JPH0137760B2 - - Google Patents

Description

[Detailed description of the invention]

[Industrial Application Field] The present invention relates to a wave generator for an electronic musical instrument that generates variable data that repeatedly changes at a cycle corresponding to a desired musical tone frequency and uses this variable data to generate musical tone signals. [Prior Art] When obtaining a musical tone signal or a sound source signal by sequentially reading sample point amplitude values of a waveform stored in a waveform memory, a constant corresponding to the frequency of the sound to be generated is repeatedly calculated (added or subtracted), An electronic musical instrument of a type in which the read address of the waveform memory is advanced in accordance with the calculation result is already known. Conventionally, in such electronic musical instruments, constants corresponding to the above-mentioned frequencies were stored in a storage device in correspondence with each note of the keyboard (Japanese Patent Application Laid-Open No. 1983-1999).
(See Publication No. 130213). [Problems to be solved by the invention] In conventional electronic musical instruments, constants corresponding to frequencies are stored in a storage device for each note on the keyboard. If there were 61 keys, 61 types of constants would have to be stored, which would require a relatively large capacity of the storage device, resulting in high costs. The present invention has been made in view of the above-mentioned points, and an object of the present invention is to provide a wave generator in which the capacity of a storage device for storing constants corresponding to frequencies is reduced. [Means and effects for solving the problem] According to the present invention, a keyboard having a plurality of keys;
key data generating means for generating key data consisting of note name data and octave data indicating the note name and octave of the operated key in response to a key operation on the keyboard; a storage means for storing a plurality of constant data consisting of a plurality of bits each corresponding to a frequency; and a reading means for reading out the constant data corresponding to the pitch name indicated by the pitch name data from the storage means based on the pitch name data. , variable data generation means for outputting variable data consisting of a plurality of bits that repeatedly changes at a cycle corresponding to the frequency of the musical tone based on the constant data read from the storage means; and a variable output from the variable data generation means. a changing means for changing the variable data by shifting the bit position of the data according to the octave data, and outputting data that changes repeatedly at a cycle corresponding to the musical tone frequency of the key indicated by the key data; and the changing means and musical tone generating means for generating a musical tone signal of a corresponding frequency based on the data output from the. [Embodiment] Hereinafter, an embodiment of the present invention will be described in detail based on the accompanying drawings. FIG. 1 is a schematic block diagram showing an example of an electronic musical instrument using the wave generator 10 of the present invention, in which a key press detection circuit 12 detects the ON or OFF operation of the key switch of each key arranged on the keyboard 11.
Outputs information identifying the pressed key. The sound generation assignment circuit 13 receives information identifying the pressed key from the pressed key detection circuit 12, and assigns the sound of the key represented by this information to one of the channels corresponding to the maximum number of simultaneous sounds (for example, 12 notes). Assign. The sound generation assignment circuit 13 has a memory location corresponding to each channel, stores a key code KC representing a certain key in the memory location corresponding to the channel to which the sound of a certain key is assigned, and stores the key code KC stored in each channel. KC is output sequentially in a time-division manner. In order to identify each key on the keyboard 11, as shown in Table 1, the key code KC is a 2-bit keyboard code K ₁ , K ₂ representing the keyboard type, and 3 representing the octave range.
It consists of a total of 9-bit codes: octave codes B ₁ , B ₂ , B ₃ and 4-bit note codes N ₁ , N ₂ , N ₃ , N ₄ representing the 12 note names within one octave. Ru. In this embodiment, it is assumed that the keyboard range of the keyboard 11 is from _C2 note to _C7 note.
The octave code "000" in the first octave range is used only for the lowest _C2 note, and the codes _B3 to _N1 are "0001110". The octave codes B ₃ , B ₂ , and B ₁ in the second octave range with the value “001” are used from C# ₂ to C ₃ , and in the same way from C# on the bass side to C on the treble side. The same octave codes B ₃ , B ₂ , and B ₁ are used, assuming that the range up to the note is one octave, and the octave code of the 6th octave range "101" is used for C# ₆ to C ₇ notes.

