JPS5846035B2 - electronic musical instruments - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is an electronic musical instrument in which the frequency of the generated sound can be specially adjusted.

Theoretically, the frequency of a note one octave higher on a musical scale is said to double, and it is technically easy to step down the frequency by half, especially with electronic musical instruments, so the generated sound increases every time it goes one octave higher. Usually, the frequency is correctly doubled.

However, psychologically, it is said that the frequency of a note that feels one octave higher is slightly higher than twice the frequency, and for this reason, a scale that correctly adjusts a note one octave higher to double or double gives the impression of a full 7. .

This invention does not give such a feeling, and aims to generate a sound with a frequency that psychologically makes the user feel that a sound one octave higher is one octave higher.

In the present invention, for the above-mentioned purpose, a simplified device will be described whose sole purpose is to uniformly adjust high tones to be higher and low tones to be lower to the lowest level.

However, the principle is that individual sounds can be adjusted arbitrarily.

Furthermore, unlike simply adjusting individual sounds as in the embodiments described later, new effects such as a certain sound having a different frequency depending on the volume of other sounds being played at the same time are also possible.

The present invention creates a pulse train that is proportional to the sound period but has an extremely short average interval (about 1/100th of a second) compared to the sound period, descends this pulse train, and counts the input pulses to obtain a fixed number of pulses. In addition to counters that operate periodically, the present invention is applied to electronic musical instruments that convert electrical waveforms associated with the periodic operations of this counter into sound.

Therefore, the frequency of the sound obtained is determined by how to create a pulse train with an average interval proportional to the period of the sound.

Therefore, the present invention describes a method for creating such a pulse train.

First, pulse trains of various average intervals are obtained by removing teeth from or packing a theoretical musical scale reference pulse train (not necessarily a pulse train of equal intervals) according to a certain rule.

For example, if m pulses are omitted for every n reference pulses, the average interval increases to 121, and if m pulses are inserted for every n reference pulses, the average interval decreases to □.

Although the pulse train obtained by n+m by such tooth cutting and filling is irregularly spaced, it is averaged as it is stepped down, and the periodic operation of the counter becomes almost evenly spaced in terms of time.

Since the average intervals obtained in this way form a simple integer ratio to each other, they are suitable for obtaining a pure tonic scale.

To obtain a pure tonic scale, please refer to Patent Application Publication No. 5038327 or Patent Application No. 52-069205 filed on June 10, 2013.
You can use a method such as the electronic musical instrument No.

This proposal is based on the pulse train for a theoretically correct scale, and attempts to obtain a psychologically satisfying scale by making slight changes to it.

Figure 1 shows the principle diagram.

In the figure, 1 is a latch circuit that stores binary numbers, 2 is a rate multiplier, 3 is a binary counter, 4 is a differentiation circuit, and 5
is an OR circuit, So is a changeover switch, Sl, S2. S3 is also a changeover switch.

These switches are electronic switches as described below.

In FIG. 1, the chain line portion excluding the latch circuit 1 is referred to as a telescopic portion as a whole.

This means expanding or contracting the average interval of pulses. Input X is a pulse train whose average interval is proportional to the period of a theoretical musical scale, and is obtained by removing teeth from a reference pulse train according to a certain rule.

Y is a pulse train that does not overlap with X at all, and is used to fill in between the X pulse trains.

An example of an X pulse train and a Y pulse train is shown in FIG.

In the figure, when the reference pulse train is a pulse train in which all teeth are not extracted, the X pulse train is shown as an example in which one pulse is missing out of every four, and the average interval is 100 times that of the standard pulse train. There is.

The Y pulse train in a is made by delaying the X pulse train and creates pulses with a short pulse width, and in b, when creating the reference pulse, two-phase pulse trains that do not overlap are created and one is made into a Y pulse train. , C is a pulse train in which pulses exist only at positions where the teeth of the X pulse are removed.

In either case, the Y pulse does not overlap with the X pulse.

Next, the operation of the principle diagram in FIG. 1 will be explained.

The latch circuit 1 stores a binary number corresponding to the sound to be generated.

The higher the pitch of the consonant, the smaller this binary number becomes, and the lower it becomes, the more it becomes a dog.

In this application example, the number of digits in the launch circuit 1 is 7 bits, and the 6 bits excluding the highest digit are input to the rate multiplier 2.

