JPH06177676A - Signal processing unit - Google Patents

Abstract

PURPOSE:To apply cross fade to audio signals of plural systems and to move spatially a sound image horizontally while keeping a sound pressure of a sound in a listening sense constant. CONSTITUTION:A signal processing unit having multipliers 25, 26 respectively multiplying audio data DD1 and DD2 with multiplication factors A and (1-A) to select any of the audio data DD1 and DD2 is provided with computing elements 23, 24 calculating multiplication factors A<1/2> and (1-A)<1/2> respectively so as to make the sound pressure constant timewise. The multiplier 25 multiplies the audio data DD1 with the multiplication factor A<1/2> and the multiplier 26 multiplies the audio data DD2 with the multiplication factor (1-A)<1/2>.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing device that executes a microprogram to perform various numerical calculation processes on an input digital signal.

[0002]

2. Description of the Related Art Conventionally, in an electronic musical instrument, an audio device, or the like, when the audio data of one tone color is switched to the audio data of another tone color while playing or playing a music source, noise is generated when the tone data is rapidly switched. Therefore, it is necessary to gradually decrease the volume of audio data of a certain tone color and gradually increase the volume of audio data of another tone color for switching. This is called crossfade. Further, conventionally, in an electronic musical instrument, an audio device, or the like, a certain sound may be spatially panned from left to right or right to left (sound image localization).

By the way, in recent years, the technology of a digital signal processor (digital signal processor (DSP)) that executes a microprogram to perform various numerical calculation processes on an input digital signal has advanced, and semiconductor manufacturing has progressed. Due to technological advances, DSPLS
I is becoming readily available. For this reason, recent electronic musical instruments, audio devices, and the like include a cross-fade circuit for performing the above-described cross-fade and a pan circuit for performing the above-mentioned pan that are configured by a DSP LSI and incorporated therein.

FIG. 12 is a block diagram showing a configuration example of a conventional crossfade circuit constructed by such a DSP LSI. In FIG. 12, 1 and 2 are digital audio data D _I1 and D of different tone colors. Input terminals 3 to which _I2 is respectively input, and 3 are multipliers for multiplying the audio data D _I1 input from the input terminal 1 by the multiplication coefficient k ₁ (0 ≦ k ₁ ≦ 1) having the characteristic shown in FIG. 4 is the audio data D _I2 input from the input terminal 2 and the multiplication coefficient k having the characteristic shown in b of FIG.
₂ (0 ≦ k ₂ ≦ 1) and a multiplier 5 for adding the output data of the multipliers 3 and 4, and 6 for one audio data in which the output data of the adder 5 is mixed. It is an output terminal that is output as D _O. With such a configuration, it is possible to smoothly switch from the audio data D _I1 input from the input terminal ₁ to the audio data D _I2 input from the input terminal 2 by linear interpolation with a simple configuration.

FIG. 14 is a block diagram showing an example of the configuration of a conventional pan circuit constructed by the above DSP LSI. In this figure, 7 is audio data D.
An input terminal for inputting _I , 8 is a multiplication coefficient A (0 ≦ A ≦
1) is an input terminal to which the multiplication coefficient A is input, and 9 is a computing unit that performs the multiplication coefficient (1-x) on the multiplication coefficient A and outputs the multiplication coefficient (1-A).

10 is audio data D _I and multiplication coefficient A
And 11 are output terminals for outputting the output data of the multiplier 10 as R channel audio data D _OR , and 12 are audio data D _I and multiplication coefficient (1
-A) is a multiplier for multiplying with, and 13 is an output terminal from which the output data of the multiplier 12 is output as the L channel audio data D _OL .

In such a configuration, the multiplication coefficient A is set to 0
When changed from 1 to 1, the sound radiated from the speaker (not shown) connected to the output terminals 11 and 13 spatially pans from left to right, and conversely changes the multiplication coefficient A from 1 to 0. Then, the sound emitted from the speaker (not shown) connected to the output terminals 11 and 13 spatially pans from right to left.

[0008]

By the way, the above-mentioned D
In the conventional cross-fade circuit configured by SPLSI, since the arithmetic operations in the multipliers 3 and 4 and the adder 5 are based on the integral sum of products, as shown at the intersection c between a and b in FIG. , And the multiplication coefficients k ₁ and k ₂ are both ½, the output data of the multipliers 3 and 4 are D _I1 / 2 and D _I2 / 2, respectively. Since the power is proportional to the square of the amplitude, the power corresponding to the output data of each of the multipliers 3 and 4 is 1/4 of the audio data D _I1 and D _I2 , respectively.
Will be.

Therefore, even if the output data of each of the multipliers 3 and 4 is added by the adder 5, the power corresponding to the audio data D _O output from the output terminal 6 is the audio data D _I1 or D _I2. It becomes 1/2 as compared with the case of only one of the above. Here, since the electric power and the sound pressure are in a proportional relationship, the sound pressure corresponding to the audio data D _O is also 1/2 as compared with the case of only one of the audio data D _I1 or D _I2. I will end up. That is, there is a drawback that the audible volume is not constant during the crossfade.

Further, in the conventional pan circuit constituted by the above-mentioned DSP LSI, since the arithmetic operations in the arithmetic unit 9 and the multipliers 10 and 12 are based on the integral sum of products, the multiplication coefficient A = 0 and a certain space is set. If the multiplication coefficient A becomes 1/2 when the auditory power of the sound radiated from the speaker (not shown) arranged on the left side of the above is assumed to be P ₀ , the reason similar to that of the crossfade circuit described above is obtained. Accordingly, the power of the acoustic audibility which is respectively emitted from the speaker (not shown) disposed respectively left and the right of a certain space, thus both become P _0/4. Therefore,
As shown in FIG. 15, the audible electric power of the sound is P at portions spatially equidistant from the left and right speakers (not shown).
_It becomes 0/2, and there is a problem that the pan operation becomes unnatural because the sound image cannot be spatially moved to the left and right while the electric power on the acoustic sense, that is, the sound pressure is kept constant. It was

The present invention has been made under such a background, and it is possible to crossfade audio signals of a plurality of systems or to keep the sound image spatially left and right while maintaining a constant sound pressure of the sound. An object is to provide a signal processing device that can be moved.

[0012]

The present invention has a plurality of multiplying means for multiplying an audio signal of one or a plurality of systems and a plurality of multiplication coefficients respectively, and mixing and switching the audio signals of the one or a plurality of systems. Alternatively, a signal processing device for performing localization of a sound image is provided with a plurality of calculating means for calculating the plurality of multiplication coefficients so that the sound pressure becomes constant temporally or spatially.

[0013]

According to the above construction, the sound pressure becomes constant temporally or spatially even if the audio signals of one or a plurality of systems are mixed, switched, or the sound image is localized.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. 1 is a block diagram showing the configuration of a pitch change circuit to which a signal processing device according to a first embodiment of the present invention is applied. In this figure, 14 is an input terminal to which audio data D _I is input, and 15 is a delay circuit. For R
An AM, along with the audio data D _I input from the input terminal 14 to output the write address WA from the address generating circuit 16 shown in FIG. 2 is written, read address RA ₁ is outputted from the address generating circuit 16 and Pitch-modulated audio data D _D1 and D _D2 are read from RA ₂ .

In the address generation circuit 16 of FIG.
Reference numeral 7 denotes a subtraction counter, which counts down in synchronization with the sampling frequency φ and outputs the write address WA. In addition, the subtraction counter 17 includes the delay RAM 15
Counting down from the last address to the first address in sequence, and counting down to the first address,
Count down again from the last address. 18 is
As shown in FIGS. 4A and 4B, low-frequency oscillators (LFOs) 19 for outputting saw-tooth-shaped low-frequency data D _L1 and D _L2 which are 180 ° out of phase with each other, and 19 are write addresses WA and An adder 20 adds the low frequency data D _L1 and outputs the read address RA _1, and an adder 20 adds the write address WA and the low frequency data D _L2 and outputs the read address RA ₂ .

With such a configuration, the audio data D _I is written in the delay RAM 15 in order from the final address to the start address, and the low frequency data D _L1 and D _L2 are written in the write address WA. The audio data D _D1 and D _D2 stored at the added read addresses RA ₁ and RA ₂ are read from the delay RAM 15. Low-frequency data D added to the write address WA to obtain the read addresses RA ₁ and RA ₂ , respectively.
_{Since L1} and D _L2 are sawtooth, the read address RA ₁
And RA ₂ are sequentially delayed with respect to the write address WA, and the pitch of the audio data D _D1 and D _D2 is lowered according to the inclination of the sawtooth low frequency data D _L1 and D _L2 .

Further, in FIG. 1, reference numeral 21 is an input terminal to which the multiplication coefficient A (0 ≦ A ≦ 1) having the characteristic shown in FIG. 4 (d) is input, and 22 is (1-x ), And the multiplication coefficient (1
-A) is an arithmetic unit, and 23 is √ for the multiplication coefficient A.
an arithmetic unit for performing an operation of x and outputting a multiplication coefficient √A, 24
Is an arithmetic unit that performs a calculation of √x on the multiplication coefficient (1-A) and outputs the multiplication coefficient √ (1-A). 25 is a multiplier for multiplying the audio data D _D1 by the multiplication coefficient √A,
26 is a multiplier for multiplying the audio data D _D2 by the multiplication coefficient √ (1-A), 27 is an adder for adding the output data of the multiplier 25 and the output data of the multiplier 26, 28 is an adder 27 The output terminal outputs the output data as audio data D _O.

Next, FIG. 3 is a block diagram showing the configuration of the arithmetic unit 23. In FIG. 3, 29 is an input terminal to which the input data x _n is input, 30 is a subtractor to which the input data x _n is input to the first input end, 31 is the output data of the subtractor 30 to the first input end Is input, and 32 is an adder 3
The output data of 1 is output as the output data y _n . 33 the output data of the adder 31 is delayed by one calculation time period is a delay circuit for outputting a data y _n-1, the data y _n-1 is input to the second input of the adder 31 It is also input to the multiplier 34. The multiplier 34 squares the data y _n-1 and outputs the output data y _n-1 ²
Is input to the second input terminal of the subtractor 30.

Here, when the calculation of the calculator 23 is expressed by a recurrence formula, the formula (1) is obtained. In _{y n = x n -y n-} 1 2 + y n-1 ··· (1) (1) formula, when x _n = A, data y _n n →
The limit for ∞ converges and becomes the root of the quadratic equation shown in equation (2), so that y _n = √A is obtained from the calculator 23. y = A−y ² + y (2) Since the arithmetic unit 24 has the same configuration as the arithmetic unit 23, its description is omitted.

In such a configuration, when the audio data D _I is input from the input terminal 14, the audio data D _I is written to the write address WA of the delay RAM 15 generated by the address generation circuit 16, and The pitch-modulated audio data D _D1 and D _D2 are read from the read addresses RA ′ ₁ and RA ′ _{2 of} the delay RAM 15 generated by the address generation circuit 16.

On the other hand, the multiplication coefficient A (0, which is changed from the waveform shown in FIG. 4D, is input from the input terminal 21.
≦ A ≦ 1), the arithmetic unit 22 performs the arithmetic operation (1-x), and after the arithmetic unit 22 outputs the multiplication coefficient (1-A), the arithmetic unit 24 calculates the √x. Then, the multiplier 24 outputs the multiplication coefficient √ (1-A) having the characteristic shown in FIG. Further, in the arithmetic unit 23, the multiplication coefficient A is calculated by √x, and the arithmetic unit 2
3 outputs the multiplication coefficient √A having the characteristic shown in FIG.

As a result, the multiplier 25 multiplies the audio data D _D1 by the multiplication coefficient √A, and the multiplier 2
6 multiplies the audio data D _D2 by the multiplication coefficient √ (1−A), the adder 27 adds the output data of the multiplier 25 and the output data of the multiplier 26, and the pitch-changed audio is added. The data D _O is output via the output terminal 28.

As described above, the multiplication coefficient A is conventionally used.
And (1-A) in multipliers 25 and 26
Since the audio data D _D1 and D _D2 are multiplied as they are, the multiplication coefficients A and (1-A) are changed from 0 to 1 and from 1 to 0, as shown in FIG. Although the sound output from the speaker (not shown) has a audible reduction in the volume periodically, and a tremolo effect is provided, but according to the first embodiment described above, the multiplication coefficients √A and √. Since (1-A) is multiplied by the audio data D _D1 and D _D2 in the multipliers 25 and 26, respectively, as shown in FIG. 5B, the multiplication coefficients A and (1-A) are 0 to 1 and 1 to 0
Even in the part where the transition to, the volume of the sound does not decrease in the sense of hearing and is smoothly connected.

Next, a second embodiment of the present invention will be described. FIG. 6 is a block diagram showing the configuration of a pan circuit to which the signal processing apparatus according to the second embodiment of the present invention is applied. In this figure, parts corresponding to the respective parts in FIG. The description is omitted. In FIG.
35 is a multiplier for multiplying the audio data D _I and the multiplication coefficient √A; 36 is an output terminal from which the output data of the multiplier 35 is output as R channel audio data D _OR ;
37 is a multiplier for multiplying the audio data D _I by the multiplication coefficient √ (1-A), 38 is output data of the multiplier 37 is L
It is an output terminal for outputting the audio data D _OL of the channel.

In such a configuration, the multiplication coefficient A is set to 0
When changed from 1 to 1, the sound radiated from the speaker (not shown) connected to the output terminals 36 and 38 is spatially panned from left to right, and conversely, the multiplication coefficient A is changed from 1 to 0. Then, the sound radiated from the speaker (not shown) connected to the output terminals 36 and 38 spatially pans from right to left.

In this case, when the power corresponding to the sound of one channel is P ₀ , the power corresponding to the sound of the R channel is A × P ₀ , and the power corresponding to the sound of the L channel is (1- A) × P _0, and therefore, the perceptual power of sound, that is, the sound pressure, at the portions spatially equidistant from the left and right speakers (not shown) is represented by the multiplication coefficient A as shown in equation (3). It is kept constant regardless of the value of. _{A × P 0 + (1-} A) × P 0 = P 0 ··· (3)

By the way, in any of the first and second embodiments described above, multiplication factors √A and √ (1- (1) for pitch changing and panning using the arithmetic units 23 and 24 in either case. By obtaining A), the audible volume of the sound was kept constant even in the portion where the multiplication coefficient A was changed from 0 to 1 and from 1 to 0.

Next, on the basis of the above-mentioned technical idea, the multiplication coefficient A changes from 0 to 1 and 1 contrary to this technical idea.
It may be possible to increase or decrease the sound volume of the auditory sense in the transition from 0 to 0. To do this, the function x ² , (2x−x ² ) or ³ √
x or the like may be used. These functions x ² , (2x-
shown in FIGS. 7-9 the structure of the arithmetic unit 39 to 41 for calculating the x ²⁾ and ³ √x.

In the arithmetic unit 39 of FIG. 7, 42 is a multiplier for squaring the input data x, and in the arithmetic unit 40 of FIG.
43 is a multiplier for multiplying the input data x by the multiplication coefficient 2;
44 is a multiplier for squaring the input data x, 45 is a multiplier 4
3 is a subtractor that subtracts the output data of the multiplier 44 from the output data of 3. In addition, in the arithmetic unit 41 of FIG.
Is a subtractor whose input data x _n is input to the first input terminal,
Reference numeral 47 is an adder to which the output data of the subtractor 46 is input to the first input terminal, and 48 is a delay circuit which delays the output data of the adder 47 by one operation time and outputs it as data y _n-1 . The data y _n-1 is input to the second input terminal of the adder 47 and the multiplier 49 that squares the input data. 50 is a multiplier for multiplying the output data y _n-1 ² and the data y _n-1 of the multiplier 49, and inputs the output data y _n-13 to the second input of the subtractor 46.

Here, when the calculation of the calculator 41 is expressed by a recurrence formula, the formula (4) is obtained. In _{y n = x n -y n-} 1 3 + y n-1 ··· (4) (4) equation, when x _n = A, data y _n n →
Extreme for ∞ is converged, (5) since the roots of a cubic equation in the expression, y = ³ {square root} A is obtained from the arithmetic unit 41. y = A−Y ³ + y (5)

The arithmetic units 39 to 41 described above are used as the above-mentioned pitch change circuit (see FIG. 1) and pan circuit (see FIG. 6).
If used instead of the arithmetic units 23 and 24 of FIG. 12 or provided at the multiplication coefficient input terminals of the multipliers 3 and 4 of the crossfade circuit of FIG. 12, the multiplication coefficients A ² , (2A
-A ² ), √A or ³ √A changes as shown in FIG. 10 with respect to the change of the multiplication coefficient A, whereby the sound pressure P on the auditory sensation corresponds to the change of the multiplication coefficient A. As shown in FIG. 11, the sound effects that are different from the conventional ones can be obtained.

For example, when the arithmetic units 39 to 41 are applied to the pan circuit shown in FIG. 6, the localization of the sound image moves left and right, and the localization of the sound image moves away from or approaches the listener. When the arithmetic units 39 to 41 are applied to the crossfade circuit shown in FIG. 12, when the audio data D _{I1 is switched} to the audio data D _I2 or when the audio data D _{I2 is switched} to the audio data D _I1 , The sound pressure can be lowered to make it less noticeable, or conversely, the sound pressure can be raised to make it more noticeable.

The pitch change circuit described above,
Since the pan circuit and the crossfade circuit are configured by the signal processing device (DSP) as described above, the microprograms corresponding to these circuits are actually executed in the DSP. Therefore, since the arithmetic units 23, 24, 39 to 41 described above can be easily realized, the conventional linear interpolation technique can be used as it is, and it is not necessary to add a special circuit.

[0034]

As described above, according to the present invention, it is possible to crossfade audio data of a plurality of systems or move a sound image spatially left and right while maintaining a constant sound pressure on the audible sound. The effect is that you can.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a pitch change circuit to which a signal processing device according to a first embodiment of the present invention is applied.

FIG. 2 is a block diagram showing a configuration of an address generation circuit 16.

FIG. 3 is a block diagram showing a configuration of a computing unit 23.

FIG. 4 is a diagram showing an example of waveforms of read addresses RA ′ ₁ and RA ′ ₂ and an example of waveforms of multiplication coefficients output from each unit of the circuit shown in FIG. 1.

FIG. 5 is a diagram for comparing and explaining the first embodiment of the present invention and the conventional technique.

FIG. 6 is a block diagram showing a configuration of a pan circuit to which a signal processing device according to a second embodiment of the present invention is applied.

FIG. 7 is a block diagram showing a configuration of an arithmetic unit 33.

FIG. 8 is a block diagram showing a configuration of a computing unit 34.

FIG. 9 is a block diagram showing a configuration of a calculator 35.

FIG. 10 is a diagram showing an example of characteristics of multiplication coefficients A, A ² , (2A−A ² ), √A and ³ √A.

FIG. 11 is a diagram showing an example of characteristics of perceptual power P with respect to multiplication coefficients A, A ² , (2A−A ² ), √A and ³ √A.

FIG. 12 is a block diagram showing a configuration example of a conventional crossfade circuit.

FIG. 13 is a diagram showing an example of characteristics of multiplication coefficients k ₁ and k ₂ .

FIG. 14 is a block diagram showing a configuration example of a conventional pan circuit.

FIG. 15 is a diagram for explaining a disadvantage of the pan circuit shown in FIG.

[Explanation of symbols]

22-24, 39-41 ... Operation unit, 25, 26, 3
4, 35, 37, 42 to 44, 49, 50 ... Multiplier,
27, 31, 47 ... Adder, 30, 45, 46 ... Subtractor, 33, 48 ... Delay circuit.

Claims (1)

[Claims] 1. A plurality of multiplying means for respectively multiplying an audio signal of one or a plurality of systems and a plurality of multiplication coefficients, and mixing, switching, or localization of a sound image of the audio signals of the one or a plurality of systems. A signal processing device for performing the signal processing device, comprising: a plurality of calculating means for calculating the plurality of multiplication coefficients so that the sound pressure becomes constant temporally or spatially.