【table】

[Table] In this embodiment, in order to be able to produce multiple sounds at the same time, various counters, logic circuits, storage devices, etc. are configured in a dynamic logical manner so that they are used in a time-sharing manner. The time relationship of the clock pulses regulating operation is extremely important. Figure 2a is a graph mainly showing the clock pulse φ ₁ , which controls the time division operation of each channel, and has a period of, for _example , 1 μs (microsecond; 10 ^-6 seconds). There is. Since the number of channels is 12, the 1 μs wide time slots successively separated by the main clock pulse _φ1 are the first
Channels 1 to 12 will be supported sequentially. As shown in FIG. 2b, each time slot is designated as a first channel time to a twelfth channel time in order. Each channel time occurs cyclically. Therefore, the key code KC representing the key to which the sound generation is assigned by the sound generation assignment circuit 13 is sequentially output in a time-division manner in accordance with the time of the assigned channel. For example, the first channel is assigned the C note (C ₃ ) in the second octave range of the pedal keyboard, the second channel is assigned the G note (G ₆ ) in the fifth octave range of the upper keyboard, and the third channel is assigned the C note (C 3 ) in the second octave range of the pedal keyboard. C note (C ₆ ) in the 5th octave range of the upper keyboard
is assigned, and the E note (E ₅ ) in the fourth octave range of the lower keyboard is assigned to the fourth channel.
Assuming that no sound is assigned to the 5th to 12th channels, the content of the key code KC output from the sound generation assignment circuit 13 in a time-sharing manner in synchronization with the time of each channel is as shown in Figure 2c. . All outputs from the 5th channel to the 12th channel are "0". Furthermore, the sound generation assignment circuit 13 time-divisionally outputs an attack start signal (or key-on signal) AS indicating that sound should be generated in the channel to which the pressed key is assigned the sound generation, in synchronization with the time of each channel. Furthermore, a decay start signal (or key-off signal) DS indicating that the key assigned to each channel is released and the sound should be attenuated is output in a time-divisional manner in synchronization with the time of each channel. . These signals AS,
DS is amplitude envelope control of musical tones (pronunciation control)
used for. Furthermore, the sound generation assignment circuit 13
Then, the decay end signal DF indicating that the sound generation in that channel has ended is received from the envelope generator 14, and based on the signal DF, a clear signal CC is generated that clears various memories related to the channel and completely cancels the sound generation assignment. Output. In the example shown in Figure 2c, the keys assigned to the first and second channels are currently being pressed, and the keys assigned to the third and fourth channels have been released and their sound is attenuated. , in the fourth channel, the sound generation ends at time slot _t1 , and the decay end signal DF is generated.
Assuming that the clear signal CC is output at time slot t ₂ with a delay of 12 channels, Fig. 2 d
As shown in ~g, each signal AS, DS, DF, CC is generated. In addition, when time slot t ₂ , the clear signal
Since CC is output, the attack start signal AS and decay start signal DS of the fourth channel are erased. At this time, the key code KC of the fourth channel time in FIG. 2c is erased, but is drawn as is for convenience of explanation. Various signals output from the sound generation assignment circuit 13
As shown in FIG. 2, the channels to which KC, AS, DS, and CC belong can be distinguished based on the channel time. Key code output from pronunciation assignment circuit 13
KC joins the wave generator 10, and the wave generator 10 receives variable data (address data) for sequentially reading sample point amplitude values of the musical tone (sound source) waveform from the waveform memory 15.
XqF is generated. In this embodiment, the waveform memory 15 is configured such that sound source waveforms having different numbers of feet are read out in parallel, 2 feet 2',
4 feet 4', 8 feet 8', 16 feet 1
Memories 15A to 15E are provided corresponding to the sounds of 6' and 32 feet 32', respectively. The envelope generator 14 is connected to the sound generation assignment circuit 1
Based on various signals AS, DS, CC, etc. given from 3, an envelope waveform EV having characteristics such as attack decay is generated. Using this envelope waveform EV, a time-dependent volume amplitude envelope is added to the sound source waveform signal read out from the waveform memory 15, and the sound generation is controlled. The sound source waveform signals of each foot 2' to 32' read out from the waveform memory 15 are applied to a distribution circuit 16, and are sent to lines 17 or 18 depending on the type of keyboard.
will be distributed accordingly. The sound source waveform signal distributed to the line 17 is applied to a voltage controlled filter (VCF) 19 to perform timbre control. The sound source waveform signal distributed to the line 18 is applied to a timbre filter group 20, where timbre control is performed. Voltage controlled filter 1
The filter characteristics of 9 are varied by a control voltage (not shown), and the filter characteristics of timbre filter group 20 are fixed corresponding to various tones. Therefore, sounds with different tonal qualities are produced in both series.
The timbre-controlled musical sound waveform signal passes through an appropriate sound control circuit 21 such as an expression, and is outputted from a sound system 22. A detailed example of the wave generator 10 is shown separately in FIGS. 3 and 4. A detailed example of the note decoder 23, constant memory 24, and key voltage generation circuit 25 of the wave generator 10 is shown in FIG.
7. A detailed example of the bit position switching circuit 28 is shown in FIG. In the circuits of FIGS. 3 and 4, various circuit elements are illustrated in the manner shown in FIG. Figure 5a is an inverter, Figures b and c are AND circuits,
Figures d and e in the figure show an OR circuit, and f in the figure shows a 1-bit delay flip-flop. If the number of inputs is small in an AND circuit or an OR circuit, b and d in the same figure
A normal display diagram as shown in Figure 1 is used, and if there are a large number of inputs, the diagrams shown in Figures c and e are used. In Figures c and e, one input line is drawn on the input side of the circuit, and multiple signal lines are crossed with this input line.
The intersection point between the signal line of the signal to be input to the circuit and the input line is surrounded by a circle. Therefore, in the case of c in the same figure, the logical formula is Q=A・B・D, and in the case of the example in e of the same figure, the logical formula is Q=A+B+C. Further, FIG. 5g shows a shift register, and the number of the numerator of the fraction shown in parentheses in the block indicates the number of stages of the shift register, and the number of the denominator indicates the number of bits of input data to the shift register. Although shift clock pulses are not particularly shown in the delay flip-flop and shift register, they are shifted by a main clock pulse φ ₁ (specifically, a two-phase clock pulse) having a period of 1 μs. Key code supplied from pronunciation assignment circuit 13
Of the KCs, no codes _N1 to _N4 are applied to the note decoder 23, and decoded outputs are generated on output lines corresponding to the note names represented by the note codes _N1 to _N4 . The constant memory 24 stores in advance constant data F in binary format that is proportional to the frequencies of the 12 note names C to B within one octave.
In response to the output of No. 3, constant data F corresponding to the note names represented by the note codes N ₁ to _{N 4} is read out. The accumulator 26 repeatedly adds the constant data F read from the memory 24 at regular time intervals T to obtain variable data qF. q is a variable that increases sequentially as 1, 2, 3, 4, . . . each time the time interval T elapses. The output qF of the accumulator 26 is applied to the bit position switching circuit 28 and the octave control circuit 2
The binary bit position is shifted to the left or right according to the value of the octave switching data X given from 7. In other words, variable data qF that changes regularly corresponding to each note frequency within one octave.
is multiplied by octave switching data X, and variable data corresponding to the frequency of the note name in the octave range specified by the octave switching data
Get XqF. Octave switching data X is octave code
Generated based on the contents of _B1 to _B3 . Since the gap in one octave has a frequency ratio of 2:1, if the octave range to which the note name stored in the constant memory 24 belongs is the basic octave, then the octave range of the note to be generated is n more than this basic octave. When the octave is above, the value of X is 2 ⁿ , and when it is n octaves below, the value of X is 2 ^{- n} .
In other words, since the frequency relationship in octave units is a power of 2, the binary bit position switching circuit 28
This allows octave switching control to be performed. By the way, the value of the constant data F stored in the constant memory 24 is the frequency f of the note name in the basic octave, the number N of times the same data F is added per second in the accumulator 26, and the value stored in the waveform memory 15. It is determined by the number M of addresses for one cycle of the waveform. That is, the decimal value of the constant data F is determined by the relational expression F=f·M/N (1). Constant data F corresponding to each pitch name determined based on equation (1) is converted into a binary number, and this binary constant data F is stored in the constant memory 24 in advance. In the variable data XqF, the value of the variable q becomes N after one second, so the above equation (1) can be rewritten as follows. XF=Xf·M/N (2) Therefore, the frequency is switched in proportion to the value of the octave switching data X. As mentioned above, since the data X is given as a power of 2, the frequency Xf is switched as a power of 2 of the frequency f of the fundamental octave, and the frequency of the generated sound is switched in octave units. That is, the octave range of the generated sound is selected. Therefore, when the bit position of the output data qF of the accumulator 26 is shifted to the left (to the upper digits) in the bit position switching circuit 28, multiplication by 2 ⁿ is performed by the data X. The octave goes up by a number n. Furthermore, when the bit position of data qF is shifted to the right (to the lower digit) in the same switching circuit 28, multiplication by 2 ^{- n} is performed by data X, so
The octave is lowered by the number n of shifted bit positions. In FIG. 3, the note decoder 23 outputs note codes N ₁ to _{N 4} as shown in Table 1 above.
It is constituted by a group of AND circuits 23A having a logic so that it can decode each pitch name C# to C separately. The constant memory 24 stores 10-bit binary constant data F of a predetermined value in accordance with the output of the decoder 23.
( _F1 to _F10 ); and a gate section 24 that selects the output of the OR circuit group 24A according to the contents of the keyboard codes _K1 and _K2 .
It has B. Since the bit position switching circuit 28 shown in FIG. 4 is designed to perform a left shift operation (multiplication by ²ⁿ ), the constant data F proportional to the frequency of each of the 12 tones in the lowest octave range is stored in the constant memory 24. I remember. However, the true lowest octave range shown in Table 1 above is the octave code B ₁ , B ₂ ,
The content of _B3 is "000" in the first octave range, but as mentioned above, the only note belonging to this first octave range is the _C2 note. Therefore, in the wave generator 10 of this embodiment, the second octave range in which the contents of the octave codes B ₃ , B ₂ , and B ₁ are "001" is handled as the lowest octave (basic octave), and the second octave range is Constant data F corresponding to the frequencies of the 12 tone names (C ₂ #, D ₂ , . . . B ₂ , C ₃ ) belonging to the note name is stored in each address of the memory 24, respectively. Constant data F is also specially stored for the _C2 note, which is the only note belonging to the first octave range. In other words, as shown in Table 2, the C note (C ₃ ) belonging to the basic octave (second octave range)
Logic is set up in the OR circuit group 24A so that data with a value of 1/2 is read out as constant data F of the _C2 note. In addition, in this embodiment, data F with slightly different values for each type of keyboard is stored in the memory 2.
4, Table 2 shows only the lower keyboard tones.

[Table] Constant data F related to _C2 note is used only for _C2 note, so the note decoder 23 determines whether note codes _N1 to _N4 are for _C2 note or other C notes. It is necessary to distinguish between Therefore, an AND circuit 29 dedicated to the _C2 note is provided in the AND circuit group 23A. This AND circuit 29
is added with data B ₁ to _{B 3} obtained by inverting the data of each bit B 1 to B ₃ of the octave code using an inverter, and if the octave code _{B 1 to B 3} is in the first octave range, the data “000” is Since it is inverted and becomes "111", the AND circuit 29 becomes operable. Therefore, the C note in the first octave range, i.e.
When the note codes N ₁ to _{N 4} of C ₂ notes are supplied, the output of the AND circuit 29 becomes “1” and the memory 24
Constant data F exclusively for the _C2 note is read from. Note that all the constant data F read from the constant memory 24 are treated as belonging to the same octave range (basic octave), so the octave codes B ₁ to _{B 3} representing the first octave range corresponding to the _C2 note are It is necessary to convert the data into a two-octave range. Therefore, the octave chord
Input B ₁ to _{B 3} to the NOR circuit 30 in Fig. 4 and set it to “000”.
When , the signal “1” generated from the NOR circuit 30 is added to the OR circuit 31 to set the data of bit _B1 to “1”.
Make it. As a result, the octave chords B ₃ , B ₂ ,
If B ₁ is “000”, it is converted to “001”, and then
The signal is input to a 3-bit 12-stage shift register 27A of the octave control circuit 27. The output "1" of the NOR circuit 30 is inverted by the inverter 34 in FIG. The AND circuit 33 is a circuit for decoding note codes N ₁ to N 4 of C note other than C note (C ₂ note) in the first octave range, and the note codes N ₁ to _{N 4 of} C note ₂ are supplied. In this case, the signal "1" from the line 32 disables the AND circuit 33. In this embodiment, the memory 2 stores constant data F with slightly different values depending on the keyboard to which it belongs even if the note name is the same.
It is designed to be read from 4. In FIG. 3, keyboard codes K ₁ and K ₂ are decoded by a decoder 24C, and depending on the contents, upper keyboard signals UE and
Generated by lower keyboard signal LE or pedal keyboard signal PE. From the OR circuit group 24A, constant data F regarding the same note name is read out in parallel for each keyboard based on the output of the note decoder 23, and these constant data F are sent to the gate section 24B as the keyboard signals UE, LE, PE. selected according to As is clear from the input connection mode in the OR circuit group 24A, the data of bits lower than bit _F4 differs depending on the keyboard, and the bit
The data for bits higher than _F5 is the same for each key. Therefore, the difference in the value of the constant data F for each keyboard is very small, and this difference causes a slight pitch shift in the musical waveform signal generated using the constant data F. In this embodiment, the constant data F for each of the note names C# to C of the lower keyboard tones set as shown in Table 2 can generate a musical sound waveform of a regular pitch. Therefore, as shown in FIG. 6, the pitch shift of the lower keyboard tone is 0 cents. Furthermore, the constant data F corresponding to each note name of the upper keyboard tone is set to a value that can generate a musical sound waveform with a pitch slightly higher than the normal pitch, and the pitch deviation of each note name is uniform. As shown in Figure 6, the pitch shift increases by +3.7 as the pitch increases from C# to C.
cents to +1.86 cents. Further, the constant data F corresponding to each note name of the pedal keyboard tone is set to a value that can generate a musical sound waveform with a pitch slightly lower than the normal pitch. Moreover, as shown in FIG. 6, as the pitch increases from C# to C note, the pitch shift amount shifts from -3.7 cents to -1.86 cents. As mentioned above, the reason why the value of the constant data F is made to be slightly different for each keyboard is to prevent the occurrence of a situation where the phases of musical sound waveforms generated at the same time are completely reversed and the generated sound is canceled out. This is to do so. In other words, it is common for two keys of the same pitch to be pressed at the same time on different keyboards, and in this case, if both keyboard tones are generated at exactly the same frequency, the phases of the two tones will be opposite. In some cases, the inconvenience arises that the sound is canceled out. However, if the value of the constant data F is slightly different as in this example, even if keys of the same pitch are pressed at the same time on different keyboards, the phases of the sounds of both pressed keys will be completely opposite. cannot occur, and the sound is never cancelled. In the gate section 24B, the keyboard signals UE, LE
Alternatively, each bit data F ₁ to F ₁₀ of constant data F is selected through an AND circuit enabled to operate by PE, and each bit is grouped by an OR circuit. Since the key codes N ₁ to N ₄ , B ₁ to B ₃ , K ₁ , and K ₂ are supplied time-divisionally for each channel as described above, the constant memory 24 stores the sounds assigned to each channel. Constant data F (F ₁
~F ₁₀ ) are read out in a time-division manner. The read constant data F is supplied to the accumulator 26 shown in FIG. Accumulator 26 is
It has a 21-bit adder group 26A and a 21-bit 12-stage shift register group 26B. The output of the adder group 26A is delayed by 12 channel times in the shift register group 26B via an AND circuit group 26C. Each bit output Q ₁ of shift register group 26B
.about.Q ₂₁ (qF) is applied to the bit position switching circuit 28 and fed back to the adder group 26A, where constant data F ₁ to _{F 10} (F) are added. Therefore, it is possible to accumulate constant data F on a time-division basis for each channel, and constant data F regarding one channel is repeatedly added every 12 channel times (12 μs). A full adder is used as the adder for the lower 10 bits of the adder group 26A, and each bit data F ₁ to _{F 10} of the constant data F becomes an addition input to the lower 10 bit full adder group. Half adders are used for the upper 11 bits of the adder group 26A, and in these half adders, the lower bits contain the carry signal applied from the adder's carry output CO to the carry input CI and the shift register. Bits returned from corresponding bit positions in group 26B are added. The AND circuit group 26C is rendered inactive by the clear signal CC applied from the sound generation allocation circuit 13, and clears the calculated value of the accumulator 26. The bit position switching circuit 28 includes an OR circuit group 28A consisting of 15 5-input type OR circuits for deriving each bit S ₁ to S ₁₅ of variable data XqF that has been octave switched, and each OR circuit group 28A in this OR circuit group 28A. A total of 5 each corresponding to the OR circuit.
The AND circuit group 28B is composed of 75 two-input type AND circuits. In the AND circuit group 28B, five AND circuits are provided corresponding to each bit _S1 to _S15 , and each of the AND circuits can be operated individually according to the signals from the five output lines of the decoder 27B of the octave control circuit 27. It has come to be regarded as Output Q ₁ to Q ₂₁ of accumulator 26
are the five outputs X ₁ , X ₂ , X ₃ ,
Predetermined bits S ₁ to _{S 15} corresponding to X ₄ and X ₅ , respectively
The signals are input to the AND circuit group 28 so as to be distributed to the following. Table 3 shows how variable data XqF (S ₁ to S ₁₅ ) is obtained by switching the bit positions of each bit Q ₁ to _{Q 21} of variable data qF in accordance with each output X ₂ to _{X 1} of the decoder 27B. It is as follows.

〔Effect of the invention〕

As explained above, according to the present invention, it is only necessary to store in the memory constant data corresponding to frequencies that correspond to each note within a desired octave. It is much smaller than conventional devices for storing constant data.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the wave generator of the present invention in relation to the overall configuration of an electronic musical instrument, and FIG. 2 is a timing chart for explaining the operation of the sound generation assignment circuit shown in FIG. 1. , Figures 3 and 4 are circuit diagrams showing detailed examples of each part of the wave generator shown in Figure 1, Figure 5 is a diagram for explaining how to illustrate various circuit elements, and Figure 6 shows constants for each keyboard. FIG. 7 is a graph showing the state of pitch deviation of each keyboard tone achieved by slightly different values of data F, and is a timing chart for explaining the highest note detection operation in the key bolt generation circuit. 10... Wave generator, 15... Waveform memory, 24... Constant memory, 26... Accumulator, 27... Octave control circuit, 28... Bit position switching circuit.

Claims (1)

[Claims] 1. A keyboard having a plurality of keys, and generating key data consisting of pitch name data and octave data indicating the pitch name and octave of the operated key in response to key operations on the keyboard. key data generation means; storage means for storing a plurality of constant data consisting of a plurality of bits each corresponding to the frequency of a musical tone of each note name in a predetermined octave; reading means for reading the constant data corresponding to the note name indicated by; and a variable for outputting variable data consisting of a plurality of bits that repeatedly changes at a cycle corresponding to the frequency of the musical tone based on the constant data read from the storage means. data generating means; changing the variable data by shifting the bit position of the variable data output from the variable data generating means according to the octave data;
A changing means for outputting data that repeatedly changes at a cycle corresponding to a musical tone frequency of a key indicated by the key data, and a musical tone generating means for generating a musical tone signal of a corresponding frequency based on the data output from the changing means. A wave generator for electronic musical instruments.