JPH08221076A - Musical sound controller - Google Patents

Abstract

PURPOSE: To perform filter processing about the frequency band of a musical sound waveform, to change wholly the change of a frequency characteristic of a musical sound, to shift the musical sound on a frequency without changing the density of respective frequency components of the musical sound waveform, to control a characteristic of a filter and to perform the filter processing independent of a pitch. CONSTITUTION: The musical sound waveform data TWj(t) is limited by a band control filter A06 in a band, and is frequency shifted to a transit band of a formant control filter A20 by a first frequency shift part A10, and the change of the frequency characteristic is performed with filter processing gradually changing from higher harmonics toward a basic wave from the basic wave toward the higher harmonics in the transit band of the formant control filter A20, and is shifted to the frequency according to the pitch by a second frequency shift part A30, and by the multiplication of cosωsj(t), sinωsj(t), the frequency band of the musical sound waveform data TWj(t)=ΣAn×cosωn(t)+ΣBn×sinωn(t) is shifted by -ωsj (low-pass) or ωsj (high-pass).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tone control device,
In particular, it relates to improvement of filter control of musical sounds and frequency shift and filter control of musical sounds.

[0002]

2. Description of the Related Art Conventionally, in the field of musical sound filter control, the pass band and stop band of the characteristics of the filter have been mainly used. Then, the cutoff frequency between the pass band and the stop band is adjusted to the upper limit or the lower limit of the frequency region to be cut.

As a result, of all the frequency components of the musical sound,
Only the frequency components above the cutoff frequency were cut, or only the frequency components below the cutoff frequency were cut. As a result, the harmonic components of the musical tone are cut, and conversely, the subharmonic components are also cut, resulting in changes in the tone color of the musical tone.

[0004]

However, in such filter control, only a part of the frequency characteristic of the musical sound changes, and it is not possible to realize an overall change in frequency.

An object of the present invention is to solve the above-mentioned problems, and a filter control having a characteristic completely different from that of the filter control focusing on the pass band and the stop band of such a filter is provided. Is to be realized.

Further, the frequency band of the musical tone to be filtered is greatly changed depending on the pitch of the musical tone. For example, the tone frequency of the note name A4 is 440 Hz, and the tone frequency of the note name A5 one octave higher is 880 Hz. Therefore, if the pitch of the tone to be filtered is changed, the characteristics such as the cutoff frequency of the filter must be changed accordingly.

An object of the present invention is to solve the above-mentioned problems, and to realize filter characteristics in a stable manner by performing filter control regardless of pitch.

[0008]

To achieve the above object, the present invention filters almost all frequency bands of a musical tone waveform only in a transition band between a pass band and a stop band of a filter. It is a thing. Further, according to the present invention, the frequency band of the musical tone waveform is shifted on the frequency without changing the density of each frequency component of this frequency band, is filtered, and is further shifted to the frequency according to the pitch. It is done.

[0009]

As a result, since the attenuation characteristic gradually changes in the transient band of the filter, the change in the frequency characteristic of the musical tone to be filtered is changed from the fundamental wave to the harmonic wave or from the harmonic wave to the fundamental wave. And change as a whole. Therefore, only a part of the frequency characteristic of the musical sound does not change, and it is possible to realize an overall change in frequency, and it is possible to perform a musical sound control that has never been achieved.

Further, this makes it possible to select a specific portion of the filter characteristic and perform the filter control only at this portion, so that the filter characteristic can be stably realized.
Further, since the frequency shift according to the pitch is performed after the filter processing, the filter processing can be performed regardless of the pitch. The characteristics of the filter can be changed.

