JPH05100669A - Electronic musical instrument - Google Patents

Abstract

PURPOSE:To obtain the electronic musical instrument which has no peak overlap of frequency characteristics even when plural musical sounds differing in timbre are generated at the same time. CONSTITUTION:Musical sound signals WAVE-A and WAVE-B of different timbre are separated by bands through band-pass filters 201-204 and 209-212 and level detectors 205-208 and 213-216 detect their peak levels. The peak levels are added, band by band, by adders 217-224 and compressibility calculation parts 221-224 generate compressibility signals R1-R4 corresponding to the addition values. Then the musical sound signals WAVE-A and WAVE-B are attenuated, band by band, by filter systems 225 and 226 corresponding to the compressibility signals R1-R4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic musical instrument, and more particularly to an electronic musical instrument capable of preventing a particular frequency component from being emphasized when a plurality of musical tones are mixed and sounded.

[0002]

2. Description of the Related Art Recent electronic musical instruments are capable of producing not only a plurality of musical tones at the same time based on a player's instruction but also a plurality of tone signals by adding a plurality of types of musical tone signals. This kind of technology is disclosed in, for example, Japanese Patent Laid-Open No. 2-173.
698 and the like.

[0003]

Musical tones used in electronic musical instruments and the like generally contain many overtones in order to obtain high-quality sound. If these musical tones are simply added as seen in the above publication, if there is a band in which the peaks of the respective frequency characteristics overlap, only that band may be overemphasized for the sense of hearing, resulting in an uncomfortable musical tone. .. The present invention has been made in view of the above-mentioned circumstances, and even in the case of simultaneously producing musical tones of a plurality of tones,
An object of the present invention is to provide an electronic musical instrument that can prevent emphasis of a specific frequency component.

[0004]

SUMMARY OF THE INVENTION In order to solve the above problems, the present invention provides a musical tone signal of a plurality of tone colors and, in an electronic musical instrument that synthesizes these musical tone signals, sets the overlapping frequencies of the respective musical tone signals. The present invention is characterized by comprising a superposition frequency detecting means for detecting and attenuating means for attenuating each musical tone signal before synthesis or the musical tone signal after synthesis with respect to the frequency detected by the superposition frequency detecting means.

[0005]

The overlapping frequency detecting means detects the overlapping frequencies of the respective timbres, and the overlapping frequency portions are attenuated according to the detection result. Therefore, it is possible to prevent the specific frequency component from being emphasized in the final synthesized tone signal.

[0006]

Embodiments of the present invention will be described below with reference to the drawings. First, the basic operation of the present invention will be described. 10A and 10B show amplitude frequency characteristics of the musical tones A and B to be synthesized, respectively. In the conventional electronic musical instrument, these musical tones A and B are simply added. For example, in the digital domain, sample values are simply added. When the musical tones A and B are added in this manner, the addition is also performed in the frequency domain in the same manner as in the time domain of the waveform. Strictly speaking, what kind of addition is to be performed must take into consideration the phase characteristics of the waveform. That is, even if there is a peak on the graph of the amplitude frequency characteristic, if the phases are shifted by 180 degrees, the amplitudes will weaken each other. However, since the musical tone itself is not stationary and the basic pitch changes significantly with time, when adding peaks on each amplitude frequency characteristic, the emphasis of the frequency in that band is stochastically compensated and almost simple addition is possible. It can be thought to be done.

As an example, FIG. 10C shows the result of addition of the musical tones A and B in the amplitude frequency region. What is shown by a broken line is a curve in which both amplitude frequency characteristics are simply added. Considering phase information, the amplitude frequency characteristic does not always look like the broken line in the figure, but it can be expected that the entire shape is stochastically preserved as described above. From the characteristic of the broken line shown in FIG. 10C, it can be seen that the frequency characteristic is emphasized at the peak on the right side of the amplitude frequency characteristic. In the present invention, this peak is suppressed like the curve shown by the solid line below so that a specific frequency band is not unnecessarily conspicuous.

