JP2011147031A - Musical tone signal processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a musical tone signal processing apparatus, capable of freely expanding or contracting an acoustic image of a desired musical tone in inputted musical tones. <P>SOLUTION: According to a musical tone signal processing apparatus, when a second reference localization, a first function stipulating expansion of the boundary of one end of a direction range and a second function stipulating expansion of the boundary of another end of the direction range are set by a second setting means, an output direction indicated by localization information of an extraction signal that exists in the direction range in which the conditions have been set in this manner is shifted by an acoustic image scaling means. Thus, the acoustic image within the direction range is expanded or contracted. Therefore, the acoustic image formed by the extraction signal extracted from the set direction range is expanded or contracted by shifting the output direction indicated by localization information for each extraction signal extracted from within the set direction range, thereby freely expanding or contracting each acoustic image indicated by a stereo sound source. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a musical tone signal processing apparatus, and more particularly to a musical tone signal processing apparatus capable of freely expanding or reducing a desired musical tone image in an input musical tone.

  Japanese Patent Laid-Open No. 2000-504526 discloses that a monaural signal is divided into left and right according to a frequency band so that even a monaural sound source is output with a slight spread as a pseudo stereo sound. A possible device is described.

  Patent Document 2 (Japanese Patent Laid-Open No. 08-123410) discloses an acoustic effect adding device that can give a sense of spread to the sound of each sound source composed of left and right channels. In such an acoustic effect adding device, specifically, the timbre data of each frequency band given by time division for each of the left and right channels is adjusted by adjusting the localization for each frequency band and reversing the phase, etc. By blurring the feeling, the sound is spread.

Special Table 2000-504526 Japanese Patent Laid-Open No. 08-123410

  However, the conventional methods such as Patent Document 1 and Patent Document 2 have a problem that it is difficult to freely expand the sound image of the input sound source as desired.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide a musical tone signal processing apparatus capable of freely expanding or reducing a desired musical tone image in an input musical tone.

Means for Solving the Problems and Effects of the Invention

  In order to achieve this object, according to the musical sound signal processing device of the first aspect, the second setting means defines the second reference localization and the degree of expansion of the boundary on one end side of the direction range by the first setting means. When a function and a second function that defines the degree of enlargement of the boundary on the other end side of the direction range are set, the localization information of the extracted signal existing in the direction range in which the condition is set indicates The output direction is moved by the sound image enlargement / reduction means, whereby the sound image within the direction range is enlarged or reduced. Specifically, with respect to the direction range in which the second reference localization, the first function, and the second function are set, out of the extracted signals extracted by the extraction means from within the direction range, the second reference localization is more than the second reference localization. The output direction indicated by the localization information of the first extraction signal located at one end of the direction range is moved by the linear mapping according to the first function based on the second reference localization by the sound image enlargement / reduction means. In addition, the output direction indicated by the localization information of the second extracted signal located on the other end side of the direction range with respect to the second reference localization is the same as that of the second reference localization with the second reference localization as a reference. It is moved by a linear mapping according to a function of 2. It should be noted that the setting of each condition (second reference localization, first function, second function) by the second setting means in claim 1 is variably set in advance or fixedly set. It is intended to include both cases. Further, the “magnification” in claim 1 includes not only enlargement but also reduction in accordance with the degree.

  Therefore, the output signal indicated by the localization information of each extraction signal extracted from the set direction range is moved, that is, the extraction signal extracted from the set direction range by moving the localization of each extraction signal. Since the sound image formed by the method is enlarged or reduced, each sound image indicated by the stereo sound source can be freely enlarged or reduced.

  According to the musical sound signal processing device according to the second, third, and fourth aspects, in addition to the effect produced by the musical sound signal processing device according to the first aspect, the following effect is obtained. The extraction unit, when the determination unit determines that the output direction indicated by the localization information is within the direction range in the output direction set by the setting unit, the frequency band corresponding to the localization information determined to be included The signal of each channel is extracted as an extraction signal from the tone signal for each direction range used for the determination. In this way, the extraction means can extract the extraction signal from the signal of each channel for each set direction range, that is, for each desired localization. Here, since the second setting means can set the second reference localization, the first function, and the second function for each direction range set by the setting means, the second setting means can set each direction range set by the setting means. The effect is that the sound image can be enlarged or reduced independently. In addition, the extraction means extracts signals for each set direction range from the signals of each channel, so after performing signal processing on the extracted signals (extraction signals), those signals (signal processing) It is possible to re-synthesize the extracted signal).

  According to the musical tone signal processing device of the fifth aspect, in addition to the effect produced by the musical tone signal processing device according to the first aspect, the following effect is produced. When the extraction signal to be processed is a monaural signal, the output direction indicated by the localization information of the extraction signal is a preset continuous output direction at one end or the other end of the direction range including the extraction signal. It is automatically distributed by the preparation means so as to alternate for each frequency range to be performed. Then, a linear map using the second reference localization, the first function, and the second function set for the direction range including the extracted signal is extracted from the extracted signal distributed by the preparation means. By means. In addition, in claim 5, “when the extraction signal to be processed is a monaural signal” means that the input tone signal is a monaural signal (one-channel signal) and that the input tone signal is a signal of a plurality of channels. However, it is intended to include both the case where the signal is synthesized into a monaural signal.

  Therefore, even if the extraction signal to be processed is a monaural signal, the output direction (localization) of each extraction signal is automatically distributed to any boundary of the direction range in advance by the preparation unit. The localization can be moved by the linear mapping using the localization, the first function, and the second function. As a result, there is an effect that the sound image can be spread. In addition, since the localization by the preparation unit is performed so as to alternate in units of groups for each predetermined frequency, there is an effect that it is possible to have a well-balanced feeling of spread.

  According to the musical tone signal processing device of the sixth aspect, in addition to the effect produced by the musical tone signal processing device according to any one of the second to fifth aspects, the following effect is produced. The signal processing means distributes the signal of each channel to be processed according to the output channel, and performs independent signal processing on each of the distributed signals. Since a plurality of output means are provided corresponding to the above independent signal processing, after extracting the extraction signal for each desired localization, the extraction signal at the desired localization (that is, one localization) is distributed. In addition, after each of the distributed signals is subjected to independent signal processing, they can be output from separate output means. At this time, the second setting means, which is a part of the signal processing means, can set the second reference localization, the first function, and the second function for each signal distributed according to the output channel. The reduction process is performed independently for each distributed signal using the second reference localization set for each signal distributed according to the output channel, the first function, and the second function. Therefore, there is an effect that the sound image can be appropriately enlarged or reduced according to the arrangement of the output means.

  According to the musical sound device of claim 7, when a monaural signal is input, an output direction of the extraction signal (an output direction equal to a predetermined first reference localization) is a direction range corresponding to the extraction signal. In the output direction of one end or the other end, the preparatory means automatically distributes the signals so as to alternate every predetermined continuous frequency range. Then, the second reference localization, the first function, and the second function set by the second setting unit with respect to the direction range corresponding to the extraction signal are used for the extraction signal distributed by the preparation unit. The linear mapping is performed by the sound image enlargement / reduction means, whereby the sound image within the direction range is enlarged or reduced.

  Therefore, even in the case of an input signal monaural signal, the output direction (localization) of each extracted signal is automatically distributed to any boundary of the direction range in advance by the preparation means, so the second reference localization and the first The localization can be moved by the linear mapping using the function and the second function. As a result, there is an effect that the sound image can be spread. In addition, since the localization by the preparation unit is performed so as to alternate in units of groups for each predetermined frequency, there is an effect that it is possible to have a well-balanced feeling of spread.

It is the block diagram which showed the effector (an example of a musical tone signal processing apparatus) which is one Embodiment of this invention. It is the figure which showed typically the process performed with DSP using a functional block. It is the figure which showed the process performed by each part S10, S20, S60-S90. It is the figure which showed the process performed by main process part S30. It is the figure which showed the detail of the process performed by 1st extraction process S100, 1st signal process S110, 2nd extraction process S200, and 2nd signal process S210. It is the figure which showed the detail of the process performed by process S300 which takes out other than designation | designated, and signal process S310 which is not designated. (A), (b) is a graph for demonstrating each coefficient determined by the localization w [f] and the target localization. It is the schematic diagram which showed a mode that a sound image was expanded or reduced by the sound image expansion / contraction process in 2nd Embodiment. It is the figure which showed the detail of the process performed by 1st signal processing S110 and 2nd signal processing S210 of 2nd Embodiment. It is a schematic diagram for demonstrating the outline | summary of the sound image expansion / contraction process in 3rd Embodiment. It is the figure which showed the main processes performed with the effector of 3rd Embodiment.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. First, a first embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a block diagram showing an effector 1 (an example of a musical tone signal processing apparatus) according to the first embodiment of the present invention. The effector 1 extracts a musical tone signal (hereinafter referred to as an “extraction signal”) to be subjected to signal processing over a plurality of conditions (each of a set of frequency, localization and maximum level). Is something that can be done.

  The effector 1 includes an Lch analog-digital converter (hereinafter referred to as “A / D converter”) 11L, an Rch A / D converter 11R, a digital signal processor (hereinafter referred to as “DSP”) 12, and an Lch A first digital-analog converter (hereinafter referred to as “D / A converter”) 13L1, an Rch first D / A converter 13R1, an Lch second D / A converter 13L2, and an Rch second D / A converter 13R2. The CPU 14, the ROM 15, the RAM 16, and the bus line 17 are included. The units 11 to 16 are electrically connected to each other via the bus line 17.

  The Lch A / D converter 11L converts the left channel signal (part of the musical tone signal) input from the IN_L terminal from an analog signal to a digital signal, and converts the digitized left channel signal to the bus line 17. It is a converter which outputs to DSP12 via this. The Rch A / D converter 11R converts the right channel signal (a part of the musical sound signal) input from the IN_R terminal from an analog signal to a digital signal, and converts the digitized right channel signal to the bus line. 17 is a converter that outputs the signal to the DSP 12 via 17.

  When the DSP 12 receives the left channel signal output from the Lch A / D converter 11L and the right channel signal output from the Rch A / D converter 11R, the DSP 12 performs signal processing on the input left channel signal and right channel signal. The left channel signal and the right channel signal subjected to the signal processing are converted into an Lch first D / A converter 13L1, an Rch first D / A converter 13R1, an Lch second D / A converter 13L2, and an Rch second D / A. It is a processor that outputs to the A converter 13R2.

  The Lch first D / A converter 13L1 and the Lch second D / A converter 13L2 convert the left channel signal subjected to signal processing by the DSP 12 from a digital signal to an analog signal, and convert the analog signal to the main speaker. It is a converter that outputs to terminals (OUT1_L terminal, OUT2_L terminal) to which the L channel side (not shown) is connected. It should be noted that the left channel signal that is output after the DSP 12 performs independent signal processing is input to the Lch first D / A converter 13L1 and the Lch second D / A converter 13L2.

  The first D / A converter 13R1 for Rch and the second D / A converter 13R2 for Rch convert the right channel signal, which has been subjected to signal processing by the DSP 12, from a digital signal to an analog signal, and convert the analog signal to the main speaker. This is a converter for outputting to terminals (OUT1_R terminal, OUT2_R terminal) to which the R channel side (not shown) is connected. The Rch first D / A converter 13R1 and the Rch second D / A converter 13R2 each receive a right channel signal that has been subjected to independent signal processing by the DSP 12 and output.

  The CPU 14 is a central control device that controls the units 11L to 13R2, 15, and 16. The ROM 15 is a non-rewritable memory that stores a control program executed by the effector 1. Processing performed by the DSP 12 to be described later with reference to FIGS. 2 to 6 is stored in the ROM 15 as a control program. The RAM 16 is a memory for temporarily storing various data.

  Next, the processing of the DSP 12 will be described with reference to FIG. FIG. 2 is a diagram schematically showing processing executed by the DSP 12 using functional blocks. The DSP 12 has a first processing unit S1 and a second processing unit S2 as functional blocks. The DSP 12 repeatedly executes the process shown in FIG. 2 while the effector 1 is powered on.

  The DSP 12 receives a left channel signal (hereinafter referred to as “IN_L [t] signal”) input from the IN_L terminal in the time domain. [T] indicates that the signal is indicated in the time domain, and the IN_R terminal. The right channel signal (hereinafter referred to as “IN_R [t] signal”) in the time domain input from is input, and the processes in the first processing unit S1 and the second processing unit S2 are executed.

