JPS6341080B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to a musical sound waveform synthesis device capable of forming musical sound waveform signals of arbitrary tones by suitably combining and synthesizing a plurality of waveform signals having frequency spectra in a specific region. 2. Description of the Related Art Conventionally, there is a so-called harmonic synthesis method as a musical tone synthesis method for electronic musical instruments in which a musical tone signal of an arbitrary tone is formed by appropriately synthesizing waveform signals. This harmonic synthesis method was constructed based on the fact that musical tone signals of various tones are formed by combining harmonic components of each order, and each waveform signal corresponding to each harmonic component is memorized. A plurality of waveform memories are provided, and the waveform signals read from these waveform memories are multiplied by predetermined harmonic coefficients selected in accordance with the designation of the timbre, and the results are synthesized to generate a musical tone signal. It forms the However, although this harmonic synthesis method can theoretically synthesize signals of any tone depending on the selection of harmonic coefficients, it requires a large number of memories to store each harmonic component. There were drawbacks such as a complicated circuit configuration for controlling the multiplication of harmonic coefficients for each harmonic. In addition to the above-mentioned harmonic synthesis method, there is also a method in which the waveform signals of various tones are directly stored in the waveform memory, and these are read out in accordance with the specified timbre. There were drawbacks such as the need for a large number of waveform memories and the difficulty in obtaining changes in timbre over time. This invention was made in order to eliminate the drawbacks of the conventional method described above, and it is possible to generate a wide variety of musical sound waveforms using a small number of waveform memories, and also to easily control changes in timbre over time. It is an object of the present invention to provide a musical sound waveform synthesis device. According to the present invention, attention is paid to the fact that a musical sound waveform of a natural musical instrument or the like generally has frequency spectrum groups in a plurality of specific regions called formants.
A plurality of waveform memories (hereinafter referred to as spectral group memories) each storing a waveform signal formed by synthesizing each frequency component corresponding to a frequency spectrum group in a specific region are provided, and the waveform signals read from these spectral group memories are appropriately read out. A musical waveform signal of a desired tone is obtained by combinatorial synthesis. Further, by changing the above combination over time, a change in timbre over time of the musical waveform signal is obtained. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 shows an embodiment of an electronic musical instrument to which a musical sound waveform synthesis device according to the present invention is applied. The key switch circuit 1 includes a key switch corresponding to each key of a keyboard section (not shown), and in response to the press of a key, key information KC indicating the pressed key and a key-on signal indicating that the key is pressed.
Generate KON. Such key switch circuits are well known. Key information KC generated from the key switch circuit 1 is applied to the address generator 2. In response to the added key information KC, the address generator 2 generates an address signal ADR with a period corresponding to the pitch of the sound indicated by the key information KC.
Note that the period of the address signal ADR generated from the address generator 2 is determined in relation to the number of sample points of the stored waveforms in the spectrum group memories SGM1 to SGMn, which will be described next.
The address signal ADR generated from the address generator 2 is the spectral group memory SGM1.
~Added to SGMn. Spectral group memories SGM1 to SGMn are
This constitutes the main part of the musical sound waveform synthesizer type according to the present invention, and each memory SGM1 to SGMn stores a waveform signal formed by synthesizing each frequency component in a specific frequency spectrum group, and stores amplitude digital values at a plurality of sample points. is remembered as. The above-mentioned specific frequency spectrum group is not particularly limited, but in order to obtain a timbre close to the sound of a natural musical instrument, it is preferable to use one that corresponds to the formants of various natural musical instruments. In order to facilitate understanding of the waveforms stored in the spectral group memories SGM1 to SGMn, an example of frequency spectrum groups regarding the waveforms stored in each memory in the case of n=10 is schematically shown in Figure 2. become. i.e. in this case the spectral group memory
SGM1 stores a waveform obtained by synthesizing each frequency component corresponding to the spectrum represented by spectrum group SG1, and the spectrum group memory
SGM2 stores a waveform obtained by synthesizing each frequency component corresponding to the spectrum represented by spectrum group SG2, and similarly, spectrum group memories SGM3 to SGM10 store each frequency component corresponding to the spectrum represented by spectrum groups SG3 to SG10, respectively. The synthesized waveform is stored. In FIG. 2, the horizontal axis shows the harmonic order, and the vertical axis shows the level. By the way, in the present invention, a waveform signal of a desired timbre is formed by synthesizing one or more of the waveforms stored in the spectral group memories SGM1 to SGMn. For example, if this is explained based on the spectrum groups SG1 to SG10 shown in FIG.
1, T2, . . . are synthesized.

[Table] In Table 1, for example, timbre T1 is obtained by combining waveform signals corresponding to spectral groups SG1, SG2, SG3, and SG4, and timbre T4 is similarly obtained by combining waveform signals corresponding to spectral groups SG5, SG5, and SG4, respectively.
It is obtained by combining waveform signals corresponding to SG2 and SG8, respectively. Note that the combination examples shown in Table 1 are merely examples of combinations, and it will be easily understood that there are many more combinations. The tone selector 3 has a plurality of tone selection switches corresponding to each tone, and the tone selection switches select the spectral group memory.
The combination of waveform signals read from SGM1 to SGMn is controlled. The output of the tone color selection switch corresponding to each tone output from the tone selector 3 is applied to the memory selector 4. The memory selector 4 is composed of, for example, a read-only memory (ROM), and applies an enable signal to each enable terminal number EN of the spectrum group memories SGM1 to SGMn in response to the selection of tone color by the tone selector 3, thereby controlling the above-mentioned combinations. For example, if tone selector 3 selects tone T2 shown in Table 1 above, memory selector 4 selects spectral group memories SGM1, SGM6, SGM8,
A signal "1" is applied to the enable terminal EN of the SGM 10, and a signal "0" is applied to each enable terminal EN of the other spectrum group memories. As a result, the spectral group memories SGM1, SGM6,
Only SGM8 and SGM10 become operational. The spectral group memories SGM1 to SGMn, which have become operational by the output of the memory selector 4, sequentially store the amplitude digital values at each sample point stored in each address in response to the address signal ADR applied from the address generator 2 mentioned above. read out. The amplitude digital values read from the spectrum group memories SGM1 to SGMn are added and synthesized by an adder 5, and then sequentially added to a multiplier 6. The multiplier 6 receives an envelope signal (digital signal) generated from the envelope generator 7.
has been added. envelope generator 7
, a predetermined envelope waveform is stored as amplitude digital values at a plurality of sample points, and these amplitude digital values are sequentially read out in response to the key-on signal KON output from the key switch circuit 1 described above and output as an envelope signal. do. A multiplier 6 multiplies the output of the adder 5 by this envelope signal to give a predetermined envelope, which is then sent to a digital-to-analog converter 8.
The signal is converted into an analog signal and added to the sound system 9 to be generated as a musical tone. As described above, in this embodiment, only a predetermined memory among the spectral group memories SGM1 to SGMn becomes operable in response to the selection of a tone by the tone selector 3, and the data is stored in this memory that has become operable. Waveform signals representing a predetermined frequency spectrum group are synthesized to form a musical tone signal of a desired tone. In the above embodiment, in order to clarify the principle of the present invention, the case of a so-called monophonic musical instrument with a maximum simultaneous number of sounds of 1 was shown, but in the case of a visual musical instrument with a maximum simultaneous number of sounds of 2 or more, For example, you can use a pronunciation assignment circuit (key assigner) to generate key information for each pressed key.
Each KC is assigned to one of multiple pronunciation channels, and the key information KC assigned to each channel is
is output on a time-division basis in synchronization with each channel time,
And spectrum group memory SGM1~
Waveform reading from SGMn, addition synthesis in adder 5, and envelope application in multiplier 6 are performed in a time-division manner for each channel time, and this is performed by an accumulator (not shown) at one time for each channel time.
It is only necessary to accumulate the sum every cycle and add it to the digital-to-analog converter 8. FIG. 3 shows another embodiment of the present invention, in which the combination of waveform signals to be synthesized is changed in accordance with the progression of the envelope to obtain a change in timbre over time. In FIG. 3, blocks that perform the same functions as those in the embodiment shown in FIG. 1 are given common reference numerals to simplify the explanation. In this example, the envelope generator 7
By further controlling the memory selector 4 by the output of an envelope counter 71 provided therein, timbre control corresponding to the progression of the envelope is performed. That is, the memory selector 4 is supplied with the output of a tone selector 3 consisting of a tone color selection switch and the output of an envelope counter 71 for advancing reading of an envelope memory 72 in an envelope generator 7. The memory selector 4 stores a spectrum group memory corresponding to the output of the tone selector 3 and the output of the envelope counter 71 of the envelope generator 7.
Send an enable signal to the enable terminal EN of SGM1 to SGMn. Therefore, the operable memory among the spectral group memories SGM1 to SGMn corresponds to the selection of tone color by the tone selector 3 and changes in accordance with the progression of the envelope. This makes it possible to form a musical tone signal whose timbre changes in accordance with the progression of the envelope. In addition, in Fig. 3, the other blocks are the first block.
Perform the same operation as shown in the figure. Further, as shown by the dotted line in FIG. 3, instead of adding the output of the envelope counter 71 to the memory selector 4, the output of the envelope memory 72 may be added to the memory selector 4. In this case, it is possible to obtain a musical tone signal whose timbre changes in accordance with the envelope signal (the amplitude value of the envelope waveform). Alternatively, the output of the envelope counter 71 is not applied to the memory selector 4, but an independent time information generation circuit is provided, and the memory selector 4 is controlled by the output of this time information generation circuit to obtain a change in tone over time. be able to. FIG. 4 shows another embodiment configured as described above, in which only the parts related to the tone selector 3 and memory selector 4 are shown. Memory selector 4 contains the output of tone selector 3 and counter 1.
An output of 0 is added. The counter 10 corresponds to the above-mentioned time information generating circuit, and is reset by the output of the one-shot circuit 11 triggered by the key-on signal KON output from the key switch circuit 1 (not shown), and is reset by the output of the one-shot circuit 11.
The output of NA is driven by a clock pulse φ of a predetermined frequency applied to the clock input CK via an AND circuit AN which joins the other inputs. Then, when each bit of the counter 10 becomes "1", the NAND condition of the NAND circuit NA to which each bit output is added is satisfied, and the output of the NAND circuit NA becomes "0", thereby making the AND circuit AN inoperable. The operation of the counter 10 then stops. That is, the counter 10 starts operating in response to the key-on signal KON, and supplies the counted value to the memory selector 4 as time information until the counted value reaches the maximum value. As a result, memory selector 4 of spectrum group memories SGM1 to SGMn
The selected memory changes in accordance with the timbre selection by the tone selector 3 and the count value of the counter 10 (that is, the elapsed time since the key-on signal KON was generated), thereby obtaining a timbre change over time. Figure 5 shows spectrum group memory SGM1~
This shows another embodiment in which the envelopes of the outputs of SGMn are individually controlled and these are added and synthesized to form musical tones. Note that what is shown in FIG. 5 shows only the essential parts of this embodiment, and the other parts are the same as what is shown in FIG. 1. Also the first
Blocks that perform the same functions as those shown in the figures are given the same reference numerals. In this embodiment, a plurality of envelope generators EG1 to EGn are provided corresponding to the spectral group memories SGM1 to SGMn. These envelope generators EG1 to EGn each convert a predetermined envelope waveform into an amplitude digital value (envelope signal) at a plurality of sample points.
and store this as key switch circuit 1 (first
Key-on signal generated in response to a key press from (Fig.)
It is read out sequentially in synchronization with KON, and is similar to the envelope generator 7 shown in FIG. Spectral group memories SGM1 to SGMn are
Memory selector 4 is connected to each enable terminal EN.
An enable signal generated from the address generator 2 (FIG. 1) is added, and only the memory selected in accordance with the tone selection in the tone selector 3 (FIG. 1) becomes operable.
The amplitude digital values at each sample point of the stored waveform are sequentially read out in response to the address signal ADR applied from FIG. spectral group memory
The outputs of SGM1 to SGMn are each multiplier MU1
~MUn is added to the multiplier MU1~MUn
Envelope signals generated from envelope generators EG1 to EGn are applied to other inputs of the envelope generators EG1 to EGn, respectively. Therefore, the multiplier corresponding to the memory that has become operational multiplies the amplitude digital value read from the memory by the envelope signal output from the envelope generator, and the outputs of these multipliers are sent to the adder. 5 and are added and synthesized. The output of the adder 5 is then converted into an analog signal by a digital-to-analog converter 8 and sent to a sound system 9 (Fig. 1).
pronounced as musical tones. With such a configuration, a change in timbre corresponding to the progression of the envelope can be obtained. That is, since each envelope signal generated from the envelope generators EG1 to EGn has different attack time, decay time, attack level, sustain level, etc., the level of each waveform signal synthesized by the adder 5 depends on the progression of the envelope. Therefore, it will change sequentially. This brings about a change in the synthesized waveform over time, resulting in a musical tone whose timbre changes over time. However, among the spectrum group memories SGM1 to SGMn, the combination of memories that can be operated is memory selector 4 (Figure 1).
, which does not vary with respect to the envelope, resulting in a different timbre change from that shown in FIGS. 3 or 4. In this embodiment, as in the case of FIG. 3 or 4, the spectral group memory SGM
If the combination of operable memories among 1 to SGMn is further changed, even greater changes in timbre over time can be obtained. In the embodiment shown in Fig. 5, envelopes are applied separately to the outputs of each spectral group memory SGM1 to SGMn by parallel processing, but in Fig. 6, envelopes are applied by time-sharing processing. This shows another example. FIG. 7 shows a timing chart of the channel signals CH1 to CHn indicating the time division channels used in this embodiment in relation to the clock pulse nφ. In FIG. 6, the address generator 2 is
It is equipped with a frequency numerical information memory 21 that stores numerical information F corresponding to the frequency of each pitch at each address, an accumulator 22 that accumulates the numerical information F read from this frequency numerical information memory 21, and a key switch circuit. 1 (FIG. 1) is added to the frequency numerical information memory 21 as an address designation signal.
, the accumulator 22 accumulates this for each channel signal CH1 indicating the time-division first channel, and outputs the accumulated value qF to the spectrum group memories SGM1 to SGMn as an address signal ADR. . Spectral group memories SGM1 to SGMn are
An enable signal is applied to the enable terminal EN from the memory selector 4 (FIG. 1) via AND circuits A1 to An that are operable at the timing of each channel signal CH1 to CHn. Therefore, the spectral group memories SGM1 to SGMn
Among them, only the memory selected corresponding to the tone selected by the tone selector 3 (Fig. 1) becomes operational at the corresponding channel time, and each sample point of the waveform stored in the memory that became operational is The amplitude digital values at are read out in a time-division manner. The amplitude digital values read out in this time-division manner are applied to a multiplier 12. The output of an envelope generator circuit 13 is added to the other input of the multiplier 12, and this envelope generator circuit 13 is connected to each spectral group memory.
Envelope signals corresponding to SGM1 to SGMn are time-divisionally output as time-division envelope signals. Therefore, the multiplier 12 sequentially time-divisionally multiplies the outputs of the spectral group memories SGM1 to SGMn, which are read out in a time-division manner, by the time-division envelope signal read out from the envelope generator circuit 13. The output of this multiplier 12 is applied to an accumulator 14. Accumulator 14
is to sequentially accumulate the values added by the clock pulse nφ, and the accumulated value is used to delay the output of the one-shot circuit 17 triggered by the channel signal CH1 indicating the first channel time by the delay circuit 16. This signal is cleared by the signal. As a result, the accumulator 14 accumulates the output values of all channels while the channels go around once, and this accumulated value is loaded into the register 15 by the output of the one-shot circuit 17 triggered by the channel signal CH1. is added to the sound system 9 via the digital-to-analog converter 8, and produced as a musical tone. With such a configuration, it is no longer necessary to provide an envelope generator for each of the spectral group memories SGM1 to SGMn, and the envelope generator circuit 13 can be configured with one envelope generator circuit. In addition, as shown in FIG. 8, the address generator 2 and the spectrum group memory SGM (SGM
1 to SGMn) between the address signal conversion circuit AC
By inserting the spectral group memory, which stores the waveform representing a certain frequency spectral group,
Waveforms representing other frequency spectrum groups can be read out from the SGM. For example, if the address signal ADR generated from the address generator 2 is x, and if this signal x is converted into a signal ax that changes at a frequency times a by the address signal conversion circuit AC, then the spectral group memory SGM Even if the spectral group corresponding to the stored waveform is the one shown by SG in Figure 9, the spectral group shown by the dotted line
The waveform corresponding to SG' is read out. i.e. 1
By using one spectrum group memory SGM, it is possible to obtain a waveform signal similar to that obtained when a plurality of spectrum group memories SGM are provided, and the number of memories can be reduced. Note that the above address signal conversion circuit AC is, for example,
If conversion of ax±b (±b is phase conversion) is performed, it can be easily configured by a combination of a bit shift circuit and an addition/subtraction circuit. As explained above, according to the present invention, it is possible to synthesize many types of tones using a small number of waveform memories, it is also possible to easily obtain changes in tones over time, and the musical waveform synthesis has a simple configuration. equipment can be provided. It should be noted that although the embodiments shown in this specification only use digital memory, it will be easily understood that the same configuration can be made using analog memory.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of this invention, FIG. 2 is a schematic diagram showing an example of a frequency spectrum group, FIG. 3 is a block diagram showing an embodiment of this invention, FIGS. 6 and 8 are main part block diagrams showing still other embodiments of the present invention, FIG. 7 is a timing chart regarding the embodiment shown in FIG. 6, and FIG. 9 is a diagram of the embodiment shown in FIG. 8. It is a graph explaining the operation. 1... Key switch circuit, 2... Address generator, 3... Tone selector, 4... Memory selector, 5... Adder, 6, 12... Multiplier,
7... Envelope generator, 8... Digital to analog converter, 9... Sound system,
10... Counter, 13... Envelope generator circuit, 14... Accumulator, 15...
Registers, SGM1 to SGMn...spectrum group memory.

Claims (1)

[Scope of Claims] 1. A plurality of spectral group memories each storing a plurality of waveform signals obtained by synthesizing frequency components corresponding to spectral groups of a plurality of specific frequency bands corresponding to specific formants of musical instrument sounds; reading means for reading out the waveform signal from the group memory; timbre instruction means for instructing the timbre of the musical waveform to be synthesized and outputting timbre instruction information corresponding to the timbre; and the spectral group memory according to the timbre instruction information. spectral group selection means for selecting a combination of at least two memories from among the spectral group memories, and for commonly selecting at least some of the spectral group memories among the spectral group memories for different timbre instruction information; , a musical tone that forms a musical waveform signal having a frequency spectrum corresponding to the timbre specified by the timbre instruction means by synthesizing waveform signals read out by the reading means from the spectral group memory selected by the spectral group selection means; A musical sound waveform synthesis device comprising a forming means. 2. A plurality of spectral group memories each storing a plurality of waveform signals synthesized from frequency components corresponding to spectral groups of a plurality of specific frequency bands corresponding to specific formants of musical instrument sounds; and a plurality of spectral group memories that store the waveform signals from the spectral group memories. reading means for reading out the timbre; timbre instructing means for instructing the timbre of the musical waveform to be synthesized and outputting timbre instruction information corresponding to the timbre; means for changing the timbre instruction information over time; A combination of at least two memories is selected from among the spectral group memories according to the timbre instruction information, and at least some of the spectral group memories are selected from among the spectral group memories for different timbre instruction information. and a frequency corresponding to the timbre designated by the timbre instructing means by synthesizing the waveform signals read by the reading means from the spectral group memory selected by the spectral group selecting means. A musical sound waveform synthesis device comprising musical tone forming means for forming a musical sound waveform signal having a spectrum and whose tone color changes over time. 3. The music waveform according to claim 2, wherein the means for changing the timbre instruction information over time changes the timbre instruction information in response to a change in the amplitude envelope given to the music waveform signal. Synthesizer. 4 A plurality of spectral group memories each storing a plurality of waveform signals synthesized from frequency components corresponding to spectral groups of a plurality of specific frequency bands corresponding to specific formants of musical instrument sounds, and a plurality of spectral group memories that store the waveform signals from the spectral group memories a reading means for reading out the timbre of the musical waveform to be synthesized, a timbre instructing means for instructing the timbre of the musical waveform to be synthesized and outputting timbre instruction information corresponding to the timbre, and an envelope controlling the envelope of each waveform signal read out from the spectral group memory. a control means for selecting a combination of at least two memories from among the spectral group memories according to the timbre instruction information, and selecting a combination of at least two memories from among the spectral group memories for different timbre instruction information; a spectral group selection means for commonly selecting a spectral group memory of the section, and a waveform signal read out by the readout means from the spectral group memory selected by the spectral group selection means and subjected to envelope control by the envelope control means. and musical tone forming means for forming a musical sound waveform signal having a frequency spectrum corresponding to the tone specified by the tone color specifying means and whose tone color changes over time. 5. The reading means reads out the waveform signals from the spectrum group memory in a time-division manner, and the envelope control means performs envelope control on each waveform signal in a time-division manner. Device.