[0011]

【Example】

<< 1 >> Overall Circuit FIG. 1 shows an overall circuit of the musical sound generating apparatus. Performance information generator 10 generates pitch information and other performance information. The performance information generation unit 10 is a sounding instruction device, an automatic performance device or an interface which is played by a manual operation. From the performance information generation unit 10, the performance information, that is, pitch information (range information (upper keyboard, upper keyboard, Lower keyboard,
Musical factor information such as footnote keyboard)), elapsed time information from the start of pronunciation, performance part information, musical sound part information, musical instrument part information, etc. is generated. The pronunciation instruction device is a keyboard instrument, a string instrument, a wind instrument, a percussion instrument, a computer keyboard, or the like. The automatic performance device automatically reproduces the stored performance information. Interface is MID
It is a device such as I (Musical Instrument Digital Interface) that receives and sends performance information from a connected device.

Further, the performance information generating section 10 is provided with various switches, and these various switches include a tone color tablet, an effect switch, a rhythm switch, a pedal, a wheel, a lever, a dial, a handle, a touch switch, etc. for musical instruments. belongs to. Musical factor information is input from these various switches, and the musical factor information includes tone color information and touch information (speed of sounding instruction operation /
Strength), effect information, rhythm information, sound image (stereo) information, quantize information, modulation information, tempo information,
Volume information, formant characteristic information, envelope information,
It is the time elapsed from the start of pronunciation.

The musical factor information is also included in the performance information, is input from the various switches, is included in the automatic performance information, and is included in the performance information transmitted and received by the interface. The touch switch is provided corresponding to each of the sounding instruction devices, and generates initial touch data and after-touch data indicating the speed and strength of the touch. The tone information includes keyboard instruments (piano, etc.), wind instruments (flute, etc.),
It corresponds to the sound of stringed instruments (such as violin) and percussion instruments (such as drums). The envelope information is an envelope level, an envelope phase, etc. The performance part information, the musical tone part information, and the musical instrument part information correspond to, for example, a melody, an accompaniment, a chord, a bass, etc., or an upper keyboard, a lower keyboard, a foot keyboard, etc. Such musical factor information is sent to the controller 20, and various signals, data, and parameters to be described later are switched.

The performance information is processed by the controller 20, and various data is converted into formant control parameter generating section 4.
0, which is sent to the formant waveform generator 50 and the accumulator 70 to generate a formant composite signal Wj (t). The controller 20 is composed of a CPU and the like. The program / data storage unit 21 includes a storage device such as a ROM and a RAM.
Programs for the controller 20 to perform various processes, the above-described various data, and other various data are stored. These various data also include data necessary for time-division processing and data for allocation to time-division channels.

Formant control parameter generator 40,
The formant waveform generation unit 50 and the formant waveform control unit 60 generate the formant composite signal Wj (t) in a time division manner. “J” of Wj (t) indicates the division order or channel number of the time division processing. From the formant control parameter generator 40, various parameters required to generate the formant composite signal Wj (t), that is, formant control parameters ωcj (t) and ωf.
j (t), aj (t), cj (t), dj (t), etc. are generated.

Details of this parameter are described in Japanese Patent Application No. 6-231.
No. 817 and the drawings. Formant waveform generator 50 and formant waveform controller 6
At 0, the formant synthesis signal Wj (t) is read out, generated, and synthesized based on the input formant control parameter. This formant composite signal Wj
In (t), various controls are performed by the shift filter unit A0, and the accumulation unit 70 accumulates and synthesizes each series channel.
The sound output unit 80 outputs the sound as a musical sound. This series indicates one musical tone of the formant synthesized signal Wj (t) which is a partial tone, or the musical factor described above.

From the timing generator 30, a timing control signal for synchronizing all the circuits of the musical tone generating apparatus is output to each circuit. The timing control signal includes a clock signal of each cycle, a signal obtained by performing a logical product or a logical sum of these clock signals, a signal having a channel division time cycle of time division processing, channel number data j, and the like.

Each frequency number data FN (pitch information) read out from the assignment memory 213 of the program / data storage unit 21 by the controller 20 in a time division manner, or the formant density from the formant control parameter generation unit 40. Parameter ωfj (t) or formant carrier parameter ωcj (t)
Are sent to the harmonic control circuit 90.

