JPH10214089A - Musical sound signal generator - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain timbre variation similar to that of a natural musical instrument in accordance with a touch even thought there exist a small amount of waveform data by adding envelopes to aperiodic signals and filtering the signals by the filter characteristics that vary in accordance with the touch. SOLUTION: A first multiplier 52 multiplies the periodic sound wave data from a first DCO 50 by the first envelope signals from an envelope generator 60. Then, the first envelope is added to the periodic sound wave data and the result of the multiplication is supplied to a first DCF 54. A second multiplier 53 multiplies the aperiodic sound wave data from a second DCO 51 by the second envelope signals from an envelope generator 60. Thus, the second envelope is added to the aperiodic sound wave data. The result of the multiplication is supplied to a second DCF 55. Then, an adder 56 adds the digital signals from the DCF 54 to the digital signals from the DCF 55.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tone signal generator, and more particularly to a technique for generating a tone signal for generating a tone similar to a natural musical instrument.

[0002]

2. Description of the Related Art Conventionally, in an acoustic piano, which is one of natural musical instruments, it is known that a tone color changes when a keystroke strength (hereinafter, referred to as "touch") changes. In order to simulate the characteristics of an acoustic piano with an electronic musical instrument, a plurality of waveform data corresponding to a plurality of musical tones generated by different touches are stored in advance, and when a key is pressed, the touch of the key is performed. In this case, one waveform data may be read out from the plurality of waveform data and reproduced. However, in this method, since waveform data corresponding to a plurality of touches must be stored in advance, there is a problem that the amount of waveform data is enormous.

[0005] Therefore, several methods have been developed for realizing a tone change in accordance with a touch despite a small amount of waveform data. The first method is a method of simulating a tone change using a filter. In other words, paying attention to the characteristic of natural musical instrument sounds, in which a high-strength sound contains a large amount of high-frequency components but a low-strength sound does not, only waveform data of a strong-strike sound containing a large amount of high-frequency components is prepared. The signal generated based on the waveform data is passed through a digital filter that lowers the signal level by several tens of decibels [dB / oct] for example, in one octave. As a result, the high-frequency component included in the strong tapping sound is removed at a cutoff frequency corresponding to the touch, and a tone color corresponding to the touch such as a medium hitting sound, a light hitting sound, or the like is created in a pseudo manner.

The second method is a method of simulating a tone change by, for example, separating a heavy blow into a plurality of characteristic components and adding them at a ratio corresponding to a touch. For example, acoustic piano sounds include periodic sounds generated by vibrations of strings and sound boards, and non-periodic sounds generated by vibrations other than strings, for example, so-called squeak sounds generated by impact at the time of keying. included. These mixing ratios change with touch, and this contributes to a change in timbre. Therefore, one musical tone is separated into a periodic sound and an aperiodic sound to generate respective waveform data, and when a key is pressed, both of these waveform data are added at a ratio according to the touch of the key, This achieves a tone change.

[0005]

SUMMARY OF THE INVENTION The present inventor has studied the relationship between touch and frequency characteristics in an acoustic piano. The results are shown in FIGS. 3A and 3B are diagrams showing the frequency characteristics at the time of a strong hit, wherein FIG. 3A shows the frequency characteristics of the original sound, FIG. 3B shows the frequency characteristics of the periodic sound component included in the original sound, and FIG. The frequency characteristics of the non-periodic sound components included in the original sound are shown. 4A and 4B are diagrams showing the frequency characteristics at the time of a light hit, wherein FIG. 4A shows the frequency characteristics of the original sound, and FIG. 4B shows the frequency characteristics of the periodic sound component contained in the original sound.
FIG. 4C shows the frequency characteristics of the aperiodic sound components included in the original sound.

Here, comparing the frequency characteristic of the periodic sound component at the time of heavy hitting shown in FIG. 3B with the frequency characteristic of the periodic sound component at the time of light hitting shown in FIG. At the time, a level drop of 10 dB is observed around 3 kHz (sixth harmonic) as compared with the time of hitting hard. On the other hand, comparing the frequency characteristic of the non-periodic sound component at the time of heavy hit shown in FIG. 3C with the frequency characteristic of the non-periodic sound component at the time of light hit shown in FIG. At the time, the level is reduced by 20 dB as compared with the time of hitting hard. This means that the change of the frequency characteristic according to the touch in the acoustic piano differs between the periodic component and the aperiodic component.

In the first method, since the frequency characteristics of the periodic component and the non-periodic component change in the same manner, there is a problem that the timbre change differs from the actual timbre change. In the second method, since the frequency characteristics of the periodic component and the aperiodic component do not change, there is a problem that the timbre change differs from the actual timbre change.

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and is capable of obtaining a tone change similar to that of a natural musical instrument in response to a touch, despite a small amount of waveform data. It is an object to provide a signal generation device.

[0009]

In order to achieve the above-mentioned object, a musical sound signal generating apparatus according to the present invention is adapted to generate periodic sound waveform data corresponding to a periodic sound component included in one musical tone and a non-periodic sound component. Waveform memory storing non-periodic sound waveform data, first signal generating means for generating a periodic sound signal based on the periodic sound waveform data from the waveform memory, and periodic sound signal from the first signal generating means A first envelope adding means for adding a first envelope; a first filter means for filtering a signal from the first envelope adding means with a first filter characteristic that changes according to a touch; Second signal generation means for generating an aperiodic sound signal based on the aperiodic sound waveform data from the waveform memory;
A second envelope adding means for adding a second envelope to the aperiodic sound signal from the second signal generating means;
A second filter unit for filtering a signal from the second envelope adding unit with a second filter characteristic that changes according to a touch; a signal from the first filter unit; and a second filter unit. And a mixing means for mixing the signals from the first and second signals to generate a musical sound signal.

In the present invention, "one tone" means a tone of one tone and one pitch. The waveform memory can be configured by, for example, a read-only memory (hereinafter, referred to as “ROM”) or a random access memory (hereinafter, referred to as “RAM”). The periodic sound waveform data and the non-periodic sound waveform data stored in the waveform memory separate a musical tone signal corresponding to one musical tone into a musical tone signal of a periodic tone component and a musical tone signal of an aperiodic tone component. Can be obtained. In this case, it is desirable that the one musical tone is a musical tone obtained by hitting hard, but the present invention is not limited to this.

The first and second signal generating means can be constituted by, for example, a digitally controlled oscillator (DCO). Further, the first and second filter means can be constituted by, for example, a digital control filter (DCF). Further, the first and second envelope adding means can be constituted by, for example, a digital multiplier. Further, the mixing means can be constituted by a digital adder.
These DCO, DCF, digital multiplier and digital adder are, for example, a digital signal processor (DSP).
Can be configured.

According to the present invention, the periodic sound signal to which the first envelope is added is filtered by the first filter means having the first filter characteristic, and the non-periodic sound signal to which the second envelope is added is filtered by the second filter. Filtering is performed by the second filter means having the above filter characteristic, thereby simulating a change in timbre of an acoustic piano.
In this case, each of the first and second filter characteristics is configured to change according to each touch.

For example, when the touch is strong, the first filter characteristic has a frequency characteristic similar to that of the original sound, that is, without a level decrease, as shown in FIG. As shown in FIG. 3 (C), it is configured to have a frequency characteristic similar to the original sound, that is, without a level decrease. On the other hand, when the touch is weak, the first filter characteristic has a frequency characteristic in which the level drops by about 10 dB near 3 kHz (sixth harmonic) as shown in FIG. 4B, and the second filter characteristic has , 3 kHz as shown in FIG.
It is configured to have a frequency characteristic in which the level drops by about 20 dB near z. In the above description, examples of the filter characteristics in two cases, that is, when the touch is strong and when the touch is weak, have been described. However, the first and second filter characteristics indicate a plurality of frequencies corresponding to each of a plurality of touch sizes. It can be configured to have characteristics.

With such a configuration, for example, a musical tone signal having a frequency characteristic substantially equivalent to the actually measured frequency characteristic shown in FIG. 3 can be obtained when hitting strongly, and shown in FIG. 4 when hitting lightly. A tone signal having substantially the same frequency characteristics as the actually measured frequency characteristics can be obtained. That is, it is possible to generate a tone signal in which the timbre changes in the same manner as a natural musical instrument according to a touch. Moreover, it is not necessary to store the waveform data for each touch size, and only the waveform data of the periodic sound component and the waveform data of the non-periodic sound component need to be stored. it can.

Further, according to the present invention, there is provided a musical sound signal generating apparatus which includes a microphone for picking up one musical sound and a signal for separating a signal from the microphone into a signal corresponding to a periodic sound component and a signal corresponding to an aperiodic sound component. Means for generating periodic sound waveform data based on a signal corresponding to the periodic sound component, and generating non-periodic sound waveform data based on a signal corresponding to the non-periodic sound component, respectively. It is possible to store the periodic sound waveform data and the non-periodic sound waveform data in the waveform memory.

The microphone converts one musical sound, for example, a sound generated by tapping a natural musical instrument such as an acoustic piano, into an electric signal and sends it to the separating means. The separating means separates the electric signal into a tone signal having a periodic component and a tone signal having an aperiodic component. For this separation, for example, a method of extracting only a periodic sound component included in a signal from a microphone using a comb filter, a method of removing a periodic sound component included in a signal from a microphone using an autocorrelation function, and the like are used. it can. Note that, in a simplified manner, a method of using a low-pass filter or a high-pass filter to separate a specific frequency or higher of a signal from a microphone as a periodic sound component and the other as a non-periodic sound component can be used.

The periodic sound waveform data and the non-periodic sound waveform data are obtained by converting the tone signals of the periodic sound component and the non-periodic sound component thus separated, for example, into pulse code modulation (PCM).
Can be created by doing In addition, the waveform data is not limited to PCM, and the differential pulse code modulation (DPC
M), adaptive predictive coding (ADPCM) and other methods. The periodic sound waveform data and the non-periodic sound waveform data created by the separating means are transferred to and stored in the waveform memory.

The present invention is configured to generate a musical tone by dividing a musical tone into a periodic tone component and an aperiodic tone component and filtering each tone with different filter characteristics. It is also possible to further divide the periodic sound component. For example, when generating the tone color of an acoustic piano, it is possible to use a sound generated by a string itself and a sound generated by a soundboard as a periodic sound component, and a so-called knuckle sound as a non-periodic sound. In this case, the waveform memory stores the periodic sound waveform data of the string sound, the periodic sound wave data of the sound of the soundboard, and the non-periodic sound wave data of the knuckle sound. The signal generating means, the envelope adding means, and the filter means are provided in three systems corresponding to each waveform data, and the signals from each system are mixed to generate a musical tone signal. Similarly, when a timbre of a trumpet is generated, for example, a resonance sound of a horn may be used as a periodic sound, and a breath sound and a lip vibration sound may be used as non-periodic sounds. In this case, the waveform memory stores the periodic sound waveform data of the horn sound, the non-periodic sound waveform data of the breath sound, and the non-periodic sound waveform data of the lip vibration sound. The signal generating means, the envelope adding means, and the filter means are provided in three systems corresponding to each waveform data, and the signals from each system are mixed to generate a musical tone signal. According to these configurations, a sound closer to a natural musical instrument can be generated.

[0019]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of a tone signal generating apparatus according to the present invention. In this embodiment, an example in which an acoustic piano sound is divided into a periodic sound and an aperiodic sound, signal processing is performed for each sound, and the sound is finally mixed will be described.

First, in order to facilitate understanding of the present invention,
An overall configuration of an electronic musical instrument in which a musical sound signal generation device according to an embodiment of the present invention is incorporated will be described. FIG. 2 is a block diagram showing the configuration of the electronic musical instrument.

In this electronic musical instrument, a central processing unit (hereinafter, referred to as a central processing unit)
A “CPU” 10, a program memory 11, a work memory 12, a touch detection circuit 13, a tone signal generator 14, and a frequency component separator 15 are interconnected by a system bus 30. The system bus 30 is used for transmitting and receiving, for example, address signals, data signals, control signals, and the like.

The CPU 10 has an operation panel 16,
The pedal 17 and the MIDI interface circuit 18 are connected. The touch detection circuit 13 includes the keyboard device 1.
9 is connected. Further, a D / A converter 20 is connected to the tone signal generation device 14, and a sound system 21 is connected to the D / A converter 20. Further, an A / D converter 22 is connected to the frequency component separation device 15,
A microphone 23 is connected to the A / D converter 22.

The CPU 10 controls the entire electronic musical instrument according to a control program stored in the program memory 11. The processing performed by the CPU 10 will be described later in detail.

The program memory 11 is constituted by, for example, a ROM. The program memory 11 stores a tone color parameter group in addition to the control program described above. This timbre parameter group includes a plurality of timbre parameters corresponding to a plurality of tone ranges of a plurality of timbres (instrument sounds). One tone color parameter is used to generate a sound of a predetermined tone range of a predetermined tone color.

Each tone color parameter comprises, for example, a first waveform address, a second waveform address, a first frequency data, a second frequency data, a first envelope data, a second envelope data, a first filter coefficient and a second filter coefficient. Have been.

The first waveform address designates a position in the waveform memory 70 in which the periodic sound waveform data of one tone is stored, and the second waveform address stores the non-periodic sound waveform data of the same tone. The designated position in the waveform memory 70 is designated. The first frequency data specifies the speed at which periodic waveform data is read from the waveform memory 70, and the second frequency data specifies the speed at which non-periodic waveform data is read from the waveform memory 70. The first envelope data specifies the envelope shape of the periodic sound waveform, and the second envelope data specifies the envelope shape of the non-periodic sound waveform. The first filter coefficient specifies a filter characteristic (first characteristic) for periodic sounds, and the second filter coefficient specifies a filter characteristic (second characteristic) for non-periodic sounds.

The timbre parameters read from the program memory 11 are sent to the tone signal generator 14.
The tone signal generator 14 generates a digital tone signal based on the tone color parameters.

The program memory 11 is a RAM
Can be configured. In this case, when the power is turned on, the control program and the tone parameter group stored in, for example, a hard disk, a floppy disk, an IC memory, a CD-ROM, or the like are stored in the program memory 11 (RA
M).

The work memory 12 is constituted by, for example, a RAM. The work memory 12 temporarily stores various data processed by the CPU 10. In the work memory 12, buffers, registers, counters, flags, and the like (not shown) are defined.

The operation panel 16 is provided with, for example, a tone color selection switch, a rhythm selection switch, a sound effect selection switch, a volume control volume, and the like for a person to operate the electronic musical instrument. The timbre selection switch is composed of, for example, a plurality of push button switches, and is used to select the timbre of a musical instrument such as a piano or a harpsichord. The rhythm selection switch is composed of, for example, a plurality of push button switches, and is used to select a rhythm of automatic accompaniment such as eight beats, waltz, and rock. The sound effect selection switch is composed of a plurality of push button switches, for example, and is used to select a sound effect such as reverb, chorus, and the like. The volume control volume is used to control the volume of a musical sound.

The operation panel 16 includes a panel scan circuit (not shown). The panel scan circuit controls data transmission and reception between the operation panel 16 and the CPU 10. That is, the panel scan circuit scans the operation panel 16 and takes in a signal indicating the open / closed state of each switch. Then, based on this signal, panel data in which each switch corresponds to 1 bit is generated and sent to the CPU 10 via the system bus 30.

The CPU 10 performs a process corresponding to a switch operation on the operation panel 16 with reference to panel data obtained from the panel scan circuit. For example, when the tone selection switch is operated, the CPU 10 performs a tone change process.
In this tone color changing process, the tone color number of the tone color corresponding to the operated switch of the tone color selection switches is stored in a predetermined area of the work memory 12. The CPU 10 selects one timbre parameter from the timbre parameter group stored in the program memory 11 by referring to the timbre encryption, and sends the timbre parameter to the tone signal generator 14.

The pedal 17 includes a damper pedal and a soft pedal. The damper pedal is used to prolong the sound release time and generate a rich sound.
The soft pedal is used for suppressing a loud sound to generate a soft sound. The pedal 17 has a pedal scan circuit, like the operation panel 16.
When a damper pedal or a soft pedal is operated, each pedal generates pedal data corresponding to one bit and sends it to the CPU 10. This pedal data is used to modify tone color parameters.

The MIDI interface circuit 18 controls transmission and reception of MIDI messages between the electronic musical instrument and external devices. As an external device, for example, another electronic musical instrument,
A personal computer, a sequencer, or the like can be used. More specifically, the MIDI interface circuit 1
8 receives the MIDI message sent from the external device and sends it to the CPU 10. The CPU 10 uses the MIDI
The sounding / muting process is performed based on the message, and further, the setting state of the operation panel 16 is changed. Further, data generated by operating the operation panel 16 and the keyboard device 19 is converted into a MIDI message,
The data is sent to an external device via the interface circuit 18.
Thus, an external device can be controlled from the operation panel 16 and the keyboard device 19 of the electronic musical instrument.

The keyboard device 19 has a plurality of keys (for example, 88 keys) for instructing pitch and touch. The keyboard device 19 includes two keys so that a touch can be detected.
A keyboard device 19 of a contact type is used. Each key of the keyboard device 19 has two key switches (first and second key switches) that open and close in conjunction with key depression and key release.

The touch detection circuit 13 scans the keyboard device 19 and fetches a signal indicating the open / closed state of the first and second key switches of each key. Then, based on this signal, key data in which each key corresponds to one bit (each bit indicates on / off of the key) is generated and sent to the CPU 10. Further, the touch detection circuit 13 measures the time from when the first key switch is turned on to when the second key switch is turned on, and uses the measurement result as a keying strength (keying speed). The touch data is sent to the CPU 10 as the touch data.

The tone signal generator 14 generates a digital tone signal according to the tone color parameters and the touch data from the CPU 10. The details of the tone signal generator 14 will be described later. The digital tone signal generated by the tone signal generator 14 is supplied to the D / A converter 20.

The D / A converter 20 is used to generate the musical tone signal
4 is converted to an analog tone signal. The output of the D / A converter 20 is supplied to a sound system 21. The sound system 21 includes an amplifier and a speaker or headphones. The sound system 21 amplifies the input analog musical tone signal and sends it to a speaker or headphones. As a result, a musical sound is generated from the sound system 21.

In FIG. 2, the above-mentioned tone signal generator 14, D / A converter 20, and sound system 2 are shown.
Although only one system 1 is shown, two systems can be provided to configure a stereo playback system for both left and right channels, and further, three or more systems can be provided to configure a multi-channel playback system.

The microphone 23 converts a musical instrument sound, for example, a musical sound generated by striking an acoustic piano into an analog electric signal. The analog signal from the microphone 23 is supplied to the A / D converter 22. The A / D converter 22 converts an analog signal into a digital signal and supplies the digital signal to the frequency component separation device 15.

The frequency component separating device 15 separates the digital signal from the A / D converter 22 into a periodic digital signal and an aperiodic digital signal. For this separation, for example, a method of removing a periodic sound component or an aperiodic sound component using a comb filter or a low-pass / high-pass filter, or a method of removing a periodic sound component using an autocorrelation function can be used. The digital signal of the periodic sound and the digital signal of the aperiodic sound separated in this way are each subjected to pulse code modulation (PCM), and further subjected to necessary waveform processing such as looping, resampling, and cutting. Thereby, periodic sound waveform data and aperiodic sound waveform data are generated. The periodic sound waveform data and the non-periodic sound waveform data from the frequency component separation device 15 are supplied to the waveform memory 70.

The function of the frequency component separating device 15 can be realized by the processing of the CPU 10. In this case, the output of the A / D converter 22 is supplied to the CPU 10 via the system bus 30. The CPU 10 uses the A / D
The digital signal from the converter 22 is subjected to the above-described comb filter processing, processing for removing periodic sound components using an autocorrelation function, or low-pass or high-pass filter processing, so that periodic sound waveform data and non-periodic sound wave data are obtained. Is created and stored in the waveform memory 70. According to this configuration, there is an advantage that the amount of hardware can be reduced.

In this embodiment, the waveform memory 7
Numeral 0 is a rewritable memory, in which the periodic waveform data and the non-periodic waveform data generated by the frequency component separating device 15 are directly written. The waveform memory 70 is a ROM. be able to. In this case, the electronic musical instrument need not include the frequency component separating device 15, the A / D converter 22, and the microphone 23. That is, the periodic sound waveform data and the non-periodic sound waveform data created using another device are written in the ROM in advance,
This may be configured to be mounted on the tone signal generation device 14 as the waveform memory 70.

Next, the detailed configuration of the tone signal generating device 14 will be described with reference to the block diagram of FIG.

The tone signal generator 14 is provided with a first DCO 5
0, second DCO 51, first multiplier 52, second multiplier 53, first DCF 54, second DCF 55, adder 56, envelope generator 60, filter coefficient generator 6
1 and a waveform memory 70.

The waveform memory 70 stores periodic sound waveform data and aperiodic sound waveform data corresponding to each tone color parameter. That is, the periodic sound waveform data and the non-periodic sound waveform data created by the above method are stored for each timbre (instrument sound). Further, for a predetermined timbre, the periodic sound waveform data and the non-periodic sound waveform data created by the above method are further stored for each sound range. In this case, for example, in the case of a keyboard instrument, instead of storing the periodic waveform data and the aperiodic waveform data for each key, one periodic waveform data and the aperiodic waveform data are stored in units of a plurality of keys. Is also good. According to this configuration, the amount of waveform data to be stored in the waveform memory 70 can be reduced. The periodic sound waveform data and the non-periodic sound waveform data read from the waveform memory 70 are:
These are sent to the first DCO 50 and the second DCO 51, respectively.

The first DCO 50 generates a digital signal of a periodic sound component. The first DCO 50 includes a CPU 1
0 to the first waveform address and the first
Frequency data is provided. The first DCO 50 receiving these data sequentially reads the periodic sound waveform data from the position specified by the first waveform address in the waveform memory 70 at a speed according to the first frequency data. The read periodic sound waveform data is sequentially supplied to the first multiplier 52.

The second DCO 51 generates an aperiodic digital signal. The second DCO 51 includes a CPU 10
From the second waveform address and the second
Are supplied. The second DCO 51 receiving these data sequentially reads the non-periodic sound waveform data from the position specified by the second waveform address in the waveform memory 70 at a speed according to the second frequency data. The read aperiodic sound waveform data is sequentially supplied to the second multiplier 53.

The envelope generator 60 includes a CPU 10
Supplies the first and second envelope data and the touch data in the timbre parameters. The first and second envelope data define the waveform envelope as described above. The touch data defines the overall level (amplitude) of the envelope shape. The envelope generator 60 generates a first envelope signal for forming a periodic sound and a second envelope signal for forming an aperiodic sound based on the first and second envelope data and the touch data. The first envelope signal is supplied to a first multiplier 52, and the second envelope signal is supplied to a second multiplier.

The first multiplier 52 multiplies the periodic waveform data from the first DCO 50 by the first envelope signal from the envelope generator 60. As a result, the first envelope is added to the periodic sound waveform data. The result of the multiplication is supplied to the first DCF 54. The second multiplier 53 multiplies the aperiodic sound waveform data from the second DCO 51 by the second envelope signal from the envelope generator 60. As a result, the second envelope is added to the aperiodic sound waveform data. The result of this multiplication is
This is supplied to the second DCF 55.

The filter coefficient generator 61 includes a first DCF
A first effective filter coefficient that determines the filter characteristic (first filter characteristic) of the filter 54 and a second effective filter coefficient that determines the filter characteristic (second filter characteristic) of the second DCF 55 are generated. This filter coefficient generator 61 includes:
First and second filter coefficients and touch data in the tone color parameters are supplied from the CPU 10. The first effective filter coefficient generated by the filter coefficient generator 61 is supplied to a first DCF 54, and the second effective filter coefficient is supplied to a second DCF 55.

When the touch data indicating a strong blow is supplied, the first effective filter coefficient is set to the first DCF 54 so as to obtain the frequency characteristic of the original sound without a level decrease as shown in FIG. 3B, for example. When the touch data indicating a weak hit is supplied, the first DCF 54 is controlled so as to obtain a frequency characteristic of a level drop of 10 dB near 3 kHz as shown in FIG. 4B, for example. Filter coefficients for Here, only the frequency characteristics in the case of a strong hit and the case of a light hit are described, but the first effective filter coefficient is generated such that the frequency characteristics are variously changed according to the size of the touch data.

When the touch data indicating a strong blow is supplied, the second effective filter coefficient is set to the second DCF 55 so that the frequency characteristic of the original sound without a level decrease as shown in FIG. 3C is obtained. When the touch data indicating a weak hit is supplied, the second DCF 55 is controlled so that a frequency characteristic of a level drop of 20 dB near 3 kHz as shown in FIG. 4C is obtained. Filter coefficients for In this case as well, the frequency characteristics are variously changed according to the size of the touch data.
Are generated.

The first DCF 54 is, for example, 12 dB / o
It can be constituted by a low-pass filter or a high-pass filter having a gradient of ct. In this case, the first effective filter coefficient may be data defining a cutoff frequency and / or a resonance frequency. The digital signal of the periodic sound filtered by the first DCF 54 is supplied to the adder 56.

The second DCF 55 is, for example, 12 dB / o
It can be constituted by a low-pass filter or a high-pass filter having a gradient of ct. In this case, the second effective filter coefficient can be data defining a cutoff frequency and / or a resonance frequency. The digital signal of the aperiodic sound filtered by the second DCF 525 is supplied to the adder 56.

The adder 56 adds the digital signal from the first DCF 54 and the digital signal from the second DCF 55. The result of the addition is D as a digital tone signal.
/ A converter 20 (see FIG. 2).

Next, the operation of the tone signal generator 14 in the electronic musical instrument having the above-described configuration will be described. The waveform memory 70 stores a plurality of periodic sound waveform data and a non-periodic sound waveform data corresponding to a plurality of tone colors.

When any key of the keyboard device 19 of the electronic musical instrument is hit, the touch detection circuit 13 detects a key number indicating the position of the hit key and touch data indicating the strength of the touch. The key number and touch data detected by the touch detection circuit 13 are sent to the CPU 10.

The CPU 10 recognizes the currently specified timbre by referring to the timbre number stored in a predetermined area of the work memory 12 and retrieves one timbre parameter corresponding to this timbre from the program memory 11. The signal is sent to the tone signal generator 14. As a result, the first DCO 50 in the tone signal generating device 14 converts the periodic sound waveform data from the position in the waveform memory 70 specified by the first waveform address in the tone color parameter at the speed specified by the first frequency data. Read sequentially. On the other hand, the envelope generator 60 sequentially generates the first envelope signal based on the first envelope data in the timbre parameters in parallel with the reading operation. The multiplier 52 is a first DCO 5
The periodic waveform data from 0 is multiplied by the first envelope signal from the envelope generator 60. As a result, the digital signal of the periodic sound with the envelope is sequentially output from the multiplier 52. This digital signal is sent to the first DCF 54.

Similarly, the second DCO 51 in the tone signal generator 14 designates the aperiodic waveform data from the position in the waveform memory 70 specified by the second waveform address in the timbre parameter as the second frequency data. At the specified speed. On the other hand, the envelope generator sequentially generates a second envelope signal based on the second envelope data in the sound management parameters in parallel with the reading operation. The multiplier 53 multiplies the non-periodic waveform data from the second DCO 51 by the second envelope signal from the envelope generator 60. As a result, the multiplier 53 sequentially outputs digital signals of enveloped aperiodic sounds. This digital signal is sent to the second DCF 55.

The filter coefficient generator 61 generates a first effective filter coefficient from the first filter coefficient in the timbre parameters and the touch data, and sends the first effective filter coefficient to the first DCF 54. As a result, the first DCF 54 outputs a digital signal of a periodic sound filtered to have a frequency characteristic corresponding to the touch data, and supplies the digital signal to the adder 56. Similarly, the filter coefficient generator 61 generates a second effective filter coefficient from the second filter coefficient in the timbre parameters and the touch data, and sends the second effective filter coefficient to the second DCF 55. As a result, the second DCF 525 outputs a digital signal of an aperiodic sound filtered to have a frequency characteristic corresponding to the touch data, and supplies the digital signal to the adder 56. Then, the digital signal of the periodic sound and the digital signal of the non-periodic sound are added by the adder 56 and supplied to the D / A converter 20 as a digital tone signal. The digital tone signal is converted into an analog signal by the D / A converter 20 as described above, sent to the sound system 21, and emitted by the sound system 21.

As described above, the tone specified by the tone selection switch is changed and emitted according to the key range and the touch.

Although the present embodiment has been described with reference to the case where a tone signal is generated based on data sent from the keyboard device 19, data sent from outside, for example, from a MIDI device, for instructing sound generation, It is also possible to generate a tone signal similar to the above based on the note-on MIDI message sent or the note-on data sent from the sequencer.

In the above-described embodiment, the case where sound is generated on one channel has been described. However, the same applies to the case where a tone signal is generated in stereo (two channels) and the case where a tone is generated on three or more channels. Applicable. In this case, the waveform data for each channel created in the same manner as described above may be stored in the waveform memory 70.

[0065]

As described in detail above, according to the present invention,
It is possible to provide a tone signal generation device capable of obtaining a tone change similar to that of a natural musical instrument in response to a touch, despite the small amount of waveform data.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a musical sound signal generation device according to the present invention.

FIG. 2 is a block diagram showing a configuration of an embodiment of an electronic musical instrument to which a tone signal generating device of the present invention is applied.

FIG. 3 shows the present invention and a conventional tone signal generation device;
It is a figure showing the relation between the touch at the time of a hard hit, and frequency characteristics.

FIG. 4 is a diagram showing a tone signal generating apparatus according to the present invention and a conventional tone signal generating apparatus;
It is a figure showing the relation between the touch at the time of a weak hit, and frequency characteristics.

[Explanation of symbols]

 Reference Signs List 10 CPU 11 Program memory 12 Work memory 13 Touch detection circuit 14 Music signal generation device 15 Frequency component separation device 16 Operation panel 17 Pedal 18 MIDI interface circuit 19 Keyboard device 20 D / A converter 21 Sound system 22 A / D converter 23 Microphone 50 1st DCO 51 2nd DCO 52 1st multiplier 53 2nd multiplier 54 1st DCF 55 2nd DCF 56 Adder 60 Envelope generator 61 Filter coefficient generator 70 Waveform memory

Claims (3)

[Claims] 1. A waveform memory storing periodic sound waveform data corresponding to a periodic sound component and non-periodic sound waveform data corresponding to an aperiodic sound component included in one musical tone, and a periodic sound waveform from the waveform memory. First signal generating means for generating a periodic sound signal based on data; first envelope adding means for adding a first envelope to the periodic sound signal from the first signal generating means; A first filter for filtering a signal from the envelope adding means with a first filter characteristic that changes in accordance with a touch; and a second for generating an aperiodic sound signal based on the aperiodic sound waveform data from the waveform memory. A second envelope adding means for adding a second envelope to the non-periodic sound signal from the second signal generating means; and Mixing a signal from the first filter and a signal from the second filter, the second filter for filtering the signal with a second filter characteristic that changes according to the touch, and A mixing means for generating a musical sound signal. A microphone for picking up one musical tone; and a separating means for separating a signal from the microphone into a signal corresponding to a periodic sound component and a signal corresponding to an aperiodic sound component. Generates periodic sound waveform data based on a signal corresponding to the periodic sound component, and generates aperiodic sound waveform data based on a signal corresponding to the non-periodic sound component. 2. The tone signal generating device according to claim 1, wherein data is stored in said waveform memory. 3. The tone signal generating apparatus according to claim 1, wherein said one tone is a natural musical instrument sound.