  Here, the processing in the first processing unit S1 and the second processing unit S2 is the same processing and is executed at predetermined intervals, but the processing in the second processing unit S2 is the first processing. Execution is started with a predetermined time delay from the start of execution of processing in the part S1. Thereby, the process in the second processing unit S2 supplements the joint from the end of the execution of the process in the first processing unit S1 to the start of the execution, and the process in the first processing unit S1 ends the execution of the process in the second processing unit S2 It supplements the joint from the start to the execution. Therefore, a signal obtained by combining the signal generated by the first processing unit S1 and the signal generated by the second processing unit S2, that is, the first left channel signal in the time domain (hereinafter, “ OUT1_L [t] signal ”, a first right channel signal in the time domain (hereinafter referred to as“ OUT1_R [t] signal ”), and a second left channel signal in the time domain (hereinafter referred to as“ OUT2_L [t] ”). The second right channel signal in the time domain (hereinafter referred to as “OUT2_R [t] signal”) is prevented from becoming discontinuous. In the present embodiment, the first processing unit S1 and the second processing unit S2 are executed every 0.1 second, and the processing in the second processing unit S2 is 0 from the start of the processing in the first processing unit S1. .Start execution after 05 seconds. However, the execution interval between the first processing unit S1 and the second processing unit S2 and the delay time from the start of the processing in the first processing unit S1 to the start of the processing in the second processing unit S2 are 0.1 seconds. The value according to the sampling frequency and the number of musical sound signals can be appropriately used.

  The first processing unit S1 and the second processing unit S2 are Lch analysis processing unit S10, Rch analysis processing unit S20, main processing unit S30, L1ch output processing unit S60, R1ch output processing unit S70, and L2ch output processing unit as functional blocks. S80 and R2ch output processing unit S90 are provided.

  The Lch analysis processing unit S10 refers to the input IN_L [t] signal as a left channel signal (hereinafter, “IN_L [f] signal”) indicated in the frequency domain. [F] indicates that the signal is indicated in the frequency domain. Rch analysis processing unit S20 converts the input IN_R [t] signal into a right channel signal (hereinafter referred to as “IN_R [f] signal”) shown in the frequency domain. To be output. Details of the Lch analysis processing unit S10 and the Rch analysis processing unit S20 will be described later with reference to FIG.

  The main processing unit S30 performs first signal processing, second signal processing, which will be described later, on the IN_L [f] signal input from the Lch analysis processing unit S10 and the IN_R [f] signal input from the Rch analysis processing unit S20. In addition, processing other than the designation is performed, and the left channel signal and the right channel signal shown in the frequency domain are output based on the output result from each processing. Details of the processing of the main processing unit S30 will be described later with reference to FIGS.

  When one of the left channel signals (hereinafter referred to as “OUT_L1 [f] signal) indicated by the frequency domain output from the main processing unit S30 is input to the L1ch output processing unit S60, the OUT_L1 [f] The signal is converted into a left channel signal (hereinafter referred to as “OUT1_L [t] signal”) shown in the time domain. When one of the right channel signals (hereinafter referred to as “OUT_R1 [f] signal) output in the frequency domain output from the main processing unit S30 is input, the R1ch output processing unit S70 receives the OUT_R1 [f]. The signal is converted into a right channel signal (hereinafter referred to as “OUT1_R [t] signal”) shown in the time domain.

  When an L2ch output processing unit S80 receives a left channel signal (hereinafter referred to as an “OUT_L2 [f] signal) output from the main processing unit S30 and indicated by another frequency domain, the OUT_L2 [f] The signal is converted into a left channel signal (hereinafter referred to as “OUT2_L [t] signal”) indicated in the time domain. The R2ch output processing unit S90 receives the OUT_R2 [f] when a right channel signal (hereinafter referred to as an “OUT_R2 [f] signal) output from the main processing unit S30 and indicated by another frequency domain is input. The signal is converted into a right channel signal (hereinafter referred to as “OUT2_R [t] signal”) shown in the time domain. Details of the processes of the L1ch output processing unit S60, the R1ch output processing unit S70, the L2ch output processing unit S80, and the R2ch output processing unit S90 will be described later with reference to FIG.

  The OUT_L1 [f] signal, the OUT_R1 [f] signal, the OUT_L2 [f] signal and the OUT_R2 [f] signal output from the output processing units S60 to S90 of the first processing unit S1, and the second processing unit S2 The OUT_L1 [f] signal, the OUT_R1 [f] signal, the OUT_L2 [f] signal, and the OUT_R2 [f] signal output from each of the output processing units S60 to S90 are respectively combined by crossfading.

  Next, referring to FIG. 3, the Lch analysis processing unit S10, the Rch analysis processing unit S20, the L1ch output processing unit S60, the R1ch output processing unit S70, the L2ch output processing unit S80, and the R2ch output processing, excluding the main processing unit 30, are performed. Details of the processing executed in unit S90 will be described. FIG. 3 is a diagram illustrating processing executed in each unit S10, S20, and S60 to S90.

  First, the Lch analysis processing unit S10 and the Rch analysis processing unit S20 will be described. In the Lch analysis processing unit S10, first, a window function process that is a process of applying a Hanning window to the IN_L [t] signal is executed (S11). Thereafter, Fast Fourier Transform (FFT) is performed on the IN_L [t] signal (S12). By this fast Fourier transform, the IN_L [t] signal is converted to an IN_L [f] signal (a spectrum signal having the horizontal axis of each frequency f subjected to Fourier transform). The IN_L [f] signal is expressed by an expression having a real part and an imaginary part (hereinafter referred to as “complex expression”). Here, the reason why the Hanning window is applied to the IN_L [t] signal in the process of S11 is to reduce the influence of the start point and the end point in the input IN_L [t] signal on the fast Fourier transform. .

  After the processing of S12, in the Lch analysis processing unit S10, the level of the IN_L [f] signal (hereinafter referred to as “INL_Lv [f]”) and the phase of the IN_L [f] signal (hereinafter referred to as “INL_Ar [f]”). ) Is calculated for each frequency f (S13). Specifically, INL_Lv [f] is a sum of a value obtained by squaring the real part of the complex expression of the IN_L [f] signal and a value obtained by squaring the imaginary part of the complex expression of the IN_L [f] signal. It is obtained by calculating the square root of the sum of the values. Further, INL_Ar [f] is obtained by calculating an arc tangent (tan ^ (-1)) of a value obtained by dividing the imaginary part in the complex expression of the IN_L [f] signal by the real part. After the process of S13, the process proceeds to the process of the main processing unit S30.

  In the Rch analysis processing unit S20, the processes of S21 to S23 are performed on the IN_R [t] signal. Note that the processing of S21 to S23 is the same as the processing of S11 to S13, and is different from the processing of S11 to S13 in that the processing target is the IN_R [t] signal, and the IN_R [f] signal is This is the output point. Therefore, the detailed description of the processing of S21 to S23 is omitted. In addition, after the process of S23, it transfers to the process of main process part S30.

  Next, the L1ch output processing unit S60, the R1ch output processing unit S70, the L2ch output processing unit S70, and the R2ch output processing unit S90 will be described.

  In the L1ch output processing unit S60, first, inverse fast Fourier transform (inverse FFT) is performed (S61). In this process, specifically, a complex expression is obtained by using the OUT_L1 [f] signal calculated by the main processing unit S30 and the INL_Ar [f] calculated by S13 of the Lch analysis processing unit S10. The inverse fast Fourier transform is performed on the complex expression. Thereafter, a window function process for applying the same window as the Hanning window used in the Lch analysis processing unit S10 and the Rch analysis processing unit S20 is executed on the value calculated by the inverse fast Fourier transform (S62). For example, if the window function used in the Lch analysis processing unit S10 and the Rch analysis processing unit S20 is a Hanning window, the Hanning window is applied to the value calculated by the inverse fast Fourier transform in the processing of S62. Thereby, an OUT1_L [t] signal is generated. The reason why the Hanning window is applied to the value calculated by the inverse fast Fourier transform in the processing of S62 is to synthesize the signals output from the output processing units S60 to S90 while cross-fading.

  In the R1ch output processing unit S70, the processes of S71 to S72 are performed. Note that the processing of S71 to S72 is the same as the processing of S61 to S62. The difference from the processing of S61 to S62 is that the value for obtaining the complex expression used in the inverse fast Fourier transform is the main processing. The OUT_R1 [f] signal calculated in the unit S30 and the INR_Ar [f] calculated in the process of S23 of the Rch analysis processing unit S20. Other than that, it is the same as the process of S61-S62. Therefore, detailed description of the processing of S71 to S72 is omitted.

  In the L2ch output processing unit S80, the processes of S81 to S82 are performed. Note that the processing of S81 to S82 is the same as the processing of S61 to S62. The difference from the processing of S61 to S62 is that the value for obtaining the complex expression used in the inverse fast Fourier transform is the main processing. This is the OUT_L2 [f] signal calculated in the unit S30 (the same applies to the INL_Ar [f] calculated in the process of S13 of the Lch analysis processing unit S10). Other than that, it is the same as the process of S61-S62. Therefore, detailed description of the processing of S81 to S82 is omitted.

  In the R2ch output processing unit S90, the processes of S91 to S92 are performed. Note that the processing of S91 to S92 is the same as the processing of S61 to S62. The difference from the processing of S61 to S62 is that the value for obtaining the complex expression used in the inverse fast Fourier transform is the main processing. The OUT_R2 [f] signal calculated in the unit S30 and the INR_Ar [f] calculated in the process of S23 of the Rch analysis processing unit S20. Other than that, it is the same as the process of S61-S62. Therefore, details of the processing of S91 to S92 are omitted.

  Next, with reference to FIG. 4, the detail of the process performed by main process part S30 is demonstrated. FIG. 4 is a diagram illustrating processing executed in the main processing unit S30.

  In the main processing unit S30, first, the localization w [f] is obtained for each frequency obtained by the Fourier transform (S12, S22) performed on the IN_L [t] signal and the IN_R [t] signal, and the INL_Lv [ The larger level of f] and INR_Lv [f] is set as the maximum level ML [f] at each frequency (S31). The localization w [f] obtained in S31 and the set maximum level ML [f] are stored in a predetermined area of the RAM 16. In S31, the localization w [f] is obtained by (1 / π) × (arctan (INR_Lv [f] / INL_Lv [f]) + 0.25. Therefore, a musical sound is received at an arbitrary reference point. In this case, that is, when IN_L [t] and IN_R [t] are input at an arbitrary reference point, if the INR_Lv [f] is sufficiently larger than the INL_Lv [f], the localization w [f] is 0.75. On the other hand, if INL_Lv [f] is sufficiently larger than INR_Lv [f], the localization w [f] is 0.25.

  Next, the memory is cleared (S32). Specifically, the 1L [f] memory, 1R [f] memory, 2L [f] memory, and 2R [f] memory provided in the RAM 16 are cleared to zero. The 1L [f] memory and the 1R [f] memory are memories used when changing the localization formed by the OUT_L1 [f] signal and the OUT_R1 [f] signal output from the main processing unit S30. The 2L [f] memory and the 2R [f] memory are memories used when changing the localization formed by the OUT_L2 [f] signal and the OUT_R2 [f] signal output from the main processing unit S30.

  After the execution of S32, a first extraction process (S100), a second extraction process (S200), and a process for extracting items other than the designation (S300) are executed. The first extraction process (S100) is a process of extracting a signal to be subjected to signal processing under a first condition set in advance, that is, an extraction signal, and the second extraction process (S200) is preset. This is a process of extracting the extraction signal under the second condition. Moreover, the process (S300) which takes out other than designation | designated is a process which extracts an extraction signal on conditions other than the 1st condition and the 2nd condition. In addition, since the process (S300) for taking out items other than the designation uses the processing results of the first take-out process (S100) and the second take-out process (S200), the first take-out process (S100) and the second take-out process (S200). It is executed after the process is completed.

  After the execution of the first extraction process (S100), a first signal process for performing signal processing on the extracted signal extracted in the process (S100) is executed (S110). Further, after the execution of the second extraction process (S200), a second signal process is performed (S210) for performing signal processing on the extracted signal extracted in the process (S200). Further, after execution of the processing for extracting non-designation (S300), non-designated signal processing for performing signal processing on the extracted signal extracted in the processing (S300) is executed (S310).

  Here, the first extraction process (S100), the first signal process (S110), the second extraction process (S200), and the second signal process (S210) will be described with reference to FIG. Referring to FIG. 4, the process for extracting non-designation (S300) and the non-designation signal process (S310) will be described.

  First, the first extraction process (S100), the first signal process (S110), the second extraction process (S200), and the second signal process (S210) will be described. FIG. 5 is a diagram showing details of processing performed in the first extraction processing (S100), the first signal processing (S110), the second extraction processing (S200), and the second signal processing (S210).

  In the first extraction process (S100), the frequency f is within a preset first frequency range, and the localization w [f] and the maximum level ML [f] in the frequency within the first frequency range are respectively It is determined whether it is within a preset first setting range, that is, whether the musical sound signal satisfies the first condition (S101). The frequency f is within a preset first frequency range, and the localization w [f] and the maximum level ML [f] at a frequency within the first frequency range are each within a preset first set range. (S101: Yes), it is determined that the tone signal (left channel signal and right channel signal) at the frequency f is an extracted signal, and 1.0 is assigned to the array rel [f] [1]. (S102). In the drawing, the “l (el)” portion in the array rel is displayed as a curly elbow. The frequency at the time point determined as Yes in S101 is substituted into f described in the array rel [f] [1]. Moreover, 1 described in the array rel [f] [1] indicates that the array rel [f] [1] is an extraction signal in the first extraction process (S100).

  The frequency f is not within the preset first frequency range, or the localization w [f] at the frequency f within the first frequency range is not within the first set range, or is within the first frequency range. If the maximum level ML [f] at the frequency f is not within the first setting range (S101: No), it is determined that the tone signals (left channel signal and right channel signal) at the frequency f are not extracted signals. 0.0 is substituted into the array rel [f] [1] (S103).

  After the process of S102 or S103, it is determined whether or not the process of S101 is completed for all the Fourier-transformed frequencies (S104). If the determination of S104 is negative (S104: No), the process of S101 On the other hand, if the determination in S104 is affirmative (S104: Yes), the process proceeds to the first signal process (S110).

  In the first signal processing (S110), the level of the 1L [f] signal that is a part of the OUT_L1 [f] signal is adjusted, and the level of the 1R [f] signal that is a part of the OUT_R1 [f] signal is adjusted. Thus, the process of S111 for adjusting the localization (the amount output from the main speaker) formed by the extraction signal in the first extraction process (S100) is performed. In parallel with the processing of S111, in the first signal processing (S110), the level of the 2L [f] signal that is a part of the OUT_L2 [f] signal is adjusted and a part of the OUT_R2 [f] signal is adjusted. By adjusting the level of the 2R [f] signal, the process of S114 for adjusting the localization (the amount output from the sub-speaker) formed by the extracted signal in the first extraction process (S100) is performed.

  In the processing of S111, a 1L [f] signal that is a part of the OUT_L1 [f] signal is calculated. Specifically, “INL_Lv [f] × ll + INR_Lv [f] × lr) × rel [f] [1] × a” is a Fourier transform performed on the IN_L [t] signal and the IN_R [t] signal ( This is performed for all the frequencies obtained in S12, S22), and a 1L [f] signal is calculated. Similarly, in the process of S111, a 1R [f] signal that is a part of the OUT_R1 [f] signal is calculated. Specifically, “INL_Lv [f] × rl + INR_Lv [f] × rr) × rel [f] [1] × a” is performed for all the frequencies subjected to Fourier transform in S12 and S22, and 1R [f ] Signal is calculated.

  Here, a is a coefficient predetermined for the first signal processing, and ll, lr, rl, rr are localization w [f] obtained from the musical sound signals (left channel signal, right channel signal); It is a coefficient determined in accordance with a target localization (for example, a value within a range of 0.25 to 0.75) predetermined for the first signal processing. In the drawing, “l (el)” is written as cursive el. Here, ll, lr, rl, and rr will be described with reference to FIG. FIG. 7 is a graph for explaining each coefficient determined according to the localization w [f] and the target localization. In the graph of FIG. 7, the horizontal axis is a value of (target localization−localization w [f] +0.5), and the vertical axis is each coefficient (ll, lr, rl, rr, ll ′, lr ′). , Rl ′, rr ′).

  The relationship between ll and rr is as shown in FIG. Therefore, when the value of (target localization−localization w [f] +0.5) is 0.5, ll and rr are the maximum coefficients. On the other hand, the relationship between lr and rl is as shown in FIG. 7B, and when the value of (target localization-localization w [f] +0.5) is 0.5, rl is the smallest (zero) coefficient between each other.

  Returning to the description of FIG. After the processing of S111, processing for performing pitch change, level change, and reverberation is performed on the 1L [f] signal (S112). Note that pitch change, level change, and reverberation (so-called convolution reverb) are all well-known techniques and will not be described in detail. When the process of S112 is performed on the 1L [f] signal, the 1L_1 [f] signal constituting the OUT_L1 [f] signal is generated. Similarly, after the processing of S111, processing for performing pitch change, level change, and reverberation is performed on the 1R [f] signal (S113). When the process of S113 is performed on the 1R [f] signal, the 1R_1 [f] signal constituting the OUT_R1 [f] signal is generated.

  In the process of S114, a 2L [f] signal that is a part of the OUT_L2 [f] signal is calculated. Specifically, “INL_Lv [f] × ll ′ + INR_Lv [f] × lr ′) × rel [f] [1] × b” was performed on the IN_L [t] and IN_R [t] signals. This is performed for all the frequencies obtained by the Fourier transform (S12, S22), and a 2L [f] signal is calculated. Similarly, in the process of S114, a 2R [f] signal that is a part of the OUT_R2 [f] signal is calculated. Specifically, “INL_Lv [f] × rl ′ + INR_Lv [f] × rr ′) × rel [f] [1] × b” is performed for all the frequencies subjected to Fourier transform in S12 and S22. A 2R [f] signal is calculated.

  Note that b is a coefficient predetermined for the first signal processing. This coefficient b may be the same as or different from the coefficient a. In addition, ll ′, lr ′, rl ′, rr ′ are the localization w [f] obtained from the musical tone signal and the target localization predetermined for the second signal processing (for example, 0.25-0) A value in the range of .75). Here, ll ′, lr ′, rl ′, rr ′ will be described with reference to FIG. The relationship between ll ′ and rr ′ is as shown in FIG. 7A. When the value of (target localization−localization w [f] +0.5) is 0.0, ll ′ is While it is the largest coefficient, rr ′ is the smallest (zero) coefficient. Conversely, when the value of (target localization-localization w [f] +0.5) is 1.0, ll ′ is the minimum (zero) coefficient, while rr ′ is the maximum coefficient. It becomes.

  Similarly, the relationship between lr ′ and rl ′ is as shown in FIG. 7B, and when the value of (target localization−localization w [f] +0.5) is 0.0, lr ′ is the largest coefficient, while rl ′ is the smallest (zero) coefficient. Conversely, when the value of (target localization-localization w [f] +0.5) is 1.0, lr ′ is the minimum (zero) coefficient, while rl ′ is the maximum coefficient. It becomes.

  Returning to the description of FIG. After the process of S114, processing for changing the pitch, changing the level, and applying reverb is performed on the 2L [f] signal (S115). When the process of S115 is performed on the 2L [f] signal, the 2L_1 [f] signal constituting the OUT_L2 [f] signal is generated. Similarly, after the process of S114, a process for changing the pitch, changing the level, and applying a reverb is performed on the 2R [f] signal (S116). When the process of S116 is performed on the 2R [f] signal, the 2R_1 [f] signal constituting the OUT_R2 [f] signal is generated.

  In the second extraction process S200 executed in parallel with the first extraction process S100, the frequency f is within the preset second frequency range, and the localization w [f] at a frequency within the second frequency range. It is then determined whether or not the maximum level ML [f] is within a preset second setting range, that is, whether or not the musical tone signal satisfies the second condition (S201). In the present embodiment, the second frequency range is a range different from the first frequency range (a range where the range start and the range end do not completely match), and the second set range is different from the first set range. A range (a range where the start and end of the range do not completely match) is assumed. The second frequency range may be a range that partially overlaps the first frequency range, or may be a range that completely coincides with the first frequency range. The second setting range may also be a range that partially overlaps the first setting range, or may be a range that completely matches the first setting range.

  The frequency f is within a preset second frequency range, and the localization w [f] and the maximum level ML [f] at a frequency within the second frequency range are each within a preset second set range. (S201: Yes), the tone signal (left channel signal and right channel signal) at the frequency f is determined to be an extracted signal, and 1.0 is assigned to the array rel [f] [2]. (S102). Note that 2 described in the array rel [f] [2] indicates that the array rel [f] [2] is an extraction signal in the second extraction process S200.

  The frequency f is not within the preset second frequency range, or the localization w [f] at the frequency f within the second frequency range is not within the second set range, or is within the second frequency range. If the maximum level ML [f] at the frequency f is not within the second setting range (S201: No), it is determined that the tone signals (left channel signal and right channel signal) at the frequency f are not extracted signals. 0.0 is substituted into the array rel [f] [2] (S203).

  After the process of S202 or S203, it is determined whether or not the process of S201 has been completed for all the Fourier-transformed frequencies (S204). If the determination of S204 is negative (S204: No), the process of S201 On the other hand, if the determination in S204 is affirmative (S204: Yes), the process proceeds to the second signal process (S210).

  In the second signal processing (S210), the level of the 1L [f] signal that is a part of the OUT_L1 [f] signal is adjusted, and the level of the 1R [f] signal that is a part of the OUT_R1 [f] signal is adjusted. Thus, the process of S211 for adjusting the localization (the amount output from the main speaker) formed by the extracted signal in the second extraction process (S200) is performed. In parallel with the processing of S211, in the second signal processing (S210), the level of the 2L [f] signal that is a part of the OUT_L2 [f] signal is adjusted, and a part of the OUT_R2 [f] signal is adjusted. By adjusting the level of the 2R [f] signal, the process of S214 for adjusting the localization (the amount output from the sub-speaker) formed by the extracted signal in the second extraction process (S200) is performed.

  Here, each process of S211 to S216 in the second signal process (S210) is the same as each process of S111 to S116 in the first signal process (S110) except that the points described below are different. Therefore, the description thereof is omitted. The first difference between the second signal processing (S210) and the first signal processing (S110) is that the signal input to the first signal processing is an extraction signal from the first extraction processing (S100). On the other hand, the signal input to the second signal processing is an extraction signal from the second extraction processing (S200), and the second difference is that in the first signal processing, the array rel [f] [1 ] Is used in the second signal processing, the array rel [f] [2] is used, and the third difference is that the signal output from the first signal processing is 1L_1 [f]. 1R_1 [f], 1L_2 [f], and 1R_2 [f], the signals output from the second signal processing are 2L_1 [f], 2R_1 [f], 2L_2 [f], and 2R_2. The point is [f].

  Note that the target localization in the first signal processing (S110) and the target localization in the second signal processing (S210) may be the same or different. That is, when the target localization is different between the first signal processing and the second signal processing, the coefficients ll, rr, rl, rr, ll ′, rr ′, rl ′, rr ′ used in the first signal processing The coefficients ll, lr, rl, rr, ll ′, lr ′, rl ′, rr ′ used in the second signal processing are different. Further, the coefficients a and b used in the first signal processing and the coefficients a and b used in the second signal processing may be the same or different. The contents of the processing processes S112, S113, S115, and S116 executed in the first signal processing and the contents of the processing processes S212, S213, S215, and S216 executed in the second signal processing (S210) are as follows. , May be the same or different.

  Next, a process for extracting non-designation (S300) and a non-designation signal process (S310) will be described. FIG. 6 is a diagram showing the details of the processing performed in the processing for extracting non-designation (S300) and non-designated signal processing (S310).

  In the process of extracting other than specified (S300), first, rel [f] [1] at the lowest frequency among the frequencies subjected to Fourier transform in S12 and S22 is 0.0, and rel [f] at the lowest frequency is selected. f] [2] is 0.0, that is, the tone signal (left channel signal and right channel signal) at the lowest frequency is used as an extraction signal, the first extraction process (S100) or the second extraction process (S200). (S301). In the present embodiment, the value of rel [f] [1] set in S102 and S103 in the first extraction process (S100) and the rel set in S202 and S203 in the second extraction process (S200). Assume that the determination in S301 is performed using the values of [f] and [2]. Further, before performing the process of S301, a process similar to the first and second extraction processes (S100, S200) is separately executed, and the value of rel [f] [1] and rel [f obtained at that time are obtained. ] The determination in S301 may be performed using the value of [2].

  When rel [f] [1] at the lowest frequency is 0.0 and rel [f] [2] at the lowest frequency is 0.0 (S301: Yes), at the lowest frequency It is determined that the tone signal has not yet been extracted as an extraction signal in the first extraction process (S100) or the second extraction process (S200), and 1.0 is substituted into the array remain [f] (S302). Here, 1.0 assigned to “remain [f]” indicates that the tone signal at the lowest frequency is an extracted signal in the process (S300) for extracting a non-designated signal. It should be noted that the frequency at the time when Yes is determined in S301 is substituted for f described in “remain [f]”.

  On the other hand, when rel [f] [1] at the lowest frequency is 1.0, or when rel [f] [2] at the lowest frequency is 1.0, or both are 1.0. In some cases (S301: No), it is determined that the musical tone signal at the lowest frequency has already been extracted as the extraction signal in the first extraction process S100 or the second extraction process S200, and the array remain [f] contains 0. .0 is substituted (S302). Here, 0.0 assigned to “remain [f]” indicates that the tone signal at the lowest frequency is not an extraction signal in the process (S300) for extracting other than specified.

  After the process of S302 or S303, it is determined whether or not the process of S301 has been completed for all the frequencies subjected to Fourier transform in S12 and S22 (S304). If the determination of S304 is negative (S304: No) ), The process returns to the process of S301, and the determination of S301 is performed for the lowest frequency among the frequencies for which the determination of S301 has not been performed. On the other hand, if the determination of S304 is affirmative (S304: Yes) ), The process proceeds to non-designated signal processing (S310).

  In the non-designated signal processing (S310), the level of the 1L [f] signal that is a part of the OUT_L1 [f] signal is adjusted, and the level of the 1R [f] signal that is a part of the OUT_R1 [f] signal is adjusted. Thus, the process of S311 for adjusting the localization (the amount output from the main speaker) formed by the extracted signal in the process (S300) of extracting other than the designation is performed. In parallel with the processing of S311, in the non-designated signal processing (S310), the level of the 2L [f] signal that is a part of the OUT_L2 [f] signal is adjusted, and a part of the OUT_R2 [f] signal is adjusted. By adjusting the level of the 2R [f] signal, the process of S314 is performed to adjust the localization (the amount output from the sub-speaker) formed by the extracted signal in the process (S300) of extracting other than the designation. .

  In the process of S311, a 1L [f] signal that is a part of the OUT_L1 [f] signal is calculated. Specifically, “(INL_Lv [f] × ll + INR_Lv [f] × lr) × remain [f] × c” is performed for all frequencies subjected to Fourier transform in S12 and S22, and the 1L [f] signal is obtained. Is calculated. Similarly, in the process of S311, a 1R [f] signal that is a part of the OUT_R1 [f] signal is calculated. Specifically, “(INL_Lv [f] × rl + INR_Lv [f] × rr) × remain [f] × c” is performed for all the frequencies subjected to Fourier transform in S12 and S22, and the 1R [f] signal Is calculated. Note that c is a predetermined coefficient for calculating 1L [f] and 1R [f] in the non-designated signal processing (S310), and this coefficient c is the same as the above-described coefficients a and b. It may or may not be.

  After the processing in S311, processing for performing pitch change, level change, and reverberation is performed on the 1L [f] signal (S312). When the process of S312 is performed on the 1L [f] signal, the 1L_3 [f] signal constituting the OUT_L1 [f] signal is generated. Similarly, after the processing in S311, processing for performing pitch change, level change, and reverberation is performed on the 1R [f] signal (S313). When the process of S113 is performed on the 1R [f] signal, the 1R_3 [f] signal constituting the OUT_R1 [f] signal is generated.

  In the process of S314, a 2L [f] signal that is a part of the OUT_L2 [f] signal is calculated. Specifically, “(INL_Lv [f] × ll ′ + INR_Lv [f] × lr ′) × remain [f] × d” is performed on all the frequencies subjected to Fourier transform in S12 and S22, and 2L [ f] The signal is calculated. Similarly, in the process of S314, a 2R [f] signal that is part of the OUT_R2 [f] signal is calculated. Specifically, “(INL_Lv [f] × rl ′ + INR_Lv [f] × rr ′) × remain [f] × d” is performed for all the frequencies subjected to Fourier transform in S12 and S22, and 2R [ f] The signal is calculated. Note that d is a predetermined coefficient for calculating 2L [f] and 2R [f] in the non-designated signal processing (S310), and this coefficient d is the coefficient a, b, c described above. They may be the same or different.

  After the processing of S314, processing for changing the pitch, changing the level, and applying reverb is performed on the 2L [f] signal (S315). When the process of S315 is performed on the 2L [f] signal, the 2L_3 [f] signal constituting the OUT_L2 [f] signal is generated. Similarly, after the processing of S314, processing for changing the pitch, changing the level, and applying reverb is performed on the 2R [f] signal (S316). When the process of S316 is performed on the 2R [f] signal, the 2R_3 [f] signal constituting the OUT_R2 [f] signal is generated.

  As described above, in the main processing unit S30, as shown in FIGS. 5 and 6, in addition to the processes of S111, S211, and S311, the processes of S114, S214, and S314 are executed. Accordingly, the left channel signal and the right channel signal in the extracted extracted signal can be distributed, and independent signal processing can be performed on the distributed left channel signal and right channel signal. Therefore, different signal processing (processing for changing the localization) can be performed on each of the left and right channel signals distributed from the extracted signal. Of course, the same signal processing can be performed on each of the left and right channel signals distributed from the extracted signal. Here, the signals generated by the processes of S111, S211, and S311 are output from the OUT1_L terminal and the OUT1_R terminal, which are terminals for the main speaker, after the processing process, and are generated by the processes of S114, S214, and S314. The processed signal is output from the OUT2_L terminal and the OUT2_R terminal, which are sub-speaker terminals, after the processing. Therefore, after extracting an extraction signal for each desired condition, different signal processing and processing are performed on the extraction signal under the desired condition (in other words, the extraction signal under the one condition). Each extraction signal subjected to signal processing and processing can be output separately from each of the OUT1 terminal and the OUT2 terminal.

  Returning to the description of FIG. When the first signal processing (S110), the second signal processing (S210), and the non-designated signal processing (S310) are completed, the 1L_1 [f] signal generated in the first signal processing (S110) and the second signal processing ( The 1L_2 [f] signal generated in S210) and the 1L_3 [f] signal generated in the non-designated signal processing (S310) are combined to generate the OUT_L1 [f] signal. When the OUT_L1 [f] signal is input to the L1ch output processing unit S60 (see FIG. 3), the L1ch output processing unit S60 converts the input OUT_L1 [f] signal to OUT1_L [t], and The signal is output to the first D / A converter 13L1 for Lch (see FIG. 1) via the line 17.

  Similarly, the 1R_1 [f] signal generated in the first signal processing (S110), the 1R_2 [f] signal generated in the second signal processing (S210), and 1R_3 [generated in the non-designated signal processing (S310). The signal f] is combined with the signal OUT_R1 [f]. When the OUT_R1 [f] signal is input to the R1ch output processing unit S70 (see FIG. 3), the R1ch output processing unit S70 converts the input OUT_R1 [f] signal into an OUT1_R [t] signal, The data is output to the first D / A converter 13R1 for Rch (see FIG. 1) via the bus line 17. Note that generation of the OUT_L2 [f] signal and OUT_R2 [f] signal and conversion of the OUT2_L [t] signal and OUT2_R [t] signal are performed in the same manner as described above.

  Thus, in the generation of the OUT_L1 [f] signal, the 1L_1 [f] signal generated in the first signal processing (S110) and the 1L_2 [f] signal generated in the second signal processing (S210) are combined. In addition, the 1L_3 [f] signal generated in the non-designated signal processing (S310) is synthesized. Similarly, in the generation of the OUT_R1 [f] signal, the 1R_1 [f] signal generated in the first signal processing (S110) and the 1R_2 [f] signal generated in the second signal processing (S210) are combined. In addition, the 1R_3 [f] signal generated in the non-designated signal processing (S310) is synthesized. Therefore, a signal that has not been extracted in the first signal processing (S110) or the second signal processing (S210) can be synthesized with the extracted signal extracted for each desired condition. Therefore, the OUT_L1 [f] signal and the OUT_R1 [f] can be made to be the same signal as the input musical sound signal, that is, a natural sound having a spread.

  As described above, in the effector 1 of the first embodiment, signal processing (S110, S210) is performed on the extracted signal extracted in the first extraction process (S100) or the second extraction process (S200). Here, the first extraction process (S100) and the second extraction process (S200) satisfy each condition for each set condition (for each condition in which the frequency, localization, and maximum level are a set). A musical sound signal (left channel signal and right channel signal) is extracted as an extraction signal. Therefore, extraction of an extraction signal to be subjected to signal processing can be performed over a plurality of conditions (each of which is a set of frequency, localization and maximum level).

  Next, with reference to FIG.8 and FIG.9, the effector 1 of 2nd Embodiment is demonstrated. In the second embodiment, the same parts as those in the first embodiment described above are denoted by the same reference numerals, and the description thereof is omitted.

  The effector 1 of the second embodiment, like the effector 1 of the first embodiment, is a musical tone signal (that is, an extracted signal) extracted based on the setting condition by the first or second extraction process (S100, S200). On the other hand, the first or second signal processing (S110, S210) can be performed independently for each setting condition of the extraction processing. In the first or second signal processing, the sound image enlargement / reduction processing is performed. , And the sound image can be enlarged (ie, enlarged at an enlargement ratio larger than 1) or reduced (ie, enlarged at an enlargement ratio larger than 0 times and smaller than 1).

  First, the outline of the sound image enlargement / reduction process performed by the effector 1 of the second embodiment will be described with reference to FIG. FIG. 8 is a schematic diagram showing how a sound image is enlarged or reduced by the sound image enlargement / reduction processing in the second embodiment.

  The condition for extracting the extracted signal by the first or second extraction process (S100, S200) (that is, the condition in which the frequency, localization and maximum level are set as a set) is formed with the frequency and localization as two axes. In the coordinate plane, two sides adjacent to a conditional frequency range (first frequency range, second frequency range) and a conditional localization range (first setting range, second setting range) are (Hereinafter, this rectangular area is referred to as “extraction area”). The extracted signal exists within the rectangular area. In FIG. 8, the frequency range is Low ≦ frequency f ≦ High, the localization range is panL ≦ localization w [f] ≦ panR, the extraction region is the frequency f = Low, and the localization w [f ] = PanL, frequency f = Low, localization w [f] = panR, frequency f = High, localization w [f] = panR, and frequency f = High And is represented as a rectangular region having four vertices with a point where the localization w [f] = panL.

  In the sound image enlargement / reduction process, the localization w [f] of the extracted signal in the extraction area is moved by linear mapping into a target area (hereinafter referred to as “target area”) for enlargement or reduction of the sound image. It is processing. The target area includes a sound image expansion function YL (f) that defines the boundary localization on one end side according to the frequency, a sound image expansion function YR (f) that defines the boundary localization on the other end side according to the frequency, and a frequency range. This is a region surrounded by (Low ≦ frequency f ≦ High).

  In the sound image enlargement / reduction processing of this embodiment, the center of the localization range (panL ≦ localization w [f] ≦ panR in FIG. 8), that is, panC in FIG. Among them, the localization of the extracted signal that is localized on the panL side from panC is moved by continuous linear mapping with reference to panC that is the reference localization using the sound image expansion function YL (f), while panR from panC. The localization of the extracted signal localized on the side is moved by a continuous linear mapping with reference to panC, which is the reference localization, using the sound image expansion function YR (f).

  Note that the case where the extraction signal localized from panC to the panL side moves to the panL side or the extraction signal localized from panC to the panR side moves to the panR side is enlarged, and the extraction signal moves to the reference localization panC side. If you want to reduce it. That is, in the frequency region where the sound image expansion function YL (f) is located outside the extraction region, the sound image formed by the extraction signal localized on the panL side from panC is expanded, and the sound image expansion function YL (f) is stored in the extraction region. In the frequency region located on the inner side, the sound image formed by the extraction signal localized on the panL side from panC is reduced. Similarly, in the frequency region where the sound image expansion function YR (f) is located outside the extraction region, the sound image formed by the extraction signal localized on the panR side from panC is expanded, and the sound image expansion function YR (f) is extracted. In the frequency region located inside the sound image, the sound image formed by the extracted signal localized on the panR side from panC is reduced.

  In the example shown in FIG. 8, the sound image expansion function YL (f) and the sound image expansion function YR (f) are both functions that draw a straight line according to the value of the frequency f, but the sound image expansion function YL (f) The sound image expansion function YR (f) is not limited to a function that draws a straight line in accordance with the value of the frequency f, and functions having various shapes can be applied. For example, a function that draws a polygonal line according to the range of the frequency f, a function that draws a parabola (that is, a quadratic curve) according to the value of the frequency f, a cubic function according to the value of the frequency f, and an arc of an ellipse or circle Representing functions, exponential functions or logarithmic functions can be applied.

  The sound image expansion functions YL (f) and YR (f) may be predetermined or set by the user. For example, the sound image enlargement functions YL (f) and YR (f) to be used are set in advance according to the frequency region and the localization range, and the sound image enlargement function YL (in accordance with the position of the extraction region (frequency region and localization range)) f) and YR (f) may be selected.

  Further, if desired, the user sets two or more coordinates (that is, a set of frequency and localization) on the coordinate plane including the extraction area, and based on the set frequency and localization set, the sound image expansion function You may comprise so that YL (f) or YR (f) may be set. For example, the user sets the point where the localization at the frequency f = Low is YL (Low) and the point where the localization at the frequency f = High is YL (High), and thereby the localization with respect to the change of the frequency f. Is set as a sound image expansion function YL (f) that is a function of linearly changing, while the localization at the frequency f = Low is YR (Low) and the localization at the frequency f = High is YR (High). The sound image expansion function YR (f), which is a function in which the localization changes linearly with respect to the change in the frequency f, may be set. Or you may comprise so that a user may each set the change pattern (a straight line, a parabola, an arc, etc.) of sound image expansion function YL (f) and sound image expansion function YR (f). Note that the frequency range of the sound image expansion function YL (f) or YR (f) may exceed the frequency range of the extraction region, as shown in FIG.

  When the sound image enlargement function YL (f) and the sound image enlargement function YR (f) are functions that draw a straight line according to the value of the frequency f, these sound image enlargement functions YL (f) and YR (f) are expressed as follows. It can be obtained as follows.

  First, BtmL and BtmR are assumed as coefficients for determining the degree of expansion on the Low side of the frequency f, and TopL and TopR are assumed as coefficients for determining the degree of expansion on the High side. Note that BtmL and TopL determine the degree of enlargement in the left direction (panL direction) from panC, which is the reference localization, and BtmR and TopR determine the degree of enlargement in the right direction (panR direction) from panC. These four coefficients BtmL, BtmR, TopL, and TopR are each in the range of about 0.5 to 10.0, for example. Note that when this coefficient exceeds 1.0, enlargement occurs, and when it is greater than 0 and less than 1.0, reduction occurs.

  Here, when the coordinates of both ends in the sound image expansion function YL (f) that is a straight line, that is, the coordinates of YL (Low) and YL (High) are obtained, YL (Low) = panC + (panL−panC) × BtmL. , YL (High) = panC + (panL−panC) × TopL. At this time, if Wl = panL−panC, YL (f) = {Wl × (TopL−BtmL) / (High−Low)} × (f−Low) + panC + Wl × BtmL.

  Similarly, for the sound image enlargement function YR (f), YR (Low) = panC + (panR−panC) × BtmR and YR (High) = panC + (panR−panC) × TopR, so Wr = panR−panC Then, YR (f) = {Wr × (TopR−BtmR) / (High−Low)} × (f−Low) + panC + Wr × BtmR.

  When the extracted signal PoL [f], which is localized in the left direction from the reference localization PanC, is moved using the sound image expansion function YL (f), the localization PtL [f] of the movement destination is based on panC. The ratio of the length from panC to PoL [f] and the length from panC to PtL [f] at a certain frequency f is the length from panC to panL and the length from panC to YL (f). It can be calculated based on being equal to the ratio. That is, the localization PtL [f] of the movement destination is (PtL [f] -panC) :( PoL [f] -panC) = (YL (f) -panC) :( panL-panC). [F] = (PoL [f] −panC) × (YL (f) −panC) / (panL−panC) + panC.

  Similarly, when moving the extraction signal PoR [f], which is localized to the right of the reference localization PanC, using the sound image expansion function YR (f), the localization PtR [f] of the movement destination is (PtR [ f] -panC): (PoR [f] -panC) = (YR (f) -panC): (panR-panC), so PtR [f] = (PoR [f] -panC) × (YR (F) −panC) / (panR−panC) + panC.

  In the sound image enlargement / reduction processing, the localizations PtL [f] and PtR [f] that are the movement destinations are set as the target localizations, and the coefficients ll, lr, rl, rr for moving the localization, or the coefficient ll ′ , Lr ′, rl ′, rr ′, thereby moving the localization of the extracted signal. As a result, the sound image in the extraction area is enlarged or reduced.

  That is, in the present embodiment, the localization of the extracted signal that is localized on the panL side of panC among the extracted signals in the extraction region is continuously performed with reference to panC that is the reference localization using the sound image expansion function YL (f). On the other hand, the localization of the extracted signal localized on the panR side from panC is moved by the continuous linear mapping with reference to panC as the reference localization using the sound image expansion function YR (f). By doing so, the sound image in the extraction area is enlarged or reduced.

  In FIG. 8, as an example, a state in which the sound image expansion functions YL (f) and YR (f) are set for one extraction region is illustrated, but the sound image expansion functions YL (f) and YR ( f) may be set for each take-out area.

  For example, sound image enlargement functions YL (f) and YR (f) that are different in the extraction region in which the high sound region is the frequency range, the extraction region in which the middle sound region is the frequency range, and the extraction region in which the low sound region is the frequency range, respectively. May be set. When enlarging the sound image of a stereo signal as a whole, the sound image enlarging functions YL (f) and YR (f ) And the sound image expansion functions YL (f) and YR (f) are set so that the expansion becomes larger as the frequency is increased for the entire localization range in the middle range, and a favorable audibility is given. Can do. On the other hand, the low sound range signal extraction may not be performed, and the sound image may not be expanded (or reduced).

  Note that when there are a plurality of extraction areas, an area for enlarging or reducing the sound image, that is, an area for setting the reference localization, the sound image enlarging function YL (f), and the sound image enlarging function YR (f), You may be limited to some extraction areas instead of all the extraction areas.

  In addition, by setting common BtmL, BtmR, TopL, and TopR for all extraction regions, the sound image enlargement functions YL (f) and YR that make all the extraction regions have the same expansion (or reduction) degree. (F) may be set.

  Further, BtmL, BtmR, TopL, and TopR may be set as a function of the position of the region to be extracted and / or the size of the region. That is, according to the extraction region, the enlargement degree (or reduction degree) may be changed based on a predetermined rule. For example, as the frequency increases, the degree of expansion increases, or the degree of expansion decreases as the localization of the extracted signal moves away from the reference position (for example, panC at the center), and so on. BtmL, BtmR, TopL , TopR may be set.

  Further, the reference localization, the sound image enlargement function YL (f), and the sound image enlargement function YR (f) may be set in common for all the extraction areas. That is, the extraction signals in all the extraction regions may be linearly mapped by the same sound image expansion functions YL (f) and YR (f) with the same reference localization as a reference. In such a case, by selecting the entire musical sound as one extraction area, the sound image of the entire musical sound is selected under the condition of one condition (that is, the reference localization set to sharing and the sound image expansion functions YL (f), YR). You may make it enlarge or reduce by (f)).

  In this embodiment, the center of the localization range in the extraction region (the range of panL ≦ the localization w [f] ≦ panR in FIG. 8), that is, panC is set as the reference localization, but the reference localization is in the extraction region or in the extraction region. It can be set to any localization outside the area. In the case where there are a plurality of extraction areas, a different reference localization may be set for each extraction area, or a common reference localization may be set for all extraction areas. The reference localization may be set in advance for each extraction area or for all extraction areas, or may be set by the user each time.

  Next, with reference to FIG. 9, the sound image enlargement / reduction processing performed by the effector 1 of the present embodiment will be described. FIG. 9 is a diagram illustrating details of processing performed in the first signal processing S110 and the second signal processing S210 of the second embodiment.

  In the first signal processing (S110) of the second embodiment, as shown in FIG. 9, after a musical sound signal satisfying the first condition is extracted as an extraction signal by the first extraction processing (S100), it is formed from the extraction signal. In order to enlarge or reduce the sound image, a process of calculating the localization movement amount of the extracted signal for the amount output from the main speaker is executed (S117). Similarly, in order to enlarge or reduce the sound image formed from the extracted signal extracted by the first extraction process (S100), a process of calculating the localization movement amount of the extracted signal for the amount output from the sub-speaker. Execute (S118).

  In the process of S117, the extraction signal located in the left direction from the reference localization is moved by the sound image expansion function YL1 [1] (f) in the extraction area (that is, the area determined by the first condition) by the first extraction process (S100). The amount of movement ML1 [1] [f] when moved and the extracted signal located to the right from the reference localization is moved by the sound image enlargement function YR1 [1] (f). And calculate.

  Note that the sound image magnification function YL1 [1] (f) and the sound image magnification function YR1 [1] (f) are both sound image magnification functions for moving the localization of the extracted signal corresponding to the amount output from the main speaker. Respectively, a function for moving the extracted signal in the left direction from the reference localization and a function for moving the extracted signal in the right direction from the reference localization.

  Specifically, in the process of S117, “{(w [f] −panC [1]) × (YL1 [1] (f) −panC [1]) / (panL [1] −panC [1]) + panC The movement amount ML1 [1] [f] is calculated by performing [1]} − w [f] ”on all the frequencies subjected to Fourier transform in S12 and S22. Similarly, “{(w [f] −panC [1]) × (YR1 [1] (f) −panC [1]) / (panR [1] −panC [1]) + panC [1]} − w The movement amount MR1 [1] [f] is calculated by performing [f] ”on all the frequencies subjected to Fourier transform in S12 and S22. Note that panL [1] and panR [1] are the localization of the left and right boundaries in the extraction region by the first extraction process (S100), and panC [1] is the reference localization in the extraction region by the first extraction process (S100). , For example, the center of the localization range in the extraction region.

  After the process of S117, using the movement amount ML1 [1] [f] and the movement amount MR1 [1] [f], the localization (from the main speaker) formed by the extraction signal extracted by the first extraction process (S100). The output is adjusted) (S111). Specifically, in the processing of S111, the movement amount ML1 [1] [f] and the movement amount MR1 [1] [f] are calculated based on the target localization (that is, the localization of the movement destination by enlargement or reduction). Since the difference is obtained by subtracting the localization w [f], the coefficients ll, lr, rl, rr for moving the localization using the movement amount ML1 [1] [f] and the movement amount MR1 [1] [f] are used. Using the determined coefficients ll, lr, rl, rr, the localization is adjusted in the same manner as in S111 in the first embodiment to obtain the 1L signal and the 1R signal. When the adjusted localization is less than 0, the localization is set to 0. On the other hand, when the adjusted localization exceeds 1, the localization is set to 1. The calculation of the movement amounts ML1 [1] [f] and MR1 [1] [f] by the process of S117 and the localization adjustment by the process of S111 correspond to the sound image enlargement / reduction process.

  Thereafter, the 1L [f] signal is subjected to processing in S112 to be a 1L_1 [f] signal, and the 1R [f] signal is subjected to processing in S113 to be a 1R_1 [f] signal.

  On the other hand, in the process of S118 for calculating the amount of localization movement of the extracted signal for the amount output from the sub-speaker, the extracted image that is to the left of the reference localization in the extraction area by the first extraction process (S100) is enlarged in the sound image. When the movement amount ML2 [1] [f] when moved by the function YL2 [1] (f) and the extracted signal in the right direction from the reference localization are moved by the sound image expansion function YR2 [1] (f) Movement amount MR2 [1] [f] is calculated.

  Note that the sound image magnification function YL2 [1] (f) and the sound image magnification function YR2 [1] (f) are both sound image magnification functions for moving the localization of the extracted signal that is output from the sub-speaker. Respectively, a function for moving the extracted signal in the left direction from the reference localization and a function for moving the extracted signal in the right direction from the reference localization. The sound image enlargement function YL2 [1] (f) is the same as or different from the sound image enlargement function YL1 [1] (f) used to move the localization of the extracted signal corresponding to the amount output from the main speaker. Also good. Similarly, the sound image enlargement function YR2 [1] (f) is the same as the sound image enlargement function YR1 [1] (f) used to move the localization of the extracted signal for the amount output from the main speaker. May be different.

  For example, when the main speaker and the sub-speaker are arranged at equal intervals, YL1 [1] (f) and YL2 [1] (f) are made the same, and YR1 [1] (f) and YR2 [ 1] Same as (f). When the main speaker and the sub-speaker are extremely separated, the movement amount ML2 [1] [f] and the movement amount MR2 [1] [f] are changed to the movement amount ML1 [1] [f] and the movement amount. Sound image expansion functions YL2 [1] (f) and YR2 [1] (f) that are smaller than the quantity MR1 [1] [f] are used.

  Specifically, in the process of S118, “{(w [f] −panC [1]) × (YL2 [1] (f) −panC [1]) / (panL [1] −panC [1]) + panC The movement amount ML2 [1] [f] is calculated by performing [1]} − w [f] ”on all the frequencies subjected to Fourier transform in S12 and S22. Similarly, “{(w [f] −panC [1]) × (YR2 [1] (f) −panC [1]) / (panR [1] −panC [1]) + panC [1]} − w The movement amount MR2 [1] [f] is calculated by performing [f] ”on all the frequencies subjected to the Fourier transform in S12 and S22. These movement amounts ML2 [1] [f] and MR2 [1] [f] are differences obtained by subtracting the localization w [f] of the extraction signal from the target localization (that is, the localization of the movement destination by enlargement or reduction). It corresponds to.

  After the process of S118, using the movement amount ML2 [1] [f] and the movement amount MR2 [1] [f], the localization (from the sub-speaker) formed by the extraction signal extracted by the first extraction process (S100). The output is adjusted (S114). Specifically, in the process of S114, using the movement amount ML2 [1] [f] and the movement amount MR2 [1] [f], coefficients ll ′, lr ′, rl ′, rr ′ for moving the localization are used. Using the determined coefficients ll ′, lr ′, rl ′, and rr ′, the localization is adjusted in the same manner as S114 in the first embodiment to obtain the 2L signal and the 2R signal. When the adjusted localization is less than 0, the localization is set to 0. On the other hand, when the adjusted localization exceeds 1, the localization is set to 1. Further, the calculation of the movement amounts ML2 [1] [f] and MR2 [1] [f] by the process of S118 and the localization adjustment by the process of S114 correspond to the sound image enlargement / reduction process.

  Thereafter, the 2L [f] signal is subjected to processing in S115 to be a 2L_1 [f] signal, and the 2R [f] signal is subjected to processing in S116 to be a 2R_1 [f] signal.

  Also, as shown in FIG. 9, in the second signal processing (S210) of the second embodiment, a tone signal that satisfies the second condition is extracted as an extraction signal by the second extraction processing (S200), and then formed from the extraction signal. In order to perform enlargement or reduction of the sound image to be performed, a process of calculating the localized movement amount ML1 [2] [f] for the amount output from the main speaker is executed (S217). Similarly, in order to enlarge or reduce the sound image formed from the extracted signal extracted by the second extraction process (S200), the localization movement amount MR2 [1] [f] corresponding to the amount output from the sub-speaker The process which calculates is performed (S218).

Here, the processing of S217 is performed in the same manner as the processing of S117 executed in the first signal processing (S110) except that the points described below are different, and thus the description thereof is omitted. The difference between the process of S217 and the process of S117 is that the sound image enlargement function for moving the localization output from the main speaker is a function for moving the extracted signal in the left direction from the reference localization as YL1 [ 2] (f),
Instead of using YR1 [2] (f), which is a function for moving the extracted signal in the right direction from the reference localization, instead of panL [1] and panR [1], the second extraction process (S200) is used. Localization of the left and right boundaries in the extraction area, that is, the point using panL [2] and panR [2], and the localization in the extraction area by the second extraction process (S200) instead of panC [1] as the reference localization ( For example, panC [2] that is the center of the localization range in the extraction region is used.

Further, the processing of S218 is performed in the same manner as the processing of S118 executed in the first signal processing (S110) except that the points described below are different, and thus the description thereof is omitted. The difference between the process of S218 and the process of S118 is that YL2 [which is a function for moving the extracted signal in the left direction from the reference localization as a sound image enlargement function for moving the localization output from the sub-speaker. 2] (f),
Instead of using the function YL2 [2] (f) for moving the extracted signal in the right direction from the reference localization, instead of panL [1] and panR [1], panL [2] and panR [2 ] And the point of using panC [2] instead of panC [1] as the reference localization.

  Then, after the processing of S217, the coefficients ll, lr, rl, rr are determined using the calculated movement amount ML1 [2] [f] and the movement amount MR1 [2] [f]. The localization (the amount output from the main speaker) formed by the extracted signal extracted by the extraction process (S200) is adjusted (S211). In the processing of S211, if the adjusted localization is less than 0 (zero), the localization is set to 0 (zero). On the other hand, if the adjusted localization exceeds 1, the localization is set to 1. . Note that the calculation of the movement amounts ML1 [2] [f], MR1 [2] [f] by the process of S217 and the localization adjustment by the process of S211 correspond to the sound image enlargement / reduction process. After that, the 1L [f] signal and the 1R [f] signal obtained by the processing of S211 are used for the processing processing of S212 and the processing processing of S213, respectively, and thereby the 1L_2 [f] signal and the 1R_2 [f] signal Get.

  On the other hand, after the processing of S218, coefficients ll ′, lr ′, rl ′, rr ′ are determined using the calculated movement amount ML2 [2] [f] and movement amount MR2 [2] [f], Thus, the localization (the amount output from the sub-speaker) formed by the extracted signal extracted by the second extraction process (S200) is adjusted (S214). In the process of S214, if the adjusted localization is less than 0, the localization is set to 0. On the other hand, if the adjusted localization is greater than 1, the localization is set to 1. Note that the calculation of the movement amounts ML2 [2] [f] and MR2 [2] [f] by the process of S218 and the localization adjustment by the process of S214 correspond to the sound image enlargement / reduction process. Thereafter, the 2L [f] signal and the 2R [f] signal obtained by the process of S214 are respectively used for the process of S215 and the process of S216, whereby the 2L_2 [f] signal and the 2R_2 [f] signal Get.

  As described above, in the effector 1 of the second embodiment, the reference localization and one end of the localization range are extracted from the extraction signal extracted from the extraction region by the first extraction processing (S100) or the second extraction processing (S200). A sound image magnification function YL (f) that defines the degree of enlargement (magnification) of the left boundary that is the side, and a sound image magnification function YR that defines the degree of enlargement of the right boundary that is the other end of the localization range (F) is set, and the extracted signal located to the left of the reference localization is moved by linear mapping with the sound localization function YL (f) with the reference localization as a reference, and is also to the right of the reference localization. The extracted image is moved by linear mapping with the sound localization function YR (f) with reference to the reference localization, thereby enlarging or reducing the sound image formed in the extraction region. Rukoto can. Therefore, according to the effector 1 of the present embodiment, each sound image represented by the stereo sound source can be freely enlarged or reduced.

  Next, a third embodiment will be described with reference to FIGS. 10 and 11. In the third embodiment, the same parts as those in the second embodiment described above are denoted by the same reference numerals, and the description thereof is omitted.

  In the second embodiment described above, the description has been given of enlarging or reducing the sound image of the extracted signal extracted from the musical sound signals (ie, stereo signals) of the left and right channels according to the setting conditions. However, in the third embodiment, 1 An explanation will be given of enlarging or reducing the sound image formed from the extracted signal extracted from the channel tone signal (ie, monaural signal) according to the setting condition.

  Specifically, since the localization of the monaural signal is located at the center (panC), in the present embodiment, before the sound image enlargement / reduction processing is performed, the monaural signal is a monaural signal because it is a monaural signal (panC). A process of distributing (distributing) the extracted signal that is localized to either the left boundary (panL) or the right boundary (panR) of the localization in the extraction region is performed.

  FIG. 10 is a schematic diagram for explaining the outline of the sound image enlargement / reduction processing in the third embodiment. In FIG. 10, ten boxes Po filled in black each indicate one or a plurality of extracted signals from a monaural signal in one frequency range. In FIG. 10, for the purpose of distinguishing the boxes Po, spaces (blanks) are provided between the boxes Po. However, actually, the boxes Po are continuous without gaps (that is, the boxes Po). The frequency range of Po is continuous).

  As shown in FIG. 10, in the present embodiment, as preparation processing, the box Po representing the extracted signal from the monaural signal is assigned to the distribution destination of the box Po adjacent to the frequency domain, and the localization in the extraction areas O1 and O2, respectively. Are distributed alternately at the left boundary (panL) and the right boundary (panR). That is, it is moved to the box PoL or the box PoR.

  Thereafter, as in the second embodiment described above, among the extraction signals in the extraction area, an extraction signal localized on the panL side from panC (that is, the extraction signal included in the box PoL) is supplied to each extraction area O1, O2. The sound image enlargement functions YL [1] (f) and YL [2] (f) provided for it are moved by linear mapping to a region having a left localization boundary. On the other hand, an extraction signal localized on the panR side of panC (that is, an extraction signal included in the box PoR) is converted into sound image expansion functions YR [1] (f), YR [ 2] Moves by linearly mapping to a region having (f) as the left localization boundary.

  As a result, the extraction signal from the monaural signal in the first extraction region O1 (f1 ≦ frequency f ≦ f2) (that is, the extraction signal included in the box Po) is alternately changed for each frequency range. [1] (f) or sound image enlargement function YR [1] (f) is moved to a position corresponding to each frequency, that is, box PtL or box PtR. Similarly, the boxes Po in the second extraction region O2 (f2 ≦ frequency f ≦ f3) are alternately changed for each frequency range by the sound image expansion function YL [2] (f) or the sound image expansion function YR [2] ( f) It is moved to the localization (box PtL or box PtR) corresponding to each frequency on the above.

  As described above, after the localization of the monaural musical sound signal is distributed (assigned) alternately to panL or panR for each predetermined continuous frequency range, the sound image is enlarged or reduced in the same manner as in the second embodiment. By performing the above, it is possible to give a balanced feeling to the monaural signal (monaural sound source).

  In the case where the first extraction area O1 is an area having the middle sound range as the frequency range and the second extraction area O2 is an area having the high sound area as the frequency range, the first extraction area O1 is the same as the example shown in FIG. The sound image magnification function YL [1] (f) and the sound image magnification function YR [1] (f) for the region O1 are functions representing the degree of magnification such that the localization is expanded on the high frequency side, and the sound image magnification for the second extraction region O2 The function YL [2] (f) and the sound image enlargement function YR [2] (f) can be given a favorable audibility by using functions that indicate the degree of enlargement such that the localization is narrowed on the high frequency side.

  In FIG. 10, the case where the localization range of the first extraction region O1 and the localization range of the second extraction region O2 are equal is illustrated, but the localization range of each extraction region O1, O2 may be different. .

  Next, the sound image enlargement / reduction processing according to the third embodiment will be described with reference to FIG. FIG. 11 is a diagram illustrating main processes executed by the effector of the third embodiment. The effector of this embodiment is an A / D converter that converts a monaural tone signal input from the IN_MONO terminal from an analog signal to a digital signal, instead of the Lch A / D converter 11L and the Rch A / D converter 11R. Is different from the effector 1 (see FIG. 1) used in the second embodiment.

  In the present embodiment, since the monaural signal is used as the input signal, the processing performed on the left channel signal and the right channel signal in the effector 1 of the second embodiment is performed on the monaural signal. That is, as shown in FIG. 11, the effector of the present embodiment converts the time domain IN_MONO [t] signal input from the IN_MONO terminal into the frequency domain IN_MONO [f] by the analysis processing unit S50 similar to S10 or S20. It converts into a signal and supplies it to main processing part S30.

  In the state of the monaural signal, the localization w [f] of each signal is all 0.5 (center), that is, panC. Therefore, the process of S31 executed in the main processing unit S30 of the second embodiment described above is Can be omitted. Therefore, in the main processing unit S30 in the present embodiment, first, after clearing the memory (S32), the first extraction process (S100) and the second extraction process (S200) are executed, and each preset condition is set. In addition to extracting signals, signals other than those specified are extracted (S300).

  Since the localization w [f] of each monaural signal is the center (panC), whether or not the localization w [f] of each signal is within the first or second setting range in S100 and S200 of the present embodiment. It is not necessary to determine whether or not. In S100 and S200 of the second embodiment, the maximum level ML [f] is used to extract a signal. In this embodiment, the level of the IN_MONO [f] signal is used. Further, as described above, in the present embodiment, the processing for obtaining the localization w [f] (that is, the processing of S31 in the second embodiment) is omitted because it is a monaural signal. Even so, the process of S31, that is, the process of obtaining the localization w [f] for each frequency obtained by the Fourier transform performed on the IN_MONO [f] signal may be executed.

  After execution of the first extraction process (S100), a preparation process for generating a pseudo stereo signal by distributing (distributing) the localization of the monaural extracted signal to the left and right is executed (S120). In this preparation process (S120), first, it is determined whether or not the frequency f of the extracted signal is within an odd-numbered frequency range in a predetermined continuous frequency range (S121). The predetermined continuous frequency range is, for example, a range obtained by dividing the entire frequency range by a cent unit (for example, 50 cent unit or 100 cent (semitone) unit) or a frequency unit (for example, 100 Hz unit). It is.

  If the frequency f of the extracted signal is within the odd-numbered frequency range by the process of S121 (S121: Yes), the localization w [f] [1] is set to panL [1] (S122). On the other hand, if the frequency f of the extracted signal is within the even frequency range (S121: No), the localization w [f] [1] is set to panR [1] (S123). After the process of S122 or S123, it is determined whether or not the process of S121 has been completed for all the Fourier-transformed frequencies (S124). If the determination of S124 is negative (S124: No), the process of S121 is performed. On the other hand, if the determination in S124 is affirmative (S124: Yes), the process proceeds to the first signal processing 110.

  Therefore, according to this preparatory process (S120), the localization of the extracted signal that satisfies the first condition is performed by localization of the left and right boundaries of the first setting range set for the localization for each continuous frequency range defined in advance. They are alternately distributed so that (panL [1], panR [1]).

  Thereafter, as in the second embodiment, by performing the processing of S117 and the processing of S111, the localization of the extracted signals output from the left and right main speakers is moved, while the processing of 118 and the processing of S114 are performed. By executing the processing, the localization of the extracted signal corresponding to the amount output from the left and right sub-speakers is moved. In the present embodiment, the preparation process (S120) and the processes of S117 and S111 or the processes of S118 and S114 correspond to the sound image enlargement / reduction process.

  On the other hand, after the execution of the second extraction process (S200), a preparation process for the extracted signal extracted by the second extraction process (S200) is executed (S220). Since this preparation process (S220) is performed in the same manner as the above-described preparation process (S110) except that the processing target is the extracted signal extracted by the second extraction process (S200), the description thereof is omitted. . With this preparatory process (S220), the localization of the extracted signal that satisfies the second condition is performed for each continuous frequency range defined in advance, and the localization of the left and right boundaries of the second setting range set for the localization (panL [2 ], PanR [2]).

  Thereafter, as in the second embodiment, by performing the process of S217 and the process of S211, the localization of the extracted signals output from the left and right main speakers is moved, while the process of 218 and the process of S214 are performed. By executing the processing, the localization of the extracted signal corresponding to the amount output from the left and right sub-speakers is moved. In the present embodiment, the preparation process (S220) and the processes of S217 and S211 or the processes of S218 and S214 correspond to the sound image enlargement / reduction process.

  As described above, in the third embodiment, after a monaural musical sound signal is distributed alternately for each predetermined continuous frequency range, the sound image is expanded or reduced, so that it is suitable for a monaural signal. A feeling can be given.

  As described above, the present invention has been described based on each embodiment. However, the present invention is not limited to the above-described embodiment, and various modifications can be easily made without departing from the gist of the present invention. It can be guessed.

  In the first and second embodiments, the extraction signal extraction in the first extraction process (S100) and the second extraction process (S200) is performed using a condition in which the frequency, localization, and maximum level are a set. It is not limited to this. That is, only the frequency, the localization, or the maximum level may be used as a condition for extracting the extraction signal. Alternatively, these may be appropriately combined (for example, a frequency and a localization may be combined).

  For example, when only the frequency is used as a condition for extracting the extraction signal, the determination content of S101 in the first extraction process (S100) is set to “frequency [f] is within a preset first frequency range. It may be changed to “whether or not.” Further, for example, when only the localization is used as a condition for extracting the extraction signal, the determination content of S101 in the first extraction process (S100) is set to “first localization w [f] set in advance. It may be changed to “whether it is within the setting range”. For example, when only the maximum level is used as a condition for extracting the extraction signal, the determination content of S101 in the first extraction process (S100) is set to “maximum level ML [f] is set in advance. It may be changed to “whether it is within the first setting range”. Here, when the determination content of S201 in the second extraction process (S200) is changed in accordance with the determination content change of S101, the same change as the determination content change of S101 may be performed.

  In the first and second embodiments, the condition for extracting the extracted signal is a condition in which the frequency, localization and maximum level are a set, so that noise having a center frequency outside the condition, It is possible to suppress the influence from noise having a level exceeding the condition or noise having a level below the condition. Thereby, an extraction signal can be extracted correctly.

  In S101 and S201 of the first and second embodiments, it is determined whether the frequency f, the localization w [f], and the maximum level ML [f] are within preset ranges. , The localization w [f], and an arbitrary function having at least two of the maximum levels ML [f] as variables, and determining whether a value obtained by the function is within a set range. It may be. Thereby, a more complicated range can be set.

  In each processing (S112, S113, S115, S116, S212, S213, S215, S216, S312, S313, S315, and S316) executed in each of the above embodiments, pitch change, level change, and reverberation were performed. However, these changes and reverberation may be set to the same contents in all the machining processes, or the contents of each machining process may be different. For example, the processing in the first signal processing (S112, S113, S115, S116), the processing in the second signal processing (S212, S213, S215, S216), and the processing in the non-designated signal processing (S312, S313) S315 and S316) may be set to different contents. In addition, when the contents of each processing in the first signal processing, the second signal processing, and the non-designated signal processing are different, different signal processing is performed on each extracted signal extracted for each condition. it can.

  In the first and second embodiments, the left and right two-channel music signal is input to the effector 1 as a signal processing target. However, the left and right channel music signals are not limited to the left and right, and are localized in two directions. The two-channel musical sound signal may be input to the effector 1 as a signal processing target. Further, the tone signal input to the effector 1 may be a tone signal of three or more channels. When a sound signal of three or more channels is input to the effector 1, a localization w [f] (localization information) corresponding to the localization of each channel is calculated, and each calculated localization w [f] falls within the setting range. Or not. For example, in addition to the left and right localization w [f], the vertical and / or front / rear localization is calculated, and whether the calculated left / right localization w [f] and the vertical and / or front / rear localization are within the set range. Determine. Taking the left and right four-channel music signal as an example, the localization of two pairs of music signals (left and right, front and rear) is calculated, and whether the calculated left and right localization and the calculated localization are within the set range Determine.

  In each of the above embodiments, in the extraction process (S100, S200), the amplitude of the tone signal is used as the level of each signal to be compared with the set range (particularly, the maximum level ML [f] in the first and second embodiments). However, a configuration using the power of a musical tone signal may be used. For example, in the first and second embodiments, in order to obtain INL_Lv [f], a value obtained by squaring the real part of the complex expression of the IN_L [f] signal and the imaginary part of the complex expression of the IN_L [f] signal are obtained. The square root of the added value is calculated, and for example, the value obtained by squaring the real part in the complex expression of the IN_L [f] signal and the IN_L [f] signal INL_Lv [f] may be obtained by adding together the value obtained by squaring the imaginary part in the complex expression.

  In the first and second embodiments, the localization w [f] is calculated based on the ratio of the left and right channel signal levels, but may be calculated based on the difference between the left and right channel signal levels.

  In the first and second embodiments, the localization w [f] is uniquely determined for each frequency band from the two-channel musical sound signals. However, based on the localization determined for each frequency band, a plurality of continuous positions are obtained. It is also possible to obtain a localization probability distribution within the frequency band as a group and use the localization probability distribution as localization information (localization w [f]). In such a case, for example, it is possible to extract a desired musical sound signal by determining whether or not a range in which the localization is equal to or greater than a predetermined probability falls within a set range (a range set as a direction range).

  In the first and second embodiments, in S111, S114, S211, S214, S311, and S314, the left and right musical sound signals (that is, the extracted signals) extracted by the respective extraction processes (S100, S200, and S300) are obtained. Based on the localization w [f] and the target localization, the localization formed by the extracted signal is adjusted. Instead, a monaural tone signal is synthesized from the extracted left and right tone signals, for example, by simply adding them together, and extracted based on the target localization for the synthesized monaural tone signal. It is good also as a structure which adjusts the localization formed with a signal.

  Further, in the second embodiment, the coefficients ll, lr, rl, rr and the coefficients ll ′, lr ′, rl ′, rr are set with the target of the target position of the localization for enlarging (or reducing) the sound image. Although 'is calculated, the target localization may be a combination of a moving destination of localization for enlarging (or reducing) a sound image and a moving destination of movement of the sound image itself (movement of the extraction area).

  In each of the embodiments described above, after extracting signals other than the extracted signal and designation by each extraction process (S100, S200, S300), each signal processing (S110, S210, S310) is performed on the extracted signal and signals other than the designation. Then, the obtained signals (that is, the extracted signal after processing and the signal other than the designation) are synthesized for each output channel, and the synthesized signals (OUT_L1 [f], OUT_R1 [f], OUT_L2 [ f] and OUT_R2 [f]), and then applying an inverse FFT process (S61, S71, S81, S91) to the combined signals, the time domain signal for each output channel is obtained. It was set as the structure to obtain.

  Instead, after extracting the extracted signal and signals other than the designation by each extraction processing (S100, S200, S300), each signal processing (processing corresponding to S110 and the like) is performed on the extracted signal and the signals other than the designation. After that, each obtained signal (ie, the extracted signal after processing and the signal other than the designated signal) is subjected to inverse FFT processing (processing corresponding to S61 etc.) to be converted into a time domain signal. After that, by synthesizing each obtained signal (that is, the extracted signal after processing represented in the time domain and the signal other than the designation) for each output channel, a signal in the time domain for each output channel is obtained. Also good. In such a case as well, signal processing on the frequency axis is possible as in the above-described embodiment.

  Alternatively, after extraction signals and signals other than specified are respectively extracted by each extraction process (S100, S200, S300), inverse FFT processing (processing corresponding to S61 and the like) is performed on the extracted signals and signals other than the specification, respectively. Is converted to a time domain signal, and then each signal processing (ie, a process corresponding to S110) is performed on each obtained signal (ie, the extracted signal represented in the time domain and a signal other than that specified). After that, by synthesizing each obtained signal (that is, the extracted signal after processing represented in the time domain and the signal other than the designation) for each output channel, a signal in the time domain for each output channel is obtained. Also good.

  In the first and second embodiments, the maximum level ML [f] is used as one of the conditions for extracting the extraction signal from the left and right channel signals, but instead of the maximum level ML [f], A configuration may be used in which the sum or average of the levels of each frequency band in a plurality of channel signals is used as the extraction condition.

  In each of the above-described embodiments, two extraction processes (first extraction process (S100) and second extraction process (S200)) are performed for extracting the extraction signal. However, three or more extraction processes may be provided. Good. That is, the extraction condition (for example, a condition in which the frequency, localization, and maximum level are set as one set) is not two, but may be three or more. Further, when there are three or more extraction processes for extracting the extraction signal, the signal processing may be increased according to the number of extraction processes.

  In each of the above-described embodiments, a process (S300) for extracting other than the designation is provided and a signal other than the extracted signal in the input musical tone signal such as a left and right channel signal or monaural signal is extracted. It is good also as a structure which does not provide. That is, it may be configured such that signals other than the extracted signal in the input musical sound signal are not extracted. In the case where the process for extracting data other than the designation (S300) is not provided, the signal process other than the designation (S310) may not be provided.

  In each of the above embodiments, one set of output terminals on the left and right is two sets (that is, a set of OUT1_L terminal and OUT1_R terminal, and a set of OUT2_L terminal and OUT2_R terminal). A set may be sufficient and three or more sets may be sufficient. For example, it may be 5.1 channel. When there is one set of output terminals, distribution of each channel signal is not performed in each signal processing, and the range of 0.25 to 0.75 in the graphs of FIGS. The calculation of S111, S211 and S311 may be performed using a graph stretched to 0 to 1.0 (that is, doubled).

  In each signal processing (S110, S210, S310) in each of the above embodiments, a processing process including a change in the localization of the extracted musical sound (extracted signal), a pitch change, a level change, and a reverb addition is performed. However, the signal processing performed on the extracted musical sounds may not always be the same processing. In other words, the execution contents of the signal processing may be appropriately selected for each extraction condition, and the execution contents of the signal processing may be different for each extraction condition. Further, as the contents of the signal processing, other known signal processing may be performed in addition to the localization change, the pitch change, the level change, and the reverb addition.

  In each of the above embodiments, the coefficients ll, rr, rl, rr, ll ', rr', rl ', rr' are linear with respect to the horizontal axis as shown in FIGS. 7 (a) and 7 (b). Although it is assumed to change, the portion that increases or decreases may not be linearly increased or decreased linearly but may be increased or decreased in a curved line (for example, a sine curve).

  In each of the above embodiments, the Hanning window is used as the window function. However, a Hamming window or a Blackman window may be used.

  In the second and third embodiments, the sound image enlargement function YL (f) and the sound image enlargement function YR (f) are functions with different degrees of enlargement or reduction depending on the frequency f, that is, the sound image according to the frequency f. Although the functions of the enlargement function YL (f) and the sound image enlargement function YR (f) are changed, the values of the sound image enlargement function YL (f) and the sound image enlargement function YR (f) are not dependent on the change of the frequency f. The function may be constant. That is, if BtmL = TopL and BtmR = TopR, the sound image enlargement YL (f) and YR (f) are functions whose enlargement or reduction does not depend on the frequency f. Also good.

  In the second embodiment, the sound image enlargement function is YL (f) and YR (f), that is, a function of the frequency f, but from the reference localization (in the above embodiment, the localization center of the extraction region). A function in which the degree of enlargement (or reduction degree) is determined according to the localization difference of the extracted signal (that is, the degree of separation of the extracted signal from panC), for example, a function in which the degree of enlargement increases as the distance from the center increases. . In such a case, the horizontal axis (frequency f) in the diagram shown in FIG. 8 is calculated as the difference in the localization of the extracted signal from panC, which is the reference localization, so that the calculation is the same as the calculation performed in the second embodiment. it can. Further, the frequency f and the difference in localization of the extracted signal from the reference localization (panC) are combined, and the expansion (or reduction) is determined according to the frequency f and the difference in localization of the extracted signal from the reference localization (panC). A function to decide may be used.

  In the second and third embodiments, the sound image expansion function is YL (f) and YR (f), that is, a function of the frequency f. However, when the extraction signal to be processed is a time domain signal. May use a sound image expansion function depending on the time t instead of the function of the frequency f.

  In the third embodiment, the description has been given of performing the sound image enlargement / reduction processing after performing the preparatory processing for alternately distributing each of the monaural input musical tone signals for each predetermined continuous frequency range. However, for example, a monaural tone signal is synthesized by simply adding the left and right channel tone signal, and the same preparation process as that of the third embodiment is performed on the synthesized monaural tone signal, and then the sound image is enlarged or reduced. Processing may be performed.

  In the third embodiment, the localization ranges of the first extraction region O1 and the second extraction region O2 are equal. However, the localization range may be different for each extraction region. Further, the left boundary (panL) and the right boundary (panR) in the extraction region may be set to be asymmetric with respect to the center (panC).

1 Effector (musical sound signal processor)
11L, 11R A / D (input means)
D / A for 13L1, 13L2 Lch
D / A for 13R1, 13R2 Rch
S12, S22 FFT processing (dividing means)
S13, S23 Processing in the analysis processing unit (level calculation means)
S31 Processing in the main processing unit (localization information calculation means)
S60, S80 Output processing unit (combining means, converting means)
S70, S90 Output processing unit (combining means, converting means)
S100, S200 Extraction processing (setting means)
S101, S201 Processing in retrieval processing (determination means)
S102, S202 Processing in extraction processing (extraction means)
S110, S210 Signal processing (signal processing means)
S111, S117 Processing in signal processing (sound image enlargement / reduction means)
S114, S118 Processing in signal processing (sound image enlargement / reduction means)
S211, S217 Processing in signal processing (sound image enlargement / reduction means)
S214, S218 Processing in signal processing (sound image enlargement / reduction means)
S120, S220 Preparation process (preparation means)

Claims (7)

Input means for inputting a musical sound signal composed of signals of one or a plurality of channels;
Dividing means for dividing each channel signal constituting the musical sound signal input to the input means into a plurality of frequency bands;
Level calculation means for obtaining the level of the signal of each channel in each of the frequency bands divided by the dividing means;
Localization information calculating means for calculating localization information indicating an output direction of the musical sound signal with respect to a predetermined first reference localization in association with each of the frequency bands based on the level obtained by the level calculating means;
Setting means capable of setting at least one direction range in the output direction of the musical sound signal;
Determining means for determining whether the output direction of the musical sound signal indicated by the localization information calculated by the localization information calculating means falls within the direction range set by the setting means;
When the determination means determines that the output direction of the musical sound signal indicated by the localization information falls within the direction range set by the setting means, the frequency band corresponding to the localization information determined to fall within Extraction means for extracting the signal of each channel in the musical sound signal as an extraction signal;
The extracted signal extracted by the extracting means is subjected to predetermined signal processing and converted into the time domain, thereby outputting the time domain for each preset output channel corresponding to each channel. Output signal generating means for generating a signal;
Output means for outputting the output signal in the time domain obtained by the output signal generating means for each of the output channels,
The output signal generation means includes
For the direction range set by the setting means, a second reference localization, a first function that defines the degree of enlargement of the boundary on one end side of the direction range, and the degree of enlargement of the boundary on the other end side of the direction range Second setting means for setting a second function that defines
Of the extracted signals extracted by the extracting means from within the direction range, the second reference localization, the first function, and the second function are set by the second setting means. The output direction indicated by the localization information of the first extracted signal located on one end side of the direction range with respect to the second reference localization is moved by a linear mapping according to the first function with respect to the second reference localization. And the output direction indicated by the localization information of the second extracted signal located on the other end side of the direction range with respect to the second reference localization depends on the second function based on the second reference localization. And a sound image enlarging / reducing means for enlarging or reducing the sound image in the direction range by moving the image by the linear mapping. The setting means is configured to be capable of setting a plurality of direction ranges in the output direction of the musical sound signal,
The determination means determines whether the output direction of the musical sound signal indicated by the localization information calculated by the localization information calculation means falls within the direction range set by the setting means for each set direction range. Is to judge
The extraction means determines the localization information determined to fall within the direction range set by the setting means when the determination means determines that the output direction of the musical sound signal indicated by the localization information falls within the direction range. The signal of each channel in the frequency band corresponding to is extracted as an extraction signal from the musical sound signal for each direction range used for the determination,
The output signal generation means includes
A signal processing unit that performs signal processing on the extraction signal extracted for each of the direction ranges by the extraction unit, and obtains an extracted signal after processing;
A synthesizing unit that synthesizes the extracted signal obtained by the signal processing unit for each of the output channels for all of the directional ranges;
Conversion means for converting each of the signals obtained by the synthesis by the synthesis means into a time domain signal to obtain a time domain output signal,
The signal processing means includes the second setting means and the sound image enlargement / reduction means,
2. The second setting unit can set the second reference localization, the first function, and the second function for each direction range set by the setting unit. The musical tone signal processing apparatus as described. The setting means is configured to be capable of setting a plurality of direction ranges in the output direction of the musical sound signal,
The determination means determines whether the output direction of the musical sound signal indicated by the localization information calculated by the localization information calculation means falls within the direction range set by the setting means for each set direction range. Is to judge
The extraction means determines the localization information determined to fall within the direction range set by the setting means when the determination means determines that the output direction of the musical sound signal indicated by the localization information falls within the direction range. The signal of each channel in the frequency band corresponding to is extracted as an extraction signal from the musical sound signal for each direction range used for the determination,
The output signal generation means includes
A signal processing unit that performs signal processing on the extraction signal extracted for each of the direction ranges by the extraction unit, and obtains an extracted signal after processing;
A conversion means for converting the post-processing extraction signal obtained by the signal processing means into a time domain extraction signal;
The time domain extraction signal converted by the conversion means is configured to include synthesis means for synthesizing all the directional ranges for each preset output channel corresponding to each channel,
The signal processing means includes the second setting means and the sound image enlargement / reduction means,
2. The second setting unit can set the second reference localization, the first function, and the second function for each direction range set by the setting unit. The musical tone signal processing apparatus as described. The setting means is configured to be capable of setting a plurality of direction ranges in the output direction of the musical sound signal,
The determination means determines whether the output direction of the musical sound signal indicated by the localization information calculated by the localization information calculation means falls within the direction range set by the setting means for each set direction range. Is to judge
The extraction means determines the localization information determined to fall within the direction range set by the setting means when the determination means determines that the output direction of the musical sound signal indicated by the localization information falls within the direction range. The signal of each channel in the frequency band corresponding to is extracted as an extraction signal from the musical sound signal for each direction range used for the determination,
The output signal generation means includes
Conversion means for converting the extraction signal extracted for each direction range by the extraction means into an extraction signal in a time domain;
Signal processing means that performs signal processing on the time domain extracted signal converted by the converting means, and obtains the extracted signal after time domain processing;
Synthesizing means for synthesizing the extracted signal in the time domain, which has been subjected to signal processing by the signal processing means, for all of the directional ranges for each output channel set in advance corresponding to each channel; Comprising and including
The signal processing means includes the second setting means and the sound image enlargement / reduction means,
2. The second setting unit can set the second reference localization, the first function, and the second function for each direction range set by the setting unit. The musical tone signal processing apparatus as described.   When the extraction signal to be processed is a monaural signal, the sound image enlargement / reduction means determines the output direction indicated by the localization information of the extraction signal as the output direction of one end or the other end of the direction range including the extraction signal. Including a preparatory means for automatically allocating every successive frequency range set in advance, and for the extracted signal distributed by the preparatory means, set for the direction range including the extracted signal. 2. The musical tone signal processing apparatus according to claim 1, wherein the linear mapping is performed using the second reference localization, the first function, and the second function. The signal processing means distributes the signal of each channel to be processed according to the output channel, and performs independent signal processing on each distributed signal,
A plurality of the output means are provided corresponding to the independent signal processing,
The signal processing means includes the second setting means and the sound image enlargement / reduction means,
The second setting means can set the second reference localization, the first function, and the second function for each signal distributed according to the output channel,
The sound image enlargement / reduction means uses the second reference localization set for each signal distributed according to the output channel, the first function, and the second function, and converts the linear mapping to the distribution 6. The musical tone signal processing apparatus according to claim 2, wherein the musical tone signal processing apparatus is performed independently for each signal. Input means for inputting a monaural tone signal;
A dividing means for dividing the monaural musical sound signal input to the input means into a plurality of frequency bands;
Level calculating means for obtaining the level of the monaural signal in each of the frequency bands divided by the dividing means;
Setting means capable of setting at least one set of a direction range including an output direction of the monaural musical sound signal equal to a predetermined first reference localization and a band range of the frequency band divided by the dividing means;
For each set of band ranges set by the setting means, extraction means for extracting a monaural signal in a frequency band that falls within each band range;
The extracted signal extracted by the extracting means is subjected to predetermined signal processing and converted into the time domain, thereby outputting the time domain for each preset output channel corresponding to each channel. Output signal generating means for generating a signal;
Output means for outputting the output signal in the time domain obtained by the output signal generating means for each of the output channels,
The output signal generation means includes
For each set of direction ranges set by the setting means, for each direction range, a second reference localization, a first function that defines the degree of expansion of the boundary on one end of the direction range, and the direction range A second setting means for setting a second function that defines the degree of enlargement of the boundary on the other end side;
A preparation means for automatically assigning the output direction of the extraction signal so as to alternate every predetermined continuous frequency range to the output direction of one end or the other end of the direction range corresponding to the extraction signal;
The direction range in which the second reference localization, the first function, and the second function are set by the second setting unit is extracted from the direction range by the extraction unit and distributed by the preparation unit. Among the extracted signals, the output direction indicated by the localization information of the first extraction signal located at one end side of the direction range from the second reference localization is the first function based on the second reference localization. The output direction indicated by the localization information of the second extracted signal located on the other end side of the direction range with respect to the second reference localization is based on the second reference localization. A musical sound signal processing apparatus comprising: a sound image enlarging / reducing means for enlarging or reducing a sound image within the direction range by moving the linear function according to the second function.

Images