This data FN (ωfj (t), ωcj
(T) is a harmonics control circuit 90, in which the formant carrier parameter ωcj (t) from the formant control parameter generator 40, the sampling correction data Sfj (t) from the controller 20, and the synthesized formant harmonics Hj (t) are shown. The formant density ωf
It is sent to the parameter formant shape waveform generation unit 50 as j (t).

By this synthesis, the formant shape signal Ff
The contrast value of each frequency of each frequency component of the formants of j (t) and Fj (t) is determined, and the harmonic degree of each frequency component of the formant is controlled. In this case, the frequency number data FN is directly used as the harmonics control circuit 9
It is sent to 0 or arithmetically processed and sent to the harmony control circuit 90. This calculation is a calculation with other data, and is various calculation (1) described later.

Each circuit 10, 20, 21, 30, 4 in FIG.
Descriptions of specific configurations and operations of 0, 50, 60, 70, 80, 90 are all described in the specification and drawings of Japanese Patent Application No. 6-231817. Therefore, a detailed description of these circuits will be referred to the specification of Japanese Patent Application No. 6-231817, and will not be described in the present specification and the drawings. The entire description of the specification and drawings of this Japanese Patent Application No. 6-231817 is considered to be entirely described in the present specification.

In the shift filter section A0, the input formant synthesized signal Wj (t) is subjected to shift processing and filter processing of the frequency band of the musical tone waveform data TWj (t), and the frequency characteristic is changed. Is output. In another example, this shift filter unit A0
Is provided between the multiplier 66 in the formant waveform controller 60 and the formant shape waveform generator 50 (or the multiplier 652).

<< 2 >> Shift Filter Unit A0 FIG. 2 shows the shift filter unit A0. Frequency modulation information is added to the pitch information (key number KN) from the performance information generating section 10 by an adder A01. The frequency modulation information is based on information such as vibrato in the musical effect information input from the performance information generating section 10.

The pitch information and the like from the adder A01 is converted into frequency number data FN in the frequency number table A03. The frequency number data FN is accumulated by the accumulator A04 in time division for each channel and sent to the tone waveform memory A05. The musical tone waveform memory A05 stores a large number of musical tone waveform data TWj (t) in tone color, tone range,
Multi-layered storage is performed for each musical factor such as touch, each sound generation elapsed time, each envelope level / phase, and each operator-selected data, and data corresponding to these musical factors and the like are read out. The reading speed of the tone waveform data TWj (t) does not have to depend on the pitch.

A part or all of these musical tone waveform data TWj (t) is a plurality of partial musical tone waveforms whose frequency bands do not substantially overlap or partially overlap in some cases, and after passing through a second frequency shift unit A30 described later. The accumulator 70 synthesizes and outputs one musical tone. Therefore, these partial tone waveforms can be originally obtained by dividing one tone waveform into a plurality of frequency bands by filtering. These stored musical tone waveform data TWj
The center frequency of each frequency band of (t) can be a virtual frequency "0" as described later. The tone waveform memory A05 may be detachable from the tone generating device, and may be a CD-ROM / RAM or ROM / RAM card or the like.

The tone waveform data TWj (t) read out from the tone waveform memory A05 in a time division manner is sent to the band control filter A06 after each sampling point of the waveform is interpolated by an interpolation circuit (not shown). To be This interpolation circuit is the same as the interpolation circuit AB3 of FIG. 13 described later.
The tone waveform data TWj (t) is the formant synthesis signal Wj from the formant waveform controller 60 described above.
(T), the formant shape signals Ffj (t), Fj (t) or the formant carrier signals Gfj (t), cosωcj (t) described above. Hereinafter, these signals will be generically referred to as musical tone waveform data TWj (t).

The band control filter A06 is a digital filter. In the band control filter A06, as shown in FIG. 3 to be described later, only the frequency band having a bandwidth BW above and below the center frequency of the musical tone waveform data TWj (t). Are extracted and other frequency bands are cut. This allows
Almost all the frequency bands of this tone waveform data TWj (t) filtered by the formant control filter A20 described later are limited only to the transient band of the formant control filter A20. Of course, the predetermined width may have any value as long as the transient band of the formant control filter A20 is wide. The band control filter A06 is a band pass filter, but it may be a high pass filter or a low pass filter, but a band pass filter is preferable.

The musical tone waveform data TWj to be input
Since the frequency band of (t) is narrow or the transient band of the formant control filter A20 is wide, the musical tone waveform data TW is generated.
If almost all the frequency band of j (t) falls within the transient band of the formant control filter A20, the band control filter A06 can be omitted. In this case, the musical tone waveform data TWj (t) passed through the band control filter A06 can be stored in the musical tone waveform memory A05.

If the musical tone waveform data TWj (t) is a partial musical tone waveform, only the narrow area of the transient band of the formant control filter A20 can be used.
Then, even if the actual frequency characteristic of the formant control filter A20 is non-linear as shown in FIG. 9, the linear frequency characteristic shown in FIG. 5 can be obtained as a result.

The tone waveform data TWj (t) from the band control filter A06 is shifted in the entire frequency band by the first frequency shift unit A10. By this shift, the frequency band of the musical tone waveform data TWj (t) is shifted to the transient band of the formant control filter A20 as shown in FIG.

This frequency-shifted tone waveform data T
Wj (t) is filtered by the formant control filter A20 via the multiplier A41 so that the variation of the frequency characteristic gradually changes from the fundamental wave to the harmonic wave or from the harmonic wave to the fundamental wave. It

The filtered musical tone waveform data T
Wj (t) is further shifted in the entire frequency band by the second frequency shift unit A30. By this shift,
As shown in FIG. 4, the frequency band of the musical tone waveform data TWj (t) is shifted to a position corresponding to the pitch. The frequency-shifted tone waveform data TWj (t) is output to the accumulator 70. In this case, the envelope level data (formant control parameters aj (t), ajk (t)) can be multiplied and combined via a multiplier. The direction of the first shift and the direction of the second shift are the specified pitch and the formant control filter A20.
May be the same or different depending on the frequency position of the transient band of.

Frequency shift data FS generated in time division
For j (t), the adder A42 adds the shift control data SC in a time division manner, and the adder A44 adds the envelope data. Frequency shift data F from adder A44
Sj (t) is converted into a linear value by the logarithmic-linear conversion circuit A45 and sent to the first frequency shift unit A10. As a result, the above-mentioned frequency shift data FSj (t)
The frequency of the musical tone waveform data TWj (t) is shifted according to the value of.

The envelope speed data and the envelope target data are supplied to the envelope generator A46 in a time division manner, whereby the envelope data is generated in a time division manner. The envelope level data (formant control parameters aj (t), ajk (t)) may be substituted for this envelope data.

This frequency shift data FSj (t) is
It is a value corresponding to the difference between the center frequency of the frequency band of the musical tone waveform data TWj (t) and the center frequency of the portion of the formant control filter A20 used for the filtering process of the transition band. The center frequency of the musical tone waveform data TWj (t) is determined according to the frequency at the time of storing the musical tone waveform data TWj (t) and each ratio of the stored sampling frequency and the read frequency, or of the musical tone waveform data TWj (t). It is the same as the generation frequency.

The generated frequency data GFj (t) is data corresponding to the ratio of the stored frequency of the musical tone waveform data TWj (t) to the stored sampling frequency, and the generated frequency data GFj (t) is the adder A47. Then, the above pitch information is added to obtain a value corresponding to the actual pitch. The generated frequency data GFj (t) is used as the musical tone waveform memory A.
05 represents the center frequency of each frequency band of the tone waveform data TWj (t) stored in 05. If the center frequency of each frequency band is a virtual frequency “0”, the generated frequency data GFj (t ) May also be "0".

The generated frequency data GFj (t) + pitch information is converted into a linear value by a logarithmic-linear conversion circuit A48, and the subtractor A49 converts the frequency shift data FSj (t) + shift control data SCj (t). ) + Envelope data is subtracted, and the second frequency shift unit A3
Sent to 0. As a result, as shown in FIG. 4, the frequency shift amount in the first frequency shift unit A10 is subtracted from the frequency shift amount of the musical tone waveform data TWj (t) to the original pitch, and only the remaining frequency is changed to the second frequency. Shift part A
A frequency shift is performed at 30.

The pitch information from the adder A01 is added with the value of the bandwidth BW in the adder A50 and converted into a linear value in the logarithmic-linear conversion circuit A51. As a result, the band passes through the band control filter A06 and the band width is ± BW.
Upper limit frequency value fj (t) + and lower limit frequency value fj (t) of the frequency band of the musical tone waveform data TWj (t) limited to
-And are required. The frequency value fj (t) is subtracted from the frequency shift data FSj (t) by the subtractor A52, and the data shifter (multiplier) A5 is used.
The data from the subtractor A52 is doubled by 3 and the data from the subtractor A52 is added by the adder, and the correction is performed according to the frequency shift in the first frequency shift unit A10.

The upper limit frequency value fj (t) + and the lower limit frequency value fj (t)-are supplied to the filter gain table A55 as read address data, and each frequency value f.
Gain data gj (t) +, gj according to j (t)
(T)-is read. Two of the same filter gain tables A55 are provided in parallel.
One input latch, multiplexer, one filter gain table A55, demultiplexer, and two output latches may be substituted for the time division processing.

Needless to say, the gain data of the filter gain table A55 shows the frequency characteristic of the formant control filter A20 and is obtained by calculation from the filter calculation parameter of the formant control filter A20.

As shown in FIG. 5, the lower end gain data gj
(T)-is the link data Linkj in the adder A56.
(T) is added, converted into a linear value by the logarithmic-linear conversion circuit A59, and sent to the multiplier A41. On the other hand, for the upper end gain data gj (t) +, the lower end gain data gj (t)-is subtracted by the subtractor A57, and the link data Linkj is added by the adder A58.
(T) is added and new link data Linkj
(T). This link data Linkj
(T) is stored in the latch A70, and the next link data L
It is input to the adder A56 as inkj + 1 (t).

The latch A70 is cleared by the trigger signal from the timing control section 30 having a constant cycle. The cycle of this trigger signal is equal to the division time for four channels if there are four partial sounds per musical sound. The multiplier A41 may be provided on the output side of the formant control filter A20.

Such a filter gain table A55
The reason why the multiplier A41 is provided is as follows. The above-mentioned musical tone waveform data TWj (t)
Is a partial sound existing in plural as described above for one pronunciation instruction or one musical tone. In the shift filter unit A0 of FIG. 1, the musical tone waveform data TWj which is each partial sound is generated.
For (t), tone control processing such as filter processing is performed in the transition band of the slope shown in FIG. Therefore, in the filter processing for each partial sound in the formant control filter A20, it is necessary to match the gains at the boundaries of the frequency bands of each partial sound as shown in FIG.

Therefore, the filter gain table A5
5, a multiplier A41, adders A56 and A58, a subtractor A57 and the like are provided. If the musical tone waveform data TWj (t) is not divided into partial sounds, these circuits A55, A41, ... Are unnecessary. In particular, this is not necessary when the transient band of the formant control filter A20 is extremely wide and the entire frequency band of the tone waveform data TWj (t) is included. When the pitch information (key number KN) from the adder A50 or the bandwidth BW changes, the gain at the boundary of the partial tone waveform also changes, and the gain matching is performed accordingly.

Further, the musical tone waveform data TWj (t)
Is a signal having a constant frequency regardless of the pitch, such as the formant shape signals Ffj (t) and Fj (t),
Pitch information (key number K
Only N) can be input to the log-linear conversion circuit A48. In this case, the shift filter unit A0 of FIG. 2 is provided between the multiplier 66 in the formant waveform control unit 60 and the formant waveform generation unit 50 (or the multiplier 652). The multipliers 66 and 652 are described in the specification and drawings of Japanese Patent Application No. 6-231817.

As described above, it is possible to generate a musical tone corresponding to the pitch by frequency shifting. The musical tone corresponding to the pitch has a left-right asymmetric formant shape as shown by F8 in FIG. 15 instead of the left-right symmetrical formant shape as shown in FIG. The above logarithm-
The addition in the adder and the subtraction in the subtractor on the input side of the linear conversion circuits A45, A48, A51, A59 are actually multiplications and divisions by log-linear conversion. Of course, the logarithmic-linear conversion circuit may be omitted and the adder and the subtractor may be replaced by a multiplier or the like. Further, a plurality of shift / filter units A0 may be provided and the time division processing may be omitted. In this case, the link data Linkj (t) is sent from the circuit A0 having a lower frequency band of the partial tone waveform to the circuit A0 having a higher frequency band.

<< 3 >> Formant Control Filter A20 FIG. 6 shows an example of the formant control filter A20. This filter is an FIR type digital filter that performs a convolution operation. Each delay device A71 ... Is composed of, for example, a CCD or BBD, and the output of each tap becomes the output of each delay device A71. The output H of each delay device A71 ...
1, H2, H3, ... are multipliers A72, respectively
Are multiplied by the multiplication data A1, A2, A3, ... And added and combined by the adder A73 to be output.

The delay time of each delay device A71 ... Is equal to the period Ts of the sampling frequency fs, and this sampling signal φs1 is output from the timing generator 30, the programmable counter, the programmable oscillator or the like to each delay device A71.
71 ... (CCD). The sampling frequency data fs1 (Ts1) is input to the programmable oscillator (or programmable counter) A74. Then, the sampling signal φs1 having a frequency corresponding to this is input to each of the delay devices A71 ..., By which the cutoff frequency is determined. The cutoff frequency is also changed and determined by the filter coefficient data A1, A2, A3, ....

FIG. 7 shows the formant control filter A2.
2 shows a flowchart when 0 is realized by a DSP (digital signal processor) or a microcomputer. In this filtering process, the 1st to nth delay data H1 to Hn are multiplied by the filter coefficients A1 to Am, and the product sum of these multiplied data and the input musical tone waveform data TWj (t) is obtained and output (step 2). .

Then, the data H1 to Hn in the registers of the RAM in the DSP are sequentially shifted from the nth delay data Hn to the delay data of the next higher order (step 4).
~ 8). Finally, the input tone waveform data TWj (t) is transferred to the first-order delay data H1 (step 10). The above process is repeated by the interrupt process at the cycle Ts of the sampling frequency fs.

FIG. 8 shows the formant control filter A2.
A frequency characteristic of 0 is shown. This characteristic has a pass band from frequency 0 to near the cutoff frequency, a stop band after the cutoff frequency, and a transient band near the cutoff frequency between the pass band and the stop band. Due to the frequency shift by the first frequency shift unit A10, almost all the frequency bands of the musical tone waveform data TWj (t) fall within this transient band and the filtering process is performed.

In this transient band, since the attenuation characteristic gradually changes, the tone waveform data TWj to be filtered is controlled.
The amount of change in the frequency characteristic of (t) gradually changes from the fundamental wave to the harmonic wave or from the harmonic wave to the fundamental wave as a whole. Therefore, the musical tone waveform data TWj
Only part of the frequency characteristic of (t) does not change, and it is possible to realize an overall change in frequency.

The formant control filter A2
The frequency characteristic of 0 has a pass band even near the frequency that is an integral multiple of the sampling frequency fs, and similarly has a stop band and a transient band. Therefore, although the low-pass filter is used in the above example, it can be used as the transient band of the high-pass filter if the area to be filtered is selected from another transient band.

FIG. 9 shows another frequency characteristic of the formant control filter A20. In this case, the formant control filter A20 is composed of a plurality of filters, the musical tone waveform data TWj (t) is input in parallel to the plurality of filters, and the outputs of the respective filters are added and combined by an adder. The formant control filter A20 may use the filter calculation means or the filter processing means shown in the specification and drawings of JP-A-3-177898, and all of the specifications and drawings of this Japanese Patent Application No. 3-177898. The content of the description is entirely described in the present specification. The frequency characteristic of FIG. 9 may be a symmetrical frequency characteristic of this frequency characteristic and may be the frequency characteristic of a high pass filter.

FIG. 10 shows the shift filter table A90.
Indicates. The shift control data SCj (t), frequency shift data FSj (t), sampling frequency data f
s1 (Ts1), filter coefficient data A1, A2, A
, ..., Envelope speed data, envelope target data, generated frequency data GFj (t), and frequency modulation information such as vibrato in musical effect information are described in the specification and drawings of Japanese Patent Application No. 6-231817. Data SP, O, Min, Ta, Ea, formant shape signal Ffj (t), formant density parameter ωfj
(T), formant carrier parameter ωcj
(T), n sets of parameters ωcjk (t), ajk
(T), cj (t), formant shape table 212
Alternatively, like the storage of the formant center table 214, it is stored in the shift filter table A85 in multiple layers for each musical factor, for each elapsed time from the start of sounding, each envelope level or each envelope phase.

This musical factor is output from the performance information generator 10 as described above, and the elapsed time from the start of sound generation is the formant control parameter Valj as described above.
(In some cases, the accumulated formant density parameter Σω
fj (t) or accumulated formant carrier parameter Σωcj (t)) or time count data is used, envelope level data is the formant control parameter alj (t), and the envelope phase is based on the count number of request data Req. . This request data Req can be found in Japanese Patent Application No. 6-231817.
And the drawings.

Data such as these musical factors are supplied to the above table as upper read address data. In addition, the data S for each musical factor etc.
Cj (t), FSj (t), fs1 (Ts1), GFj
(T), A1, A2 ..., Envelope speed data,
Information such as envelope level data and vibrato is also selectively read by selection data (higher-order read address data) input by the operator from the panel switch group of the performance information generator 10. Further, the data SCj (t), FSj (t), fs1 (Ts1), GF
Frequency modulation information such as j (t), A1, A2, ..., Envelope speed data, envelope target data, and vibrato may be input from the performance information generator 10 by the operator.

The musical factors include the formant control parameter Valj, which changes according to the envelope information or changes with time,
The time count data and the like may be modified and combined by various calculations (1) described later.

The storage for each elapsed time from the start of sound generation or each envelope level is omitted, and each data SCj (t), FSj (t), fs1 (Ts1), G is stored.
For Fj (t), A1, A2, etc., the elapsed time from the start of sound generation or the envelope level may be modified and combined. This correction synthesis is based on various calculations (1) described later, and the output end of the table is provided with a calculation device for correcting synthesis of the sound generation elapsed time or the envelope level.

The data SCj (t), FSj (t), fs read from the table as described above.
1 (Ts1), GFj (t), A1, A2, etc. are written in each channel area of the assignment memory 213 described in the specification of Japanese Patent Application No. 6-231817 and FIG. 26 in the drawings. In rare cases, they may be sequentially read in time division and supplied to the shift filter unit A0 in FIG.

By the control as described above, the shift amount of the frequency band of the musical tone waveform data TWj (t) is changed according to the musical factor, the elapsed time from the start of sound generation or the envelope level / phase. Further, by changing control of the sampling frequency data fs1 (Ts1) and the filter coefficient data A1, A2, A3, ..., The frequency characteristics and the cutoff frequency of the formant control filter A20 are changed, whereby the formant control filter A20 is changed. The transient band itself is also shifted, and as a result, the portion (area) in the transient band where the tone waveform data TWj (t) is filter-controlled is changed.

<< 4 >> First Frequency Shift Unit A10 (Second Frequency Shift Unit A30) FIG. 11 shows the first frequency shift unit A10 and the second frequency shift unit A30. The above frequency shift data FS
j (t) etc. or generated frequency data GFj (t) etc.
The accumulator A70 accumulates each channel in a time division manner and supplies the accumulated signals to the cosine table A60 and the sine table A61, respectively. Here, this frequency shift data FSj (t) or the like or generated frequency data GFj
If (t) and the like are ωsj, the shift data cosω is obtained from the cosine table A60 and the sine table A61.
sj (t) and sinωsj (t) are output. This shift data cosωsj (t), sinωsj (t)
Is supplied to the multipliers A62 and A63, multiplied by the musical tone waveform data TWj (t), and filtered by the filters A64 and A63.
After 65, the adder A64 performs addition synthesis.

The musical tone waveform data TWj (t) can be expressed as follows based on the principle of Fourier analysis.

TWj (t) = ΣAn × cosωn (t) + ΣBn × sinωn (t) (A1) Σ is a symbol indicating accumulation from n = 1 to n = m (m is an arbitrary number). . The shift data cosωsj
Multiplying (t) and sinωsj (t) gives the following.

TWj (t) × cosωsj (t) = {ΣAn × cosωn (t) + ΣBn × sinωn (t)} × cosωsj (t) = 1/2 × ΣAn × cos {(ωn + ωsj) (t)} + 1 / 2 × ΣBn × sin {(ωn + ωsj) (t)} + 1/2 × ΣAn × cos {(ωn−ωsj) (t)} + 1/2 × ΣBn × sin {(ωn−ωsj) (t)} (A2 ) TWj (t) × sin ωsj (t) = {ΣAn × cosωn (t) + ΣBn × sin ωn (t)} × sin ωsj (t) = 1/2 × ΣAn × sin {(ωn + ωsj) (t)} −1/2 × ΣBn × cos {(ωn + ωsj) (t)} −1 / 2 × ΣAn × sin {(ωn−ωsj) (t)} + 1/2 × ΣBn × cos {(ωn−ωsj) (t)} (A3 ) These two are passed through a low pass filter and cos {(ωn ω
The components of sj) (t)} and sin {(ωn + ωsj) (t)} are cut as follows.

Icj (t) = 1/2 × ΣAn × cos {(ωn−ωsj) (t)} + 1/2 × ΣBn × sin {(ωn−ωsj) (t)} (A4) Isj (t) = 1/2 × ΣBn × cos {(ωn−ωsj) (t)} −1 / 2 × ΣAn × sin {(ωn−ωsj) (t)} (A5) When these two are added and combined, the following is obtained. become.

Iwj (t) − = (A6) 1/2 × (ΣAn + ΣBn) × cos {(ωn−ωsj) (t)} + 1/2 × (ΣBn−ΣAn) × sin {(ωn−ωsj) ( t)} This is the original tone waveform data TWj except for the amplitude.
The frequency band of (t), that is, all frequency components are similar-
This means that only ωsj has been shifted.

Further, the above two are passed through a high-pass filter and cos {(ωn-ωsj) (t)} and sin {(ωn-
The following is obtained by cutting the components with ωsj) (t)} and adding and combining them.

Iwj (t) + = ... (A7) 1/2 × (ΣAn−ΣBn) × cos {(ωn + ωsj) (t)} + 1/2 × (ΣAn + ΣBn) × sin {(ωn + ωsj) (t)} Also, the original musical tone waveform data TWj except for the amplitude
The frequency band of (t), that is, all frequency components are exactly +
This means that only ωsj has been shifted. These ± ωsj
The frequency shift does not change the density of each frequency component in the frequency band of the musical tone waveform data TWj (t). But,
The overtone ratio of each frequency component of the frequency band changes and the timbre (sound quality) changes subtly.

When the frequency shift of the tone waveform data TWj (t) is set to -ωsj, the filters A64 and A are used.
65 is a low-pass filter with a frequency shift of + ω
When it is set to sj, it becomes a high-pass filter. The cutoff frequency of the low-pass filter and the high-pass filter is ωsj. Note that normally, the shift angular frequency ωsj
Is the frequency bandwidth ωn of the tone waveform data TWj (t).
A value larