JP2001356800A - Formant addition device - Google Patents

Abstract

(57) [Problem] To provide a formant adding device capable of changing the formant characteristics of a modulated audio signal continuously and intuitively with a simple operation. SOLUTION: A waveform table 2 for storing a plurality of waveform elements in advance, a two-dimensional operation is possible, and a synthesis ratio of the waveform elements is set in the X-axis direction, and a reading speed of the synthesized waveform elements is set in the Y-axis direction. A pad controller 3 and an FIR filter 4 are provided. When the operator designates two-dimensional coordinates on the touch panel 3a of the pad controller 3, the composition ratio of the waveform elements and the reading speed of the composite waveform elements corresponding to the designated coordinates are set, and the FIR filter is accordingly set. The coefficients of the four multipliers 4a-1 to 4a-n are rewritten. Therefore, the input audio signal passing through the FIR filter 4 is
It is modulated by the formant characteristics of the synthesized waveform element.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a formant adding device capable of adding desired formant characteristics to an audio input signal and outputting the added signal, and more particularly to a formant adding device such as a pad controller.
The present invention relates to a technique that enables formant switching and formant shift with a simple operation by using a dimension controller.

[0002]

2. Description of the Related Art A vocoder is known as a device for adding a desired formant to an input audio signal. A vocoder obtains a unique effect by adding a formant characteristic obtained from a modulated audio signal to a modulated audio signal (general term for all sounds such as musical instrument sounds, human voices, and animal sounds). For example, enter the sound of the piano as the modulated audio signal, if you enter the sound that you say "Hello" in the human voice as a modulated audio signal, because the sound of the piano is modulated by the formant characteristics of the human voice, as if piano output sound, such as talking with the "Hello", and performance effects in social fields such as music hall, is extremely useful recording, in live performance or the like.

FIG. 6 is a block diagram showing a configuration of a conventional vocoder 101. As shown in FIG. As shown in the figure, the modulated audio signal input from the input terminal T1 is band-passed by a plurality of (in this case, n) first BPFs (band-pass filters) 102-1 to 102-n having different pass frequency bands. The first and second BPFs 102-1 to 102-n are further divided into envelope followers 103-1 to 103.
-n determines the amplitude envelope.

Further, the modulated voice signal input from the input terminal T2 is divided into bands by second BPFs 104-1 to 104-n, and envelope followers 103-1 to 103-1 are provided for each band.
03-n is multiplied by the amplitude envelope, and the signal after the multiplication by the amplitude envelope is added by the adder 105. Therefore, the output sound output from the output terminal T3 is the modulated sound signal. It becomes a signal modulated by the modulated audio signal. That is, it is possible to add a modulated sound signal, such as a musical instrument sound or an animal cry, to the modulated sound signal and output the modulated sound signal.

On the other hand, a harmonizer is known as a device capable of independently changing the formant characteristics and the pitch of an input voice input signal.

FIG. 7 is a block diagram schematically showing a configuration of a conventional harmonizer. As shown, the harmonizer 110 includes a buffer 111 for temporarily storing a voice input signal provided from an input terminal T10, a pitch detector 112 for detecting a pitch of the voice input signal, and a voice input signal stored in the buffer 111. A waveform element cutout processing unit 113 for performing processing to cut out a waveform element having a basic period of a signal (or a period twice as long as this) using a method such as applying a window function, an FIR filter 114, and a pulse generator 11
5 and a read speed control unit 119.
The keyboard 117 further includes a keyboard 117 capable of inputting a chord. The output of the keyboard 117 and the output of the pitch detector 112 are selectively switched by a switch 116 and connected to a pulse generator 115.

The harmonizer 11 configured as described above
At 0, when a voice input signal is given to the input terminal T10, the fundamental period is detected by the pitch detector 112, and this voice input signal is temporarily stored in the buffer 111. Based on the basic period detected by the pitch detector 112, the waveform element cutout processing unit 113
11 performs a process of extracting a waveform element from the audio input signal stored in the FIR filter 114. Further, each of the multipliers 118-1 included in the FIR filter 114 corresponds to the amplitude of the extracted waveform element.
The coefficient of ~ 118-n is set in the range of -1 to +1.

Next, a pulse signal having the same cycle as the basic cycle detected by the pitch detector 112 or a cycle having a predetermined magnification is output from the pulse generator 115, and the pulse signal is subjected to a filtering process by the FIR filter 114. Therefore, it is possible to obtain the output sound at the output terminal T11 while maintaining the formant characteristics of the sound input signal. Furthermore, the pitch of the output sound can be changed by appropriately changing the pulse interval of the pulse signal output from the pulse generator 115.

Further, the formant characteristic can be shifted by adding the processing of time compression or time expansion of the waveform extracted by the waveform element extraction processing section 113 by the read speed control section 119. Further, if the switch 117 is switched to select the keyboard 117, a chord can be output. This operation enables various operations such as, for example, converting a male voice into a female voice and outputting it, and outputting a chord based on a voice uttered by one person.

[0010]

However, the conventional vocoder 101 shown in FIG. 6 receives a modulated voice signal, modulates a modulated voice signal such as a musical instrument sound with the modulated voice signal, and modulates the modulated voice signal. This is a configuration in which a formant characteristic is added to the modulated audio signal. In order to change the formant characteristic, it is necessary to input the modulated audio signal each time.

The conventional harmonizer shown in FIG. 7 also has a disadvantage that an input audio signal must be input each time the formant is changed in characteristics. The present invention has been made to solve such a conventional problem, and an object thereof is to change a formant characteristic without inputting a modulation signal from the outside and to shift a formant characteristic. It is an object of the present invention to provide a formant adding device that can be used.

[0012]

In order to achieve the above object, an invention according to claim 1 of the present application is directed to a formant adding apparatus for adding a desired formant characteristic to an input modulated audio signal. A two-dimensional manipulator having an operation axis and a second operation axis, a waveform table in which a plurality of formant waveforms are stored in advance, and one formant waveform or two or more formant waveforms stored in the waveform table. The modulated voice, comprising: a waveform synthesizing means for outputting a waveform synthesized at a ratio of; a read speed control means for controlling a read speed of a formant waveform output from the waveform synthesizing means; and a plurality of multipliers. A digital filter for applying a filter process to the signal to output an output audio signal,
The operation axis is assigned a synthesis ratio of the formant waveform stored in the waveform table, the second operation axis of the two-dimensional operation element is assigned a reading speed in the reading speed control means, and the two-dimensional operation element is The coefficient of each multiplier mounted on the digital filter is changed according to the synthesis ratio of the formant waveform obtained by the operation and the readout speed of the formant waveform to change the formant characteristics of the output audio signal. is there.

According to a second aspect of the present invention, the two-dimensional operation device includes a planar touch panel for detecting a two-dimensional position touched by a finger of an operator, and the first operation axis. And the second operation axis is set to be orthogonal to each other on the touch panel.

According to a third aspect of the present invention, the first operation axis direction is divided into a plurality of regions, each region is assigned a formant waveform, and the formant waveform is The combination ratio is set to change within a range of a region adjacent to the region.

According to a fourth aspect of the present invention, the composite ratio of the formant waveform has a peak value at a substantially central portion of a region to which the formant waveform is assigned, and gradually attenuates when reaching a region adjacent to this region. It is characterized by forming a trapezoidal shape. The invention according to claim 5 is characterized in that an equalizer circuit for enhancing a high-frequency component included in the modulated audio signal is provided at a stage preceding the digital filter.

According to a sixth aspect of the present invention, there is provided an effect balance circuit for synthesizing an audio signal obtained through the digital filter and the modulated audio signal at a desired ratio and outputting the synthesized signal. And According to a seventh aspect of the present invention, the digital filter is an FIR filter.

In the present invention configured as described above, by operating the two-dimensional operation element in the direction of the first operation axis, the synthesis ratio of each formant waveform stored in the waveform table is set, and the synthesized waveform is changed. Generated. By operating the two-dimensional operation element in the direction of the second operation axis, the read speed of the generated composite waveform is set. Then, a process of rewriting the coefficient of the multiplier of the digital filter is performed according to the generated synthesized waveform and the reading speed.

The input modulated signal passes through the digital filter, thereby adding a formant characteristic. Thereby, the operator can change the formant characteristics continuously and intuitively.

[0019]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration of a formant adding apparatus according to an embodiment of the present invention. As shown in FIG. 1, the formant adding apparatus 1 includes a waveform table 2 for storing a plurality of waveform elements (formant waveforms) having desired formant characteristics, and a pad controller (two-dimensional operation element) capable of inputting two-dimensional coordinates. ) 3, an FIR filter (digital filter) 4, an equalizer circuit 5,
And an effect balance circuit 6, a waveform synthesis processing section (waveform synthesis processing means) 7, and a read speed control section (read speed control means) 8.

The waveform table 2 stores a waveform element having a desired formant characteristic in the form of a frequency spectrum signal, and performs a spectrum analysis on an arbitrary sound source (a voice uttered by a human or a musical instrument). And a waveform element obtained by adding a window process (window function process) to two periods of the periodic waveform generated by the sine wave addition synthesis. In the present embodiment, as an example, FIGS.
The vowels "A", "I",
Five types of waveform elements (formant waveforms) of “U”, “E”, and “O” are stored. All stored waveform elements have the same time length, and the overtone components of each waveform element are unified with an initial phase angle of zero.

The pad controller 3 has a touch panel 3a having a planar shape, and detects two-dimensional coordinates of a position touched by an operator on the touch panel 3a. FIG.
FIG. 3 is an explanatory diagram showing a manner in which a waveform element (formant waveform) stored in the waveform table 2 is assigned to the X axis (first operation axis) of the touch panel 3a. As shown in FIG. The X-axis direction on 3a has five regions R
1 to R5, and five waveform elements “A”, “I”, “U”, “E”, and “O” are assigned to each area.

Further, as shown in FIG.
The formant waveforms assigned to 1 to R5 are set so that the gain of the formant waveform becomes a trapezoidal shape with a peak at the center of the region. Further, the gains of the adjacent formant waveforms are assigned so as to overlap. In this case, the sum of the gains is set to 1.

The waveform synthesizing unit 7 includes a plurality (five in this case) of multipliers 7a-1 to 7a-5 and an adder 7b. In accordance with the synthesis ratio of, the processing for synthesizing the respective waveform elements stored in the waveform table 2 to generate a synthesized waveform element is performed.

FIG. 4 is an explanatory view showing the manner in which the waveform readout speed is assigned to the Y axis (second operation axis) of the touch panel 3a. The waveform readout speed increases as the Y coordinate increases. Is set. That is, the lower end position of the touch panel 3a (the position where the Y coordinate is the smallest)
In the example shown in FIG. W1, the waveform reading speed is set to be 50% of the original waveform, and at the center of the touch panel 3a, as shown in FIG. %, And at the upper end position of the touch panel 3a (the position where the Y coordinate is the largest), the waveform reading speed is set to be 200% of the original waveform.

The read speed controller 8 changes the read speed of the synthesized waveform element output from the waveform synthesizer 7. The read speed controller 8 actually performs the waveform synthesis at the read speed set by the pad controller 3. The processing of reading out the synthesized waveform elements output from the processing unit 7 is performed. Therefore, the operator moves the surface of the touch panel 3a in the vertical direction (Y-axis direction).
, The reading speed of the synthesized waveform element becomes 50% ~
It changes continuously in the range of 200%.

The number of FIR filters 4 is n (for example, n =
171) multipliers 4a-1 to 4a-n and m (m = n-
1) delay units 4b-1 to 4b-m and m adders 4c-1 to
4c-m. Then, each multiplier 4a-1
The multiplication coefficients of 4a-n are appropriately rewritten according to the reading speed of the synthesized waveform element set by the reading speed control unit 8.

The equalizer circuit 5 emphasizes high-frequency components contained in an audio input signal (modulated audio signal) such as a musical instrument sound and removes unnecessary low-frequency components. The output signal of the equalizer circuit 5 is amplified by an amplifier 9 and then applied to an FIR filter 4.

The effect balance circuit 6 includes two multipliers 6a-1 and 6a-2 and an adder 6b. By changing the multiplication coefficient of each of the multipliers 6a-1 and 6a-2. , An output audio signal obtained through the FIR filter 4 and an original audio input signal not passing through the equalizer circuit 5 and the FIR filter 4 at a desired ratio.

Next, the operation of this embodiment configured as described above will be described with reference to the flowchart shown in FIG. When the operator touches an arbitrary point on the touch panel 3a of the pad controller 3, the touch panel 3a
The two-dimensional coordinates of the touched position are detected. As described above, the X axis of the touch panel 3a is divided into a plurality of regions, and the gains of different waveform elements (formant waveforms) are set in the respective regions. When the position of the indicated point P1 is designated, the gain of the waveform element according to the X coordinate of this position is selected (step ST1).

In this case, as shown in FIGS. 3C to 3G, the gain of the waveform element “A” and the gain of the waveform element “I” are 0.5, respectively, and the waveform elements “U” and “E”. , "O" are zero, so that the coefficients of the multipliers 7a-1 to 7a-5 in the waveform synthesizing unit 7 are "0.
5,0.5,0,0,0 ".

Accordingly, the adder 7b synthesizes the respective waveform elements shown in FIG. 2 at the above-mentioned ratios, so that "A" and "I" are synthesized by 50%, respectively, as shown by reference numeral S1 in FIG. It is possible to obtain a synthesized waveform element having the characteristics described above (step ST2).

At this time, all the waveform elements stored in the waveform table 2 have the same period and the same time length, and the initial phase angles of the respective harmonics between the waveform elements are unified. No offset occurs when combining elements.

Further, according to the Y coordinate detected by the pad controller 3, the reading speed control unit 8 shown in FIG.
Since the reading speed of the synthesized waveform element generated by the waveform synthesis processing unit 7 is set, the synthesized waveform element is read at this reading speed. That is, as shown by W1 to W3 in FIG. 4, a process of compressing or expanding the synthesized waveform element is added. Then, according to the level of the read composite waveform element, the FI
The coefficients of the multipliers 4a-1 to 4a-n of the R filter 4 are set (step ST3).

When a modulated voice signal such as a musical instrument sound, an animal squeal, a plosive sound, or a human utterance sound is input from the input unit indicated by reference numeral T21 in FIG. 1 (step ST4), the modulated voice signal is input. The high frequency component is emphasized by the equalizer circuit 5 (step ST5). After that, it is supplied to the FIR filter 4 and subjected to convolution processing (step ST).
6). At this time, the multipliers 4a-1 to 4a-4 of the FIR filter 4
Since the coefficient of an is set by the above processing,
The output signal of the FIR filter 4 is applied to the pad controller 3
The signal has the formant characteristics set in (1) and is a formant-shifted signal.

Next, the signal subjected to the convolution processing is applied to the effect balance circuit 6. On the other hand, the original modulated audio signal that does not pass through the equalizer circuit 5 and the FIR filter 4 is also supplied to the effect balance circuit 6. Then, the two signals are determined for their synthesis ratio in accordance with the multiplication coefficients of the multipliers 6a-1 and 6a-2, added by the adder 6b, and output as an audio output signal from the output unit T22 to the outside. (Step ST7).

According to such a configuration, when a modulated voice signal such as a musical instrument sound is provided to the input unit T21, the modulated voice signal is modulated by the synthesized waveform element output from the waveform synthesis processing unit 7. Therefore, it is possible to obtain an output audio signal having a formant characteristic of the waveform element stored in the waveform table 2 or a synthesized waveform element obtained by synthesizing a plurality of waveform elements. In addition, since the reading speed can be appropriately changed by the reading speed control unit 8, the formant characteristics of the output audio signal can be shifted.

For example, when a piano sound is input as a modulated voice signal and the pad controller 3 is operated to select a waveform element “A” set in the waveform table 2, it is as if the piano is “A”. ”Can be obtained. When the finger is traced rightward on the X axis on the touch panel 3a of the pad controller 3, the selected waveform element changes in the order of "A, I, U, E, O" and is adjacent to each other. Since the waveform elements overlap in some areas, for example, when the point touched by the finger is moved from “A” to “I”, the formant characteristic gradually changes from “A” to “I”, and the formant characteristic changes. The characteristics can be changed continuously. At this time, as described above, since each waveform element does not cancel each other, it is possible to continuously change between different formant characteristics without a sense of incongruity. Further, when a finger touches a position between “A” and “I” with a finger, a formant characteristic obtained by combining two formant characteristics can be obtained.

Further, when the finger is traced on the touch panel 3a in the Y-axis direction (vertical direction), the reading speed of the synthesized waveform element changes, so that the synthesized waveform element can be compressed or expanded temporally. Thereby, the formant characteristic can be shifted. For example, when a waveform element of “A” is given, an operation is performed so that “A” uttered by a male is switched to “A” uttered by a female. Can be.

As described above, according to the formant adding apparatus according to the present embodiment, the formant characteristic of the output audio signal can be changed and the formant characteristic can be shifted by operating the pad controller 3. This allows continuous and intuitive operation of the formant characteristics.

In the above embodiment, the composition ratio of formants is allocated in the X-axis (horizontal axis) direction, and the readout speed of the composite waveform element is allocated in the Y-axis (vertical axis) direction. However, the present invention is not limited to this, and it is also possible to adopt a configuration in which the composition ratio of formants is assigned in the Y-axis direction and the reading speed of the combined waveform element is assigned in the X-axis direction.

In the above-described embodiment, an example in which the X axis and the Y axis are orthogonal to each other has been described. However, it is not always necessary that the X axis and the Y axis are orthogonal, and the directions of the X axis and the Y axis may be arbitrarily set on the touch panel 3a. Can be.

Further, in the above-described embodiment, an example in which the pad controller 3 is used as a two-dimensional operation element has been described. However, the present invention is not limited to this.
Other operators that can input dimensional information can be used.

In the above embodiment, five waveform elements "A, I, U, E, O" have been described as examples of waveform elements stored in the waveform table 2, but the present invention is not limited to this. Instead, various formant waveforms can be stored in the waveform table 2 as waveform elements.

Further, in the above-described embodiment, an example in which two waveform elements are combined to obtain a combined waveform element has been described.
The present invention is not limited to this, and it is also possible to combine three or more waveform elements to generate a synthesized waveform element.

[0045]

As described above, according to the formant adding apparatus according to the present invention, the formant waveform stored in the waveform table is selected by operating the two-dimensional operation element in the direction of the first operation axis. However, this can be changed continuously. Further, by operating the two-dimensional operation element in the direction of the second operation axis, the formant characteristics obtained from the synthesized waveform element can be continuously shifted. Therefore, there is no need to input a modulated signal having formant characteristics to be added to the modulated signal as an audio signal as in a conventional vocoder or a harmonizer, and a simple operation of operating a two-dimensional operation unit allows a simple operation. A process for adding a formant characteristic to the modulated signal can be performed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a formant adding apparatus according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing characteristics of waveform elements stored in a waveform table.

FIG. 3 is an explanatory diagram showing an operation position of an X-axis direction of a pad controller and a ratio of a gain of each waveform element.

FIG. 4 is an explanatory diagram showing a relationship between an operation position of a pad controller in a Y-axis direction and a waveform element reading speed.

FIG. 5 is a flowchart showing the operation of the embodiment.

FIG. 6 is a block diagram showing a configuration of a conventional vocoder.

FIG. 7 is a block diagram showing a configuration of a conventional harmonizer.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 formant adding device 2 waveform table 3 pad controller (two-dimensional controller) 3a touch panel 4 FIR filter (digital filter) 4a-1 to 4a-n multiplier 5 equalizer circuit 6 effect balance circuit 7 waveform synthesis processing section (waveform synthesis processing) Means) 8 Read speed control unit (Read speed control means) 9 Amplifier

Claims (7)

[Claims] 1. A formant adding apparatus for adding a desired formant characteristic to an input modulated audio signal, comprising: a two-dimensional operation device having a first operation axis and a second operation axis; A waveform table stored in advance, one formant waveform stored in the waveform table,
Alternatively, waveform synthesizing processing means for outputting a waveform obtained by synthesizing two or more formant waveforms at a predetermined ratio; reading speed control means for controlling a reading speed of the formant waveform output from the waveform synthesizing processing means; And a digital filter for applying a filtering process to the modulated audio signal to output an output audio signal. The digital filter being stored in the waveform table on a first operation axis of the two-dimensional operator. Assigning the read-out speed in the read-out speed control means to the second operation axis of the two-dimensional operator, and synthesizing the formant waveform obtained by operating the two-dimensional operator, And changing the coefficient of each multiplier mounted on the digital filter in accordance with the read speed of the formant waveform and outputting the output audio signal. Formant additional apparatus characterized by varying the formant characteristics. 2. The two-dimensional operation device includes a planar touch panel for detecting a two-dimensional position touched by a finger of an operator, and the first operation axis and the second operation axis are 2. The formant addition device according to claim 1, wherein the setting is performed on the touch panel at right angles to each other. 3. The first operation axis direction is divided into a plurality of regions, each region is assigned a formant waveform, and the formant waveform is divided into the region and a region adjacent to the region. The formant addition device according to claim 1, wherein the combination ratio is set to change within a range. 4. The composition ratio of the formant waveform is such that a peak value is obtained at a substantially central portion of a region to which the formant waveform is assigned, and a trapezoidal shape is obtained in which a region adjacent to the region is gradually attenuated. The formant adding device according to claim 3, characterized in that: 5. An equalizer circuit for enhancing a high-frequency component included in the modulated audio signal is provided in a stage preceding the digital filter. Item 7. The formant adding device according to item 1. 6. An effect balance circuit for synthesizing an audio signal obtained through the digital filter and the modulated audio signal at a desired ratio and outputting the synthesized signal. Item 6. The formant adding device according to any one of items 5. 7. The formant adding apparatus according to claim 1, wherein said digital filter is an FIR filter.