The 6-digit rate multiplier 2 allows n (n is the input binary number from the launch circuit 1) pulses to pass for every 64 (i.e., 26) X pulse trains shown in the figure. A similar mechanism can be used to block the passage of every 64 n pieces and allow the others to pass, so the rate multiplier 2 in this figure outputs both of these, and these are denoted as Z and U, respectively. .

As can be seen from the way these Z and U pulses are made, the average interval is inversely proportional to the stored binary number or its complement.

That is, Z is a pulse train in which n pulses pass every 64 X pulses, and U is a pulse train in which (64-n) pulses pass.

The output of the highest digit of the latch circuit 1 is set to the selector switch S1. S2
゜S3. When the highest digit is O, everything is switched to the lower side as shown in the figure, and when it is ■, it is switched to the upper side.

3 is a binary count, which in this example has 4 bits, so it returns to the initial state every 16 input pulses.

The differentiating circuit 4 receives the most significant digit of this binary count as input, and when this digit changes from O to 1, only one of the pulse trains is switched S.

You can switch with . This differential circuit 4 and switch S.

FIG. 3 shows an example of the configuration of the portion (within the chain line) consisting of the following.

In the figure, 6 and 7 are J and K flip-flops, and when the pulse applied to the Ct terminal changes from H (high level) to L (low level), the outputs Q and Q are input to J. Let us invert them so that they are equal.

8 is an inverter, 9 is an AND circuit, and these commonly used symbols will not be given their names hereafter.

When the output R of the highest digit of the counter 3 becomes O (this happens rarely compared to the frequency with which the pulse train of 3 is input).

) The JK flip-flop 6 is inverted by the next ■ pulse, and the JK flip-flop 7 is inverted by the next ■ pulse.

This period when 6 is inverted and 7 is not inverted, that is, 1 of V pulse
The output of the AND circuit 9 becomes H for the duration of the pulse interval, the AND circuit 10 blocks the passage of the pulse, and the AND circuit 11
Allow the pulse to pass.

In this way, the chain line S.

The inner part as a whole acts as a selector switch.

(81,S2.Ss is also electronically shown in Figure 3S.

It is configured as follows. ) In FIG. 1, when the sound to be produced is high, the binary number held by the latch circuit 1 is small and the highest digit is 0, and the selector switch S1. S2. Since S3 is tilted downward, the pulse train applied to the counter 3 is a U pulse train, and the smaller the stored binary number of 1, the more pulses pass through and are added to the counter 3.

Every time this counter 3 operates one cycle, the selector switch S2
Only one Y pulse passing through is added to the OR circuit 5 and packed into the X pulse train.

The output pulse train W is the output of the OR circuit 5 and has a shorter average interval than the X pulse train.

S because the U pulse train becomes sparse as the number of binary numbers stored in the latch circuit 1 increases.

The number of times of switching is reduced, and the number of pulses stuffed into the X pulse train is reduced.

Therefore, the average interval of the W pulse train approaches the average interval of the X pulse train.

The number of binary numbers stored in latch circuit 1 increases further, and the best technique is O.
When it becomes 1, it becomes 81. S2. S3 is tilted upward, and the input pulse train to the counter 3 is a Z pulse train, that is, a kneading with a small number of binary numbers stored in the latch circuit 1, and becomes denser as the number of binary numbers increases.

- The pulse train is the X pulse train itself, which is changed over by the changeover switch S every time the counter 3 operates one cycle.

Since only one pulse is blocked from passing by, the output W pulse train becomes the X pulse train with teeth removed, and the average interval increases, and the extent of this increase is so remarkable that the binary number stored in the latch circuit 1 becomes a dog.

In this way, while the number of binary numbers stored in the latch circuit 1 is small, the output W pulse train has a shorter average interval than the X pulse train, and as the number of stored binary numbers increases, the average interval becomes longer than the X pulse train.

When the number of binary numbers stored is the smallest, it is oooooooo, and all of the X pulses become U pulses, and one out of 16 X pulses is packed into a W pulse, so the average interval is I, and as a 7 note. It becomes almost a semitone higher, and 1111111, which has the largest stored binary number, becomes about a semitone lower in the same way.

Based on the above principle, we create a pulse train X with an average interval proportional to the period of the theoretical scale of the pitch of the sound we want to generate, and at the same time store a binary number corresponding to the height in the latch circuit 1. The frequency can be changed slightly by using the circuit shown in FIG.

The extent of the change can be determined by how many input pulses the binary counter 3 in FIG. 1 returns to the original value.

Next, an embodiment for generating this X pulse train and generating a binary number to be stored in the latch circuit 1 will be described.

In the figure, 12 is a scanning counter, 13 is a selection circuit, 14 is a keyboard, 15 is a second selection circuit, 16 is a theoretical pulse train generation circuit, and 1 is a latch circuit, which also serves as the latch circuit in FIG.

17 is the portion previously referred to as the expansion/contraction section excluding the latch circuit 1 in FIG. 1, and S4 is a distribution circuit.

Pulses are constantly applied to the scanning counter 12, which is counted, and the displayed number increases one by one one after another. In this example, since there are 7 bits, 128 input pulses complete one cycle.

It is assumed that the time for one rotation is sufficiently fast compared to the movement of the finger.

On the keyboard 14, a pressed key outputs an H signal while being pressed.

The selection circuit 13 selects and outputs the signal from the keyboard according to the output of the scanning counter, so when the signal from the pressed key is selected, the output becomes H, and this signal is sent to the distribution circuit S4.
It is applied to the latch circuit 1 via , and stores the input that is applied from the upper side of the figure at that time, that is, the output of the scan counter.

The theoretical pulse train generation circuit 16 simultaneously generates a pulse train in parallel with an average period proportional to the period of each note of the theoretical scale.
The second selection circuit 15 selects and outputs this pulse train according to the output of the latch circuit 1.

As described above, the pulse train corresponding to the pressed key on the keyboard 14 is selected and completed.

This pulse train is designated as X, and the latch circuit 1 for selecting this pulse train also serves as the latch circuit 1 of FIG. 1, and teeth are packed into the pulse train X or teeth are removed to obtain a W pulse train.

If the configuration is configured so that keys of high to low notes are selected as the number displayed on the scanning counter 12 changes from small to large, the high parts will be somewhat higher and the lower parts will be somewhat lower than the theoretical scale. You can get the sound.

The distribution circuit is for generating multiple sounds at the same time, and as shown in the figure, it is connected to the a terminal, and when it finishes sending pulses to the latch circuit 1, it is transferred to the b terminal, and is connected to another latch circuit ( A selection circuit (not shown) connected to this latch circuit selects a pulse train.

This distribution circuit S4 part is patent application number 48-138.
The technique disclosed in No. 266 may be used, but the details are omitted since it is not related to the present invention.

In the case described above, Table 1 shows an example of the binary numbers stored in the latch circuit 1 and the reciprocal of the ratio of expansion/contraction of the average interval (frequency expansion/contraction ratio).

The frequency expansion/contraction ratio is 1 if the binary number stored in latch circuit 1 is 1. It forms an arithmetic series that changes direction every time it changes.

A geometric series is better in order for the pitch to change uniformly, but if we divide 1 1 + hill and 116 into 128 in a geometric series, the values are shown in the rightmost column. The difference is small.

There, the pitch of the zo-tone may vary uniformly.

When selecting by the output of latch circuit 1, 1 octave is 4
The pulse train generation circuit 16 generates a pulse train for one octa, and the output of the fifth digit or higher from the bottom of the latch circuit 1 may specify the number of downshift stages. As shown in the figure, the W pulse train is passed through the down-down circuit 18.
Descend at

In the figure, 16' is a pulse train generation circuit, and 15' is a selection circuit, which has the same function as in Figure 4, but is different from the selection circuit 1 in Figure 4.
5. The pulse train generation circuit 16 handles a pulse train for one octave of eight octaves, and in FIG. 5, one octave is used for distinction.

In this way, when 4 bits correspond to 1 octave, 2 bits correspond to 1 octave.
Some base numbers do not correspond to sounds, and if we look at the correspondence between sounds and binary numbers, it is possible that even though the sounds are next to each other, the corresponding binary numbers may be a little apart. , the difference is so small that it doesn't affect the ears much.

FIG. 6 shows another embodiment in which the distribution circuit S4 is a scanning counter 2.
It returns to terminal a every time it reaches its final value.

Therefore, the latch circuit 1 stores the binary number corresponding to the highest note among the pressed keys.

Reference numeral 19 denotes a reference pulse train generator which expands or contracts the average interval thereof according to the binary number stored in the latch circuit 1.

The obtained pulse train W is used as a reference pulse train, and from this pulse train generation circuit 16, teeth are removed or teeth are added to create a pulse train with the average interval of each note of the scale.The mutual ratio at this time is correct and theoretical. It is the ratio of the notes of the scale.

Then, from among these pulse trains, the pulse train corresponding to the key pressed is selected and made into sound.

As a result, when two or more tones are played at the same time, the ratio of the periods of those tones forms a theoretical simple integer ratio, and (2) a tone an octave higher has exactly twice the period.

However, the frequency of the sound as a whole becomes higher or lower, and the degree of this will depend on the highest sound among the keys being pressed.

In other words, when the highest note of the pressed key is high, the frequency is higher than the theoretical value, and when it is low, the frequency is lower than the theoretical value, so that the overall frequency does not give a "cluttered feeling", and the frequency is higher than the theoretical value. It can produce multiple sounds that are well harmonized with each other.

Note that this method also applies when producing sounds from multiple cylinders at the same time.
Since only one step is required, this method is more economical than the methods shown in FIGS. 4 and 5.

In this case, the binary number stored in the latch circuit 1 does not depend on the highest note among the notes pressed at the same time;
It is also possible to configure the latch circuit to detect a key pressed most deeply or a specific key among keys pressed at the same time, and store a binary number corresponding to that key.

According to the above method, the higher notes of the keyboard over several octaves can be made higher, and the lower ones can be made lower, and the extent to which this occurs depends on how many times the counter 3 counts before being reset.

In this explanation, the counter 3 is assumed to have 4 bits, but in reality, the number of bits is increased to minimize the deviation from the theoretical scale.

The rate multiplier 2 does not necessarily have to be 6 bits, and may have a smaller number of bits.

At this time, the lower digit of latch circuit 1 is not connected to the rate multiplier.

In this case, the expansion/contraction ratio will be the same for consecutive sounds, but even if the expansion/contraction ratio is changed one by one, the difference will be minute, so even if it is simplified as above, it will not affect the ears. be.

As described above, according to the present invention, it is possible to slightly deviate the scale of an electronic musical instrument from the theoretical frequency so as to better match the human senses, and the extent of the deviation can be easily changed.

It is also possible to shift the sound frequencies all at once according to the highest sound of the plurality of sounds produced, or a specific sound, while keeping the mutual ratio constant.

In addition, in Fig. 1, Fig. 4, Fig. 5, and Fig. 6, the launch circuit 1
Instead of inputting the output directly to the S rate multiplier 2, a code converter is inserted between the two, and by appropriately determining the code conversion, the pitch of the sound can be simply uniformly increased.
Instead of lowering the lower one, it is also possible to change the frequency of each sound from the theoretical frequency.

[Brief explanation of drawings]

1, 4, 5, and 6 are block diagrams explaining embodiments of the present invention, FIG. 3 is a partial circuit diagram thereof, and FIG. 2 is a waveform diagram explaining the present invention in detail. It is.

Claims (1)

[Claims] 1 Create a pulse train with an average interval proportional to the period of the sound to be generated, but extremely short compared to the period of the sound, step down this pulse, and add it to a counter that operates periodically every time a certain number of pulses are counted. In an electronic musical instrument that converts the electrical waveform generated by the periodic operation of the counter into sound, a reference pulse train is first created and the higher the value of one sound that satisfies a specific condition among the sounds to be generated, the more pulses are generated. The lower the pitch, the more pulse teeth are missing, and as a result, the average interval of the reference pulse train is slightly shifted, and from the pulse train whose average interval has shifted, the theoretical scale of the sound to be generated is calculated. When generating a large number of sounds at the same time by extracting or filling teeth according to the period ratio to obtain a pulse train with an average interval proportional to the period of the sound to be generated, and then creating sounds from this, the above specific conditions must be met. This electronic musical instrument is characterized in that the higher the satisfying note, the higher all the sounds generated at the same time are at the same ratio, and the lower the satisfying note is, the lower they are all at the same ratio.