Next, examples will be described. Each of the embodiments described below is an example in which the present invention is applied to an electronic musical instrument that synthesizes and outputs a plurality of sounds for one key depression. FIG. 1 is a block diagram showing the overall configuration of an electronic musical instrument which is a first embodiment of the present invention. The keyboard 1 is composed of a number of keys required for performance, and outputs information such as a key code KC indicating a pitch and an initial touch IT indicating a key operation speed to the bus line 10 via the keyboard interface 1a.
The CPU 2 controls the operation of the entire musical instrument based on the program information and data information stored in the ROM 3.
Further, various data and the like in this case are stored in the RAM 4. The operation panel 5 detects operations by the performer such as tone color switching, and outputs various information to the panel interface 5a,
It is given to the CPU 2 via the bus line 10. The tone synthesis section 6 is divided into two tone generator circuits, a tone synthesis circuit A and a tone synthesis circuit B, and synthesizes tones of the tone colors designated by the operation panel 5, respectively. Each tone synthesis circuit A, B
Is configured to be capable of simultaneously producing 16 tones by time division processing. The mixing control circuit 7 synthesizes different types of musical sounds given by the musical sound synthesizer 6, and controls the mixing ratio of each frequency band so that only a specific frequency band is not overemphasized. (Details will be described later). The sound system 8 accumulates the musical tones output from the mixing control circuit 7 to release the time-division state, converts it into an analog signal through a D / A converter, amplifies it, and outputs it from a speaker 9.

FIG. 2 is an internal block diagram of the mixing control circuit 7 in FIG. The outputs of the tone synthesis circuit A and tone synthesis circuit B are given to the mixing control circuit 7 as WAVE_A and WAVE_B, respectively. The waveform signal WAVE_A obtained from the tone synthesis circuit A is directly applied to the filter system 225 and also applied to a series of bandpass filter groups 201 to 204 having different bands. The waveform signal WAVE_B given by the tone synthesis circuit B is given directly to the filter system 226 and also to the bandpass filter groups 209 to 212. Symbols BPF1 to BPF4 shown in these drawings show characteristics, and those having the same symbols have the same characteristics. For example, the reference numerals 201 and 209 are both provided with the symbol BPF1 and therefore have the same frequency characteristics. Here, FIG. 11 shows an outline of the filter characteristics of BPF1 to BPF4. Reference numerals 205 to 2 shown in FIG.
Reference numeral 08 denotes a level detector that outputs the peak level of the signal that has passed through the bandpass filters 201 to 204, that is, the maximum value of the amplitude frequency characteristic after passing through the filter. Reference numerals 213 to 216 are also level detectors similar to the above, and output the maximum values of the amplitude frequency characteristics of the signals that have passed through the bandpass filters 209 to 212, respectively. 217
˜220 are adders, which add the respective peak levels in each band of the series A and B and give them to the compression ratio calculation units 221 to 224, respectively. Compression rate calculation unit 221
To 224 form compression rate signals R1 to R4 that indicate how much the signals in the corresponding bands should be compressed and output according to the outputs of the adders 217 to 220, and filter systems 225 and 225 of each series. It is a circuit to be given to 226.

Here, the compression rate calculators 221 to 224 will be described. These have the same circuit configuration as shown in FIG. The nonlinear circuit 301 shown in FIG. 3 has input / output characteristics (horizontal axis shows input signal, vertical axis shows output signal magnitude) as shown in the block. That is, when the input is a small value, the value is output as it is, and as the input is increased, the compressed value is output. In order to realize this characteristic, a predetermined table may be stored in the ROM and read out, or may be realized by calculation. Adder 3
Reference numeral 02 denotes the output signal of the non-linear circuit 301, that is, the deviation between the non-linear characteristic-added input signal x ′ and the original input value x, and detects the amount to be compressed (x−x ′). When the input value is small, the compression amount is naturally 0, and when the input value is large, the compression amount is large. That is, the compression amount is calculated according to the magnitude of the amplitude frequency characteristic in the input signal. Next, the non-linear circuit 303 has the input / output characteristics shown in the block (the horizontal axis represents the input signal and the vertical axis represents the magnitude of the output signal). That is, the non-linear circuit 303 outputs the inverse number (1 / x) of the input signal value x. By multiplying this value by the above-mentioned compression amount in the multiplier 304, the compression ratio (x−x ′) / x for the input signal is obtained. The compression rate calculation units 221 to 224 output the obtained compression rates as compression rate signals R1 to R4, respectively.

Next, the filter system 225 shown in FIG.
And 226 perform filter processing for each band based on the compression ratio signals R1 to R4 corresponding to each band output from each compression ratio calculation unit 221 to 224. These filter systems 225 and 226 have the same circuit configuration, and the configuration is shown in FIG. In the figure, a to d are band pass filters, respectively, and the symbols BPF1 to BPF1 in the block are indicated.
BPF4 shows each characteristic. Further, those having the same symbols as the bandpass filters 201 to 204 and 209 to 212 shown in FIG. 2 have the same characteristics. That is, those corresponding to the same band are configured to have the same characteristics. The band-pass filters a to d are configured such that the amount of attenuation in their frequency range is controlled by the compression rate signals R1 to R4, respectively. Therefore, the waveform signals input to the filter systems 225 and 226 are compressed ratio signals R1 to R4 when passing through the bandpass filters a to d.
The amount of attenuation is controlled by each value of. Then, the adder 227 shown in FIG. 2 adds the signals filtered for each band, and the musical tone signal WA whose mixing ratio is controlled for each band.
Output VE.

Next, the operation of this embodiment having the above-mentioned structure will be described. First, the musical tones of the two tones generated in the musical tone synthesizing circuits 6a and 6b are divided into respective bands by the bandpass filters 201 to 204 and 209 to 212, respectively, and the peak levels are determined by the level detectors 205 to 208 and 213 to 216. To be detected. The peak levels are added by the adders 217 to 224 for each of the two timbres, and the compression rate signals R1 to R4 corresponding to the added values are added to the compression rate calculation unit 221.
~ 224. Then, the tone synthesis circuit 6
tone signals WAVE_A and WAVE generated by a and 6b
_B undergoes band-by-band attenuation in the filter systems 225 and 226 according to the compression ratio signals R1 to R4. That is, it is attenuated according to the value of the amplitude frequency characteristic for each frequency band. Then, since the tone signals after the attenuation control are added in the adder 227, the final output WA
The VE is a signal in which the peak of the frequency characteristic does not come only in the specific band.

In the present embodiment, in order to easily confirm the configuration, the tone synthesis circuit 6 for generating different tones.
Although a and the tone synthesis circuit 6b are physically shown separately, they may be realized by time division multiplexing processing in the same circuit, and since there are many other common circuits, they are time division multiplexed. It goes without saying that you can do it. Further, when there is a circuit that can perform division, or when this circuit is realized by software, the nonlinear circuit 303 and the multiplication circuit 304 shown in FIG. 3 may be realized by one division circuit. Further, the tone synthesis method of each tone synthesis circuit in the tone synthesis section is the same as the existing waveform memory read synthesis method,
Any method such as frequency modulation combining method may be used,
Further, they may be mixed and used. Furthermore, the first
In the embodiment, the number of divisions of the band is set to 4, but it may of course be more than that, and depending on the pitch of the musical tone to be output,
The center frequency of the pass band of the BPF may be changed. In addition, in FIG. 2, the adder 227 is performed after each waveform is analyzed.
However, similar processing can be performed after the combination. In this case, the above-mentioned phase information can be considered and combined. Further, the number of additions of the musical tones is not limited to two, and may be more than two, and reverb for giving a reverberation effect to the final musical tone may be applied.

Next, a second embodiment of the present invention will be described. The second embodiment is almost the same as the first embodiment in terms of hardware, but the configuration and processing method of some circuits are different in the following points. In the first embodiment, the mixing of musical tones and the overlapping of peaks are controlled in real time, which requires high-speed calculation. Therefore, in the second embodiment, the filter coefficient is obtained in advance from the musical sound to be synthesized, and the filter coefficient is given as a fixed value. In this case, since there is a time margin in the calculation, highly accurate calculation can be performed. In the case of the second embodiment, the waveform memory reading method is used as the tone synthesis circuit. Here, the waveform memory reading method will be described. FIG. 9 is an operation explanatory diagram of the waveform memory reading method used in the second embodiment. FIG. 9A shows a waveform reading situation, in which the attack waveform is read in response to the KON signal and thereafter the loop waveform is repeatedly read. FIG. 9B shows the output of the envelope at that time. In many cases, the waveform of the attack portion suddenly changes from a very small amplitude to a large amplitude. Therefore, the envelope is not controlled, and the attack waveform itself stores the envelope-added waveform data. In the subsequent loop waveform reading section, the envelope data smoothly decreases with time, and the degree of decrease changes according to the KOFF signal. It goes without saying that the absolute magnitudes of these envelope signals are scaled according to the key pressing speed, the key code, and the like. As described above, in the waveform memory reading method, generally, different waveforms are prepared for the rising portion of the musical tone that exhibits non-stationary behavior and the loop portion of the musical tone that outputs the steady waveform, and the key depression is performed. In many cases, the waveform of the rising portion is output according to the above, and then the waveform of the loop portion is repeatedly read. In this case, since it is possible to predict what kind of frequency characteristic a musical tone will actually be obtained before synthesizing the musical tone, it is ideal as a target for implementing the present invention. Further, it is generally considered that the loop portion continuously loops not only at the waveform level but also at the phase level, which is advantageous when analyzing the frequency characteristics. The tone synthesis section shown in FIG. 1 shows an example in which two tone synthesis circuits for producing different sounds are physically provided in parallel.
In the second embodiment, "single note" and "duplex" can be switched by software processing. In this way, when the "single note" and "overtone" are switchable, the musical tone synthesizer itself is realized by a time division multiplex system. In the case of single tone pronunciation, the musical tones are synthesized on all the channels, and In this case, a duo is performed by dividing all channels into two. In the case of this method, it is possible to construct an electronic musical instrument having no sound generation channel. Further, the mixing control circuit 7 in FIG. 1 is as shown in FIG. 8 unlike the first embodiment. That is, the single filter system 801 filters the waveforms added in the adder circuit e, and the filter characteristic is controlled by the coefficient k and the key code KC. Furthermore, the filter system 801 of FIG.
Since it is a filter for suppressing the overlap of peaks when two waveforms are added, it becomes unnecessary in the case of a single tone, but in the case of this embodiment, the filter system 801 is not wasted even in the case of a single tone. Therefore, it is used as a filter for tone color adjustment or effect addition.

Next, the operation of this embodiment will be described. FIG. 5 is a flowchart of the main program executed by the CPU 2. First, step S100
In, the initial setting of each register and the like used in the subsequent programs and the initial setting of the peripheral circuits such as the musical tone synthesizer are performed. Next, in step S102, the keys are scanned based on the information from the keyboard interface 1a. If a key event (change in key operation) is detected, it is determined as "YES" in step S104, and the process proceeds to step S106. In step S104,
If no event is detected, step S11
Go to 4. In step S106, an information key event KEV indicating a key event, an information key code KC designating a pitch, and an information initial touch IT designating a key pressing speed are set based on the detected information. In the following description, when the name of a register is described, it means that register and, if necessary, indicates the contents of the register. Next, in step S108, the set information key event KEV is K
If it is ON, the process proceeds to KON processing in step S110, and if it is KOFF, KO in step S112.
Proceed to FF processing. In the KON processing in step S110,
Each value is output to the musical sound synthesizer according to the detected information.
In the present embodiment, since the operation at the time of sound generation is divided into "single-tone sound generation" and "overtone sound generation", processing is performed according to these states inside the KON processing. Here, the key code KC is output to the filter system 801. In the KOFF processing of step S112, the KOFF signal is output to the musical sound synthesizing unit to carry out the processing of ending the generation of the corresponding musical sound. that's all,
Through the processing from step S102 to step S112, the processing from the generation of the musical sound to the mute of the musical sound is performed. In step S114, the panel is scanned according to the information from the panel interface 5a. If there is a panel event, it is determined as "YES" in step S116, and the process proceeds to step S118. Step S118
Then, it is determined whether the event detected in step S114 is a timbre-related event. The judgment result is "YE
When it is "S", the process proceeds to the tone color setting process of step S120, and when the determination result is "NO", the process proceeds to step S122 and other processes are performed. Step S12
In the other processing of 2, for example, control of other devices attached to the electronic musical instrument, such as ON / OFF of reverb for giving a reverberation effect to a musical sound, is performed. As described above, the CPU 2 repeats the processing from step S102 to step S122 as long as the power is ON, and synthesizes a musical sound. FIG. 6 shows details of the tone color setting process in step S120 in FIG. In step S200, the processing is branched depending on whether or not a duo is currently selected on the panel.
In the case of a single note, the process proceeds to step S202 and step S1.
The first tone color information TC1 is set according to the event detected in 14. After that, it advances to step S204,
The filter coefficient k corresponding to TC1 is set to the filter system 80.
Set to 1.

On the other hand, when the duo is currently specified,
Proceeding from step S200 to step S206, the playing process is performed. Here, FIG. 7 shows details of the duo process. First, in step S300, a first tone color TC1 and a second tone color TC2 are set according to the event detected in step S114. In step S302, a predetermined sample value is taken from the beginning of the loop waveform in the midtone range of TC1 and is further multiplied by the window function. The predetermined sample value may be the number of sample points required for the subsequent FFT (Fast Fourier Transform) operation, and basically, the larger the sample value, the higher the frequency resolution. This is because the multiplication of the window function does not necessarily match the loop waveform with the number of sample points required by the FFT, and therefore the number of sample points corresponding to the value is cut out. In step S304, the waveform multiplied by the window function is analyzed by FFT. Steps S306 and S308 are steps S30 except that the object of analysis is the loop waveform of TC2.
2 and S304. In step S310, the frequencies at which the peaks overlap are obtained from the two FFT results obtained above. Here, of course, only the frequencies where the peaks largely overlap are detected. In step S312, the basic filter coefficients of the filter having a trough at the detected frequency are given to the filter system 801 of FIG. Here, the characteristic of the valley of the filter is determined in step S3.
It shall correspond to the height of the peak to be removed detected in 10. Here, the basic filter coefficient k is given because the sound source of the waveform memory reading system has a moving formant, so that when the pitch changes, the above filter characteristics must be changed accordingly. Is. Regarding this point, as shown in FIG. 8, a key code KC is given to the filter system 801, and the filter characteristic control according to the pitch is performed according to this key code KC. After performing these processes, the process returns to the tone color setting process of FIG. FIG. 8 is an internal block diagram of the mixing control circuit in FIG. The tone signal WAVE_A corresponding to the first tone color information TC1 and the tone signal WAVE_B corresponding to the second tone color information TC2 provided by the tone synthesizer 6 are added by the adder and provided to the filter system 801. Since the filter coefficient for attenuating the unnecessary peak is set in the filter system 801, the tone signal WAVE from which the unnecessary band emphasis is removed is output from the output of the filter system 801.

In the above embodiment, in order to simplify the processing, the filter coefficient is determined by referring to the loop waveform in the middle frequency range. However, a plurality of filter coefficients are obtained from the loop waveform at a typical position. Then, different filter coefficients may be obtained by interpolating them according to the pitch after the reproduction. Further, in the above-mentioned embodiment, although the processing concerning the time division is not particularly mentioned, the time division operation is carried out by storing each value in the shift register or the like and giving it to each circuit by the time division and accumulating the final musical sound output. It is also possible to let.

Further, when the second embodiment is applied to a tone synthesis system other than the waveform memory reading system, a plurality of internally corresponding tone sounds are generated and synthesized together with the tone color selection.
It is also possible to detect the degree of overlap of peaks to obtain the filter coefficient and then start the sounding operation. in this case,
Since it takes the same time as actually generating a musical tone to obtain the filter coefficient, the start of the sounding operation is delayed, so while internally generating and synthesizing a musical tone,
A synthetic musical tone may be obtained by raising the operation clock of the waveform.

[0019]

As described above, according to the present invention, when a plurality of musical tones are mixed, the musical tones can be synthesized without overemphasizing only a specific band, and a plurality of types of musical tones are often synthesized. In the recent electronic musical instruments, there is an effect that a high quality ensemble sound can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a first exemplary embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a mixing control circuit 7 in the embodiment.

FIG. 3 is a diagram showing compression rate calculation units 221 to 22 according to the embodiment.
4 is a block diagram showing the configuration of FIG.

FIG. 4 is a filter system 225 according to the same embodiment.
It is a block diagram which shows the structure of 226.

FIG. 5 is a flowchart showing the operation of the second embodiment of the present invention.

FIG. 6 is a flowchart showing the operation of the embodiment.

FIG. 7 is a flowchart showing the operation of the embodiment.

FIG. 8 is a block diagram showing a configuration of a mixing control circuit in the example.

FIG. 9 is an explanatory diagram for explaining a read operation of the waveform memory read method used in the embodiment.

FIG. 10 is a frequency characteristic diagram showing the synthesis of musical tone signals.

FIG. 11 is a characteristic diagram showing characteristics of the bandpass filter used in each of the above-described embodiments.

[Explanation of symbols]

201-204, 209-212 ... Band pass filter (stacking frequency detecting means), 205-208, 213-
216 ... Level detector (polymerization frequency detection means),
217-220 ... Adder (polymerization frequency detecting means), 2
21 to 224 ... Compressibility calculating section (overlap frequency detecting means), 225, 226 ... Filter system (attenuating means).

Claims (1)

[Claims] 1. An electronic musical instrument for generating tone signals of a plurality of tones and for synthesizing these tone signals, a superposition frequency detecting means for detecting overlapping frequencies of the respective tone signals, and a superposing frequency detecting means for detecting the superposing frequency. An electronic musical instrument, comprising: attenuating means for attenuating each musical tone signal before synthesis or the musical tone signal after synthesis for the reproduced frequency.