JP2008048324A - Automatic panning adjusting apparatus and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an automatic panning adjusting apparatus for adjusting the panning deviation of stereo signals that causes a listener to feel a sense of incompatibility. <P>SOLUTION: A residual signal generating section 1L detects a residual signal ERL of the audio signal XL of a left channel and the audio signal XR of a right channel, by adding the residual signal ERL to the audio signal XL, to generate an audio signal YL of the left channel. Furthermore, a residual signal generating section 1R detects a residual signal ELR of the audio signal XR of the right channel, and the audio signal XL of the left channel and adds the residual signal ELR to the audio signal XR, and thereby generating an audio signal YR of the right channel. The automatic panning adjusting apparatus is equipped with an adaptive equalizer ADFL, which performs filtering processing on the audio signal YL and inputs the resulting signal to the residual signal generating section 1L, and an adaptive equalizer ADFR, which performs filtering processing on the audio signal YR and makes the resulting signal input to the residual signal generating section 1R, to make the residual signal ELR much smaller. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an automatic panning adjustment apparatus and an automatic panning adjustment method for adjusting a panning deviation in reproduced sound of a left channel and a right channel when performing stereo reproduction based on a stereo signal recorded in stereo.

  As is well known, in a stereo audio playback apparatus, for example, when stereo playback is performed based on a stereo signal recorded in, for example, two-channel stereo recorded on a recording medium such as a music CD (Compact Disc) or DVD (Digital Versatile Disc). The left channel audio signal and the right channel audio signal are reproduced by separate speakers, or reproduced by a pair of earphones or headphones, so that a realistic reproduced sound can be obtained.

  In other words, since humans have auditory characteristics such as detecting the phase difference and volume difference of the sound entering the left and right ears and grasping the direction of the sound source, stereo recording that takes into account the auditory characteristics When stereo playback is performed by a stereo audio playback device, stereo playback is performed in consideration of human auditory characteristics.

  However, when panning processing (editing) is performed during stereo recording, there is a case where the panning bias occurs in the reproduced sound of the left channel and the right channel, and the listener may feel uncomfortable. For example, when the band performance is recorded in stereo, panning (editing) is set so that the guitar sound can be heard from the left and the vocal sound from the right. Sometimes it was difficult to listen to.

  It is an object of the present invention to provide an automatic panning adjustment apparatus and an automatic panning adjustment method that improve such a panning bias and reduce a listener's uncomfortable feeling.

  The invention according to claim 1 is an automatic panning adjustment device that adjusts the panning bias of a stereo signal having a panning bias, wherein the stereo signal includes a left channel audio signal and a right channel audio signal. First residual signal generating means for generating one residual signal, and adding the first residual signal and the left channel audio signal to adjust the panning bias, the left channel audio signal A second adder for generating a second residual signal of a right channel audio signal and a left channel audio signal of the stereo signal; and The residual signal and the right channel audio signal are added to generate a right channel audio signal in which the panning bias is adjusted. Characterized by comprising a second adder, a.

  According to a fourth aspect of the present invention, there is provided an automatic panning adjustment device for adjusting the panning bias of a stereo signal having a panning bias, wherein the first stereo channel audio signal is adaptively filtered. Adaptive equalizer, first residual signal generating means for generating a first residual signal between the output of the first adaptive equalizer and the audio signal of the right channel, and the first residual A first adder for adding a signal and the left channel audio signal to generate a left channel audio signal in which the panning bias is adjusted, and adaptively filtering the right channel audio signal of the stereo signal A second adaptive equalizer for processing; a second residual signal for generating a second residual signal between the output of the second adaptive equalizer and the left channel audio signal; A second adder for adding the second residual signal and the right channel audio signal to generate a right channel audio signal in which the panning bias is adjusted; And the first adaptive equalizer automatically adjusts a filter characteristic for the audio signal of the input left channel so that the first residual signal is smaller, and the second adaptive equalizer The converter automatically adjusts a filter characteristic for the input audio signal of the right channel so as to make the second residual signal smaller.

  A preferred embodiment of the present invention will be described with reference to FIG. FIG. 1A is a block diagram illustrating a configuration of an automatic panning adjustment device according to the first embodiment, and FIG. 1B is a block diagram illustrating a configuration of an automatic panning adjustment device according to the second embodiment. In FIGS. 1A and 1B, the same or corresponding parts are denoted by the same reference numerals.

[First Embodiment]
In FIG. 1A, the panning automatic adjustment apparatus according to the first embodiment includes residual signal generators 1L and 1R and adders 2L and 2R. For example, information such as a CD, a DVD, and a memory card is provided. It is provided in a stereo audio reproducing apparatus having a sound source such as an information reproducing apparatus that reproduces information recorded on a recording medium and a receiving apparatus connected to a network such as the Internet. Then, a stereo signal (left channel audio signal XL and right channel audio signal XR) output from the sound source selected and operated by the listener is input, and the left channel audio signal YL subjected to panning adjustment is input. The right channel audio signal YR is generated and output to the left channel and right channel speakers or earphones.

  The residual signal generation unit 1L detects a difference (XR-XL) between the audio signal XL of the right channel (R channel) and the audio signal XL of the left channel (L channel), and further calculates the difference (XR-XL). Amplification is performed at a predetermined amplification factor A (where 0 <A <1), and is supplied to the adder 2L as a residual signal ERL.

  The adder 2L adds (synthesizes) the audio signal XL and the residual signal ERL, and outputs the synthesized audio signal YL as an L-channel audio signal obtained by panning adjustment as represented by the following equation (1). To do. However, in Formula (1), t is time.

  The residual signal generator 1R detects the difference (XL-XR) between the audio signals XL and XR, and further amplifies the difference (XL-XR) with a predetermined amplification factor B (where 0 <B <1). The residual signal ELR is supplied to the adder 2R.

  The adder 2R adds (synthesizes) the audio signal XR and the residual signal ELR, and outputs the synthesized audio signal YR as an R-channel audio signal obtained by panning adjustment as represented by the following equation (2). To do. However, in Formula (2), t is time.

  Furthermore, in this embodiment, the amplification factors A and B of the residual signal generators 1L and 1R are fixed to the same amplification factor. By making the amplification factors A and B equal to each other in this way, the change rate of the amplitude of the residual signal ERL and the amplitude of the residual signal ELR is substantially the same (that is, when the amplitudes of ERL and ELR are viewed as absolute values). To almost the same rate of change).

  Then, the audio signal YL output from the adder 2L is reproduced by an L channel speaker or earphone, and the audio signal YR output from the adder 2R is reproduced by an R channel speaker or earphone.

  According to the panning automatic adjustment device having such a configuration, when the audio signals XL and XR supplied from the sound source are original sound signals that generate a sound field having a large panning bias, depending on the degree of the panning bias, By synthesizing the residual signal ERL with the audio signal XL, an L-channel audio signal YL including not only the audio signal XL but also the component of the R-channel audio signal XR is generated, and further, the residual signal ELR is added to the audio signal XR. , The R channel audio signal YR including the components of the L channel audio signal XL as well as the audio signal XR is generated.

  For this reason, the audio signals YL and YR can be made into signals having a correlation, and when reproduced in stereo with speakers or earphones, the reproduced sound can be localized near the center with respect to both ears of the listener. Therefore, a listener's discomfort can be reduced.

  In addition, when the audio signals XL and XR supplied from the sound source generate an original sound signal that generates a sound field with a small panning bias, that is, the audio signals XL and XR are originally highly correlated signals, a residual signal is included in the audio signal XL. By synthesizing ERL, it is possible to generate an audio signal YL that is almost the same as the audio signal XL. Furthermore, by synthesizing the residual signal ELR with the audio signal XR, an audio signal YR that is almost the same as the audio signal XR is generated. Therefore, when stereo playback is performed using speakers or earphones, playback sound that is faithful to the original sound can be played back.

  As described above, the automatic panning adjustment apparatus according to the present embodiment automatically generates the audio signals YL and YR that reduce the panning deviation in accordance with the panning deviation caused by the audio signals XL and XR, and reproduces the sound without a sense of incongruity. When the sound can be reproduced in stereo and the audio signals XL and XR are original sound signals that do not inherently cause panning bias, the panning adjustment is not substantially performed, and the audio signals XL and XR are converted into the audio signals YL and YR. By outputting, the reproduced sound that is faithful to the original sound can be reproduced in stereo, and a sound that does not give the listener a feeling of strangeness can be provided.

  Furthermore, since the automatic panning adjustment apparatus of the present embodiment can be realized with a simple configuration, the circuit scale can be reduced.

  Note that the panning automatic adjustment apparatus of the present embodiment can be formed by an analog circuit when the audio signals XL and XR are analog signals. When the audio signals XL and XR are digital signals that are analog-digital converted at a predetermined sampling frequency fs in accordance with the sampling theorem, the digital signals are formed by a digital circuit and synchronized with the sampling period (1 / fs). (1) Digital signal processing corresponding to (2) may be performed.

  Also, by creating a computer program that demonstrates the functions of the residual signal generators 1L and 1R and the adders 2L and 2R, and causing a digital signal processor (DSP) and a microprocessor (MPU) to execute the computer program, You may make it implement | achieve the panning automatic adjustment apparatus corresponded to the above-mentioned digital circuit.

[Second Embodiment]
Referring to FIG. 1B, the panning automatic adjustment apparatus according to the second embodiment includes residual signal generators 1L and 1R, adders 2L and 2R, and adaptive equalizers ADFL and ADFR, and is stereo. An L channel audio signal XL and an R channel audio signal XR output from a sound source such as an information playback device or a receiver provided in the audio playback device are input, and panning adjustment is performed on the L channel. The audio signal YL and the R channel audio signal YR are generated and output to the left channel and right channel speakers, earphones and the like.

  Here, the panning automatic adjustment apparatus of the present embodiment inputs audio signals XL and XR that have been analog-digital converted into digital signals at a predetermined sampling frequency fs according to the sampling theorem, and performs panning adjustment processing by digital signal processing. It is like that.

  The adaptive equalizer ADFL includes an adaptive digital filter (ADF: Adaptive Digital Filter) that adaptively performs digital filtering on the input audio signal XL, and is an inverse of the sampling frequency fs (1 / fs). The filter coefficient (tap coefficient) of the adaptive digital filter is automatically adjusted so as to make the residual signal ERL, which will be described later, smaller in synchronization with the period τ corresponding to), and the digitally filtered signal AXL is generated. This is supplied to the residual signal generator 1L.

Further, when the above-described adaptive digital filter is formed by a FIR (Finite Impulse Response) digital filter expressed by a transfer function H _L (z) expressed by z conversion shown in the following equation (3), the following equation (4) ), The signal AXL (k) is generated by convolution of the audio signal XL and the impulse response of the adaptive digital filter. Note that t is time, a _k (t) is a filter coefficient, and k is a time step for each period τ.

Further, when an LMS (the Least-Mean-Square) algorithm is applied as an adaptive algorithm for adjusting the characteristics of the adaptive digital filter, the filter coefficient a _k is calculated by the operation represented by the coefficient update equation of the following equation (5). Are sequentially obtained and adjusted. The constant μ is a correction coefficient (0 <μ) for correcting the filter coefficient a _k at each time step, and k is k = 0, 1, 2,..., M−1.

  Next, the residual signal generation unit 1L calculates a difference (XR−AXL) between the R-channel audio signal XR and the signal AXL, and supplies it as a residual signal ERL to the adder 2L and the adaptive equalizer ADFL.

  The adder 2L adds (synthesizes) the audio signal XL and the residual signal ERL, and outputs the synthesized audio signal YL as an L-channel audio signal subjected to panning adjustment.

  Next, the adaptive equalizer ADFR, the residual signal generator 1R, and the adder 2R are formed to have the same configuration as the adaptive equalizer ADFL, the residual signal generator 1L, and the adder 2L, respectively. .

That is, when the adaptive equalizer ADF _R is formed by an FIR digital filter represented by a transfer function H _R (z) in z conversion notation shown in the following equation (6), it is expressed by the following equation (7). Thus, the signal AXR (k) is generated by convolution of the audio signal XR and the impulse response of the adaptive digital filter. Assuming that the LMS algorithm is applied as an adaptive algorithm for adjusting the characteristics of the adaptive digital filter, the filter coefficient b _k is sequentially obtained and adjusted by an operation represented by the coefficient update expression of the following equation (8).

  The residual signal generator 1R calculates a difference (XL-AXR) between the L-channel audio signal XL and the signal AXR, and supplies it as a residual signal ELR to the adder 2R and the adaptive equalizer ADFR.

  The adder 2R adds (synthesizes) the audio signal XR and the residual signal ELR, and outputs the synthesized audio signal YL as an L-channel audio signal subjected to panning adjustment.

  According to the panning automatic adjustment device having such a configuration, the adaptive equalizer ADFL on the L channel side automatically adjusts the characteristics of the adaptive digital filter so that the residual signal ERL becomes a signal of a smaller level, thereby reducing the audio. A more uncorrelated component between the signals YR and YL is extracted as a residual signal ERL by the residual signal generator 1L.

  Therefore, when the audio signals XL and XR are original sound signals that generate a sound field with a large panning bias, that is, the audio signals XL and XR are original sound signals with low correlation, the residual signal ERL includes only the audio signal YR. On the contrary, when the audio signals XL and XR are highly correlated original sound signals, the residual signal ERL converges to a signal of an extremely small level.

  When the adder 2L adds the residual signal ERL and the audio signal XL to generate the audio signal YL, when the audio signals XL and XR are original sound signals having low correlation, the audio signal XR is almost only included. Since the contained components are added to the audio signal XL, the audio signal YL becomes a signal having a high correlation with the audio signal XL, and conversely, the audio signals XL and XR described above are the original sound signals having a high correlation with each other. In some cases, the residual signal ERL is a signal of an extremely small level, so that the audio signal YL is almost the same signal as the audio signal XL.

  In other words, the adaptive equalizer ADFL, the residual signal generator 1L, and the adder 2L on the L channel side perform processing for increasing the correlation only when the correlation between the audio signals XL and XR is low, and perform the audio signal YL. On the contrary, when the correlation between the audio signals XL and XR is originally high, the processing is not substantially performed, and the audio signal XL is almost directly output as the audio signal YL.

  On the other hand, the adaptive equalizer ADFR on the R channel side similarly adjusts the characteristics of the adaptive digital filter so that the residual signal ELR becomes a signal of a lower level, thereby reducing the difference between the audio signals YL and YR. The correlated signal is extracted as a residual signal ELR by the residual signal generator 1R.

  Therefore, when the audio signals XR and XL are original sound signals that generate a sound field with a large panning bias, that is, the audio signals XR and XL are original sound signals with low correlation, the residual signal ELR includes only the audio signal YL. On the contrary, when the audio signals XR and XL are original signals having high correlation, the residual signal ELR converges to an extremely small level signal.

  When the adder 2R adds the residual signal ELR and the audio signal XR to generate the audio signal YR, when the audio signals XR and XL are original sound signals with low correlation, they are included only in the audio signal XL. Therefore, the audio signal YR becomes a signal having a high correlation with the audio signal XL, and conversely, when the audio signals XR and XL are original sound signals having a high correlation, Since the residual signal ELR is an extremely small signal, the audio signal YR is almost the same signal as the audio signal XR.

  That is, the R channel side adaptive equalizer ADFR, the residual signal generator 1R, and the adder 2R perform processing for increasing the correlation only when the correlation between the audio signals XR and XL is low and perform the audio signal YR. On the contrary, when the correlation between the audio signals XR and XL is originally high, the processing is not substantially performed, and the audio signal XR is almost directly output as the audio signal YR.

  When the L-channel and R-channel audio signals YL and YR generated in this way are reproduced in stereo by speakers or earphones, when the correlation between the audio signals XR and XL is low, the audio signal YL having a high correlation is obtained. Since YR is reproduced in stereo by a speaker or earphone, the reproduced sound can be localized near the center with respect to both ears of the listener, and the listener's uncomfortable feeling can be reduced. Conversely, when the audio signals XL and XR are originally highly correlated, the audio signals YL and YR are also highly correlated. Can be provided.

  The panning automatic adjustment apparatus of the present embodiment may form the configuration shown in FIG. 2B with a digital circuit, but a computer program is created to create a digital signal processor (DSP) or a microprocessor (MPU). By executing this computer program, an automatic panning adjustment device corresponding to a digital circuit may be realized.

  Next, more specific examples according to the first embodiment shown in FIG. 1A will be described with reference to FIGS. 2A and 2B, the same or corresponding parts as those in FIG. 1A are denoted by the same reference numerals. 2A is a circuit diagram showing a basic configuration of the present embodiment, and FIG. 2B is a circuit diagram showing a configuration as a modified example.

[Automatic panning adjustment device shown in FIG. 2 (a)]
In the automatic panning adjustment apparatus shown in FIG. 2A, the L channel residual signal generator 1L subtracts the L channel audio signal XL from the R channel audio signal XR, and the subtractor. The difference (XR−XL) output from the SBL is provided with an amplifier A that amplifies at a predetermined amplification factor A (where 0 <A <1). Then, the adder 2L adds (synthesizes) the residual signal ERL output from the amplifier A and the audio signal XL to generate an L-channel audio signal YL subjected to panning adjustment.

  The residual signal generator 1H on the R channel side performs predetermined amplification on a subtracter SBR that subtracts the R channel audio signal XR from the L channel audio signal XL and a difference (XL−XR) output from the subtractor SBR. And an amplifier B that amplifies at a rate B (where 0 <B <1). The amplification factor of the amplifier B and the amplification factor of the amplifier A may be the same or different, but it is preferable to set the same amplification factor (A = B).

  The adder 2R adds (synthesizes) the residual signal ELR and the audio signal XR output from the amplifier B, thereby generating an R channel audio signal YR subjected to panning adjustment.

  According to the panning automatic adjustment device having such a configuration, the same panning adjustment function as that described in the first embodiment is exhibited, and the listener's discomfort can be reduced.

  In other words, when the audio signals XL and XR supplied from the sound source are original sound signals that generate a sound field with a large panning bias, the residual signal ERL is synthesized with the audio signal XL, thereby not only the audio signal XL. By generating an L-channel audio signal YL including a component of the R-channel audio signal XR, and further synthesizing the residual signal ELR with the audio signal XR, not only the audio signal XR but also the L-channel audio signal XL is generated. An R channel audio signal YR including the components is generated.

  For this reason, the audio signals YL and YR can be made into signals having a correlation, and when reproduced in stereo with speakers or earphones, the reproduced sound can be localized near the center with respect to both ears of the listener. Therefore, a listener's discomfort can be reduced.

  In addition, when the audio signals XL and XR supplied from the sound source generate an original sound signal that generates a sound field with a small panning bias, that is, the audio signals XL and XR are originally highly correlated signals, a residual signal is included in the audio signal XL. By synthesizing ERL, it is possible to generate an audio signal YL that is almost the same as the audio signal XL. Furthermore, by synthesizing the residual signal ELR with the audio signal XR, an audio signal YR that is almost the same as the audio signal XR is generated. Therefore, when stereo reproduction is performed using a speaker or earphone, reproduction sound that is faithful to the original sound can be reproduced, and sound that does not give the listener a sense of incongruity can be provided.

  Furthermore, the amplifiers A and B may be formed by variable amplifiers that change in conjunction with the same amplification factor, and the listener or the like may perform adjustment operations. With such a configuration, it is possible to perform panning adjustment suited to the subjectivity (preference) of the listener or the like.

[Automatic panning adjustment device shown in FIG. 2 (b)]
In the automatic panning adjustment apparatus shown in FIG. 2B, a predetermined amplification factor a ₀ (where 0 <a ₀ <1) is applied to the input stage of the subtracter SBL in the automatic panning adjustment apparatus shown in FIG. an amplifier a ₀ having, subtractor predetermined amplification factor in the input stage of SBR b ₀ (however, 0 <b ₀ <1, and a ₀ = b ₀₎ is configured to have an amplifier b ₀ is provided with Yes.

Then, the signal obtained by amplifying the audio signal XL by the amplifier a ₀ is supplied to the subtracter SBL, and the signal obtained by amplifying the audio signal XR by the amplifier b ₀ is supplied to the subtracter SBR.

  According to the panning automatic adjustment device having such a configuration, the same panning adjustment function as that of the first embodiment and the above-described automatic panning adjustment device shown in FIG. 2A is exhibited, thereby reducing the listener's uncomfortable feeling. it can.

Further, in the panning automatic adjustment apparatus shown in FIG. 2B, the amplitudes of the audio signals XL and XR are adjusted by the amplifiers a ₀ and b ₀ and then supplied to the subtracters SBL and SBR. It is possible to finely adjust the correlation between the audio signal XR of the opposite channel with respect to the signal XL and the correlation between the audio signal XL of the opposite channel with respect to the audio signal XR in the residual signal ELR.

In addition, the amplifiers A and B are formed by variable amplifiers that change in conjunction with the same amplification factor, and the amplifiers a ₀ and b ₀ are also formed by variable amplifiers that change in association with the same amplification factor. By adopting a configuration in which the adjustment operation is performed, it is possible for the listener or the like to perform panning adjustment that is more subjective (preferred).

  Next, a more specific example according to the second embodiment shown in FIG. 1B will be described with reference to FIG. In FIG. 3, the same or corresponding parts as those in FIG.

In the automatic panning adjustment apparatus of the present embodiment, the adaptive equalizer ADFL on the left channel side receives the audio signal XL and shifts in synchronization with the sampling period τ, and includes a shift register having m stages of delay circuits DL1 to DLm. , FIR with the multiplier a ₀ ~a _m for multiplying the filter coefficients a ₀ ~a _m input or output of the delay circuit DL1 to DLm, and an adder SUML for adding outputs of the multipliers a ₀ ~a _m in addition to comprising a digital filter, by performing coefficient updating operation expressed by the formula (5), configured to include an estimation controller EML updating adjusting the filter coefficients a ₀ ~a _m by the LMS algorithm Has been.

  Here, the shift register of this embodiment includes eight stages (m = 8) of delay circuits DL1 to DLm.

Further, the residual signal generator 1L includes a delay circuit DRN that outputs the right-channel audio signal XR with a delay time (z ⁻ⁿ ) that is n times the sampling period τ, and an adder from the output of the delay circuit DRN. And a subtracter SBL for subtracting the output signal AXL of the SUML. Here, the delay time (z ⁻ⁿ ) of the delay circuit DRN is set to a longer delay time than the total delay time (z ^−m ) of the shift register having the m-stage delay circuits DL1 to DLm described above. Yes. By making the delay time (z ⁻ⁿ ) longer than the delay time (z ^−m ) in this way, the component of the right channel audio signal XR is reproduced when the reproduced sound is reproduced in stereo based on the audio signal YL described later. This makes it possible to reproduce sounds with a sense of depth.

The adaptive equalizer ADFL on the right channel side has the same configuration as the adaptive equalizer ADFR, and has m stages (m = 8) of delay circuits that receive the audio signal XR and shift in synchronization with the sampling period τ. adding a shift register having a DR1～DRm, a multiplier b ₀ ~b _m for multiplying the filter coefficients b ₀ ~b _m input or output of the delay circuit DR1～DRm, the output of the multiplier b ₀ ~b _m In addition to the FIR digital filter including the adder SUMR, the coefficient update calculation represented by the above equation (8) is performed to update and adjust the filter coefficients b _{0 to} b _m by the LMS algorithm. And a control unit EMR.

  Here, the shift register of this embodiment includes eight stages (m = 8) of delay circuits DL1 to DLm.

Further, the residual signal generator 1R includes a delay circuit DLN that outputs the left channel audio signal XL with a delay time (z ⁻ⁿ ) that is n times the sampling period τ, and an adder from the output of the delay circuit DLN. And a subtracter SBR that subtracts the output signal AXR of the SUMR. Here, the delay time (z ⁻ⁿ ) of the delay circuit DLN is set to be longer than the total delay time (z ^−m ) of the shift register having the m-stage delay circuits DR1 to DRm described above. Yes. By making the delay time (z ⁻ⁿ ) longer than the delay time (z ^−m ) in this way, the component of the left channel audio signal XL is reproduced when the reproduced sound is reproduced in stereo based on the audio signal YR described later. This makes it possible to reproduce sounds with a sense of depth.

  Then, the adder 2L on the L channel side adds (synthesizes) the residual signal ERL and the audio signal XL to generate an L channel audio signal YL that has undergone panning adjustment, and the adder 2R on the R channel side However, the residual signal ELR and the audio signal XR are added (synthesized) to generate an R-channel audio signal YR subjected to panning adjustment.

  According to the automatic panning adjustment apparatus of the present example having such a configuration, the same panning adjustment function as that described in the second embodiment is exhibited, and the listener's discomfort can be reduced.

That is, in the adaptive equalizer ADFL, estimation control unit EML that is automatically adjusting the filter coefficients a ₀ ~a _m of the adaptive digital filter to a smaller level of the signal a residual signal ERL, from the audio signal YL The signal AXL correlated with the audio signal YR is extracted and supplied to the subtracter SBL. Therefore, in the subtractor SBL, a more uncorrelated component between the audio signals YR and YL is generated as the residual signal ERL.

  Therefore, when the audio signals XL and XR are original sound signals that generate a sound field with a large panning bias, that is, the audio signals XL and XR are original sound signals with low correlation, the residual signal ERL includes only the audio signal YR. On the contrary, when the audio signals XL and XR are highly correlated original sound signals, the residual signal ERL converges to a signal of an extremely small level.

  When the adder 2L adds the residual signal ERL and the audio signal XL to generate the audio signal YL, when the audio signals XL and XR are original sound signals having low correlation, the audio signal XR is almost only included. Since the contained component is added to the audio signal XL, the audio signal YL becomes a signal having a high correlation with the audio signal XR, and conversely, the above-described audio signals XL and XR are the original sound signals having a high correlation with each other. In some cases, the residual signal ERL is a signal of an extremely small level, so that the audio signal YL is almost the same signal as the audio signal XL.

  In other words, the adaptive equalizer ADFL, the residual signal generator 1L, and the adder 2L on the L channel side perform processing for increasing the correlation only when the correlation between the audio signals XL and XR is low, and perform the audio signal YL. On the contrary, when the correlation between the audio signals XL and XR is originally high, the processing is not substantially performed, and the audio signal XL is almost directly output as the audio signal YL.

On the other hand, the adaptive equalizer ADFR on the R channel side also automatically adjusts the filter coefficients b _{0 to} b _m of the adaptive digital filter so that the estimation control unit EMR makes the residual signal ELR a signal of a lower level. Thus, a signal AXL having a correlation with the audio signal YL is extracted from the audio signal YR and supplied to the subtracter SBR. For this reason, in the subtractor SBR, a more uncorrelated component between the audio signals YL and YR is generated as the residual signal ELR.

  Therefore, when the audio signals XL and XR are original sound signals that generate a sound field with a large panning bias, that is, the audio signals XL and XR are original sound signals with low correlation, the residual signal ELR includes only the audio signal YL. On the contrary, when the audio signals XR and XL are original signals having high correlation, the residual signal ELR converges to an extremely small level signal.

  When the adder 2R adds the residual signal ELR and the audio signal XR to generate the audio signal YR, when the audio signals XR and XL are original sound signals with low correlation, they are included only in the audio signal XL. Therefore, the audio signal YR becomes a signal having a high correlation with the audio signal XL, and conversely, when the audio signals XR and XL are original sound signals having an originally high correlation, Since the residual signal ELR is an extremely small signal, the audio signal YR is almost the same signal as the audio signal XR.

  That is, the R channel side adaptive equalizer ADFR, the residual signal generator 1R, and the adder 2R perform processing for increasing the correlation only when the correlation between the audio signals XR and XL is low and perform the audio signal YR. On the contrary, when the correlation between the audio signals XR and XL is originally high, the processing is not substantially performed, and the audio signal XR is almost directly output as the audio signal YR.

  When the L-channel and R-channel audio signals YL and YR generated in this way are reproduced in stereo by speakers or earphones, when the correlation between the audio signals XR and XL is low, the audio signal YL having a high correlation is obtained. Since YR is reproduced in stereo by a speaker or earphone, the reproduced sound can be localized near the center with respect to both ears of the listener, and the listener's uncomfortable feeling can be reduced. Conversely, when the audio signals XL and XR are originally highly correlated, the audio signals YL and YR are also highly correlated, so that sound that does not give the listener a sense of incongruity can be reproduced by stereo playback of the reproduced sound that is faithful to the original sound Can be provided.

It is a block diagram showing the structure of the panning automatic adjustment apparatus which concerns on 1st, 2nd embodiment of this invention. 1 is a circuit diagram illustrating a configuration of an automatic panning adjustment device according to a first embodiment. FIG. 5 is a circuit diagram illustrating a configuration of an automatic panning adjustment apparatus according to a second embodiment.

Explanation of symbols

1L, 1R: Residual signal generator 2L, 2R: Adder SBL, SBR: Subtractor A, B, a ₀ , b ₀ ... Amplifier ADFL, ADFR: Adaptive equalizer DRN, DLN ... Delay circuit

Claims (8)

A panning automatic adjustment device for adjusting the panning bias of a stereo signal having panning bias,
First residual signal generating means for generating a first residual signal of a left channel audio signal and a right channel audio signal of the stereo signal;
A first adder for adding the first residual signal and the left-channel audio signal to generate a left-channel audio signal in which the panning bias is adjusted;
Second residual signal generating means for generating a second residual signal of a right channel audio signal and a left channel audio signal of the stereo signal;
A second adder that adds the second residual signal and the right-channel audio signal to generate a right-channel audio signal in which the panning bias is adjusted;
An automatic panning adjustment apparatus comprising: The first residual signal generation means includes first amplification means having a predetermined amplification factor that adjusts the amplitude of the first residual signal and supplies the first residual signal to the first adder.
The second residual signal generating means includes second amplifying means having a predetermined amplification factor that adjusts the amplitude of the second residual signal and supplies the second residual signal to the second adder;
The panning automatic adjustment apparatus according to claim 1, wherein: The first residual signal generating means further includes third amplifying means for amplifying and inputting the left channel audio signal at a predetermined amplification rate,
The second residual signal generating means further comprises fourth amplifying means for amplifying and inputting the right channel audio signal at a predetermined amplification rate;
The panning automatic adjustment device according to claim 2, wherein: A panning automatic adjustment device for adjusting the panning bias of a stereo signal having panning bias,
A first adaptive equalizer for adaptively filtering an audio signal of a left channel of the stereo signal;
First residual signal generating means for generating a first residual signal between the output of the first adaptive equalizer and the audio signal of the right channel;
A first adder for adding the first residual signal and the left-channel audio signal to generate a left-channel audio signal in which the panning bias is adjusted;
A second adaptive equalizer that adaptively filters the audio signal of the right channel of the stereo signal;
Second residual signal generating means for generating a second residual signal between the output of the second adaptive equalizer and the audio signal of the left channel;
A second adder that adds the second residual signal and the right channel audio signal to generate a right channel audio signal in which the panning bias is adjusted;
The first adaptive equalizer automatically adjusts a filter characteristic for an input audio signal of the left channel so as to make the first residual signal smaller,
The second adaptive equalizer automatically adjusts a filter characteristic of the input right channel audio signal so as to make the second residual signal smaller;
An automatic panning adjustment device. The first and second adaptive equalizers automatically adjust the filter characteristics according to an LMS algorithm;
The panning automatic adjustment device according to claim 4, wherein: The first residual signal generating means includes first delay means for inputting the right channel audio signal with a predetermined time delay, and outputs the first adaptive equalizer and the first delay. Generating the first residual signal with the output of the means;
The second residual signal generation means includes second delay means for inputting the left channel audio signal with a predetermined time delay, and outputs the second adaptive equalizer output and the second delay. Generating the second residual signal with the output of the means;
The panning automatic adjustment apparatus according to claim 4 or 5, characterized in that: A panning automatic adjustment method for adjusting the panning bias of a stereo signal having panning bias,
A first residual signal generating step of generating a first residual signal of a left channel audio signal and a right channel audio signal of the stereo signal;
A first addition step of adding the first residual signal and the left channel audio signal to generate a left channel audio signal in which the panning bias is adjusted;
A second residual signal generating step of generating a second residual signal of a right channel audio signal and a left channel audio signal of the stereo signal;
A second adding step of adding the second residual signal and the right channel audio signal to generate a right channel audio signal in which the panning bias is adjusted;
A panning automatic adjustment method comprising: A panning automatic adjustment method for adjusting the panning bias of a stereo signal having panning bias,
A first adaptive equalization step of adaptively filtering the audio signal of the left channel of the stereo signal;
A first residual signal generation step of generating a first residual signal between the output generated in the first adaptive equalization step and the right channel audio signal;
A first addition step of adding the first residual signal and the left channel audio signal to generate a left channel audio signal in which the panning bias is adjusted;
A second adaptive equalization step of adaptively filtering the right channel audio signal of the stereo signal;
A second residual signal generation step for generating a second residual signal between the output generated in the second adaptive equalization step and the left channel audio signal;
A second addition step of adding the second residual signal and the right channel audio signal to generate a right channel audio signal in which the panning bias is adjusted;
In the first adaptive equalization step, a filter characteristic for the left-channel audio signal to be input is automatically adjusted so as to make the first residual signal smaller;
The second adaptive equalization step automatically adjusts a filter characteristic of the input right channel audio signal so as to make the second residual signal smaller;
Panning automatic adjustment method characterized by

Images