5, wherein each of W and Z is input to a delay block, and an arbitrary distance difference h is delayed from 80 to 120% of the distance difference (a + aa) with respect to X and Y. A crosstalk canceling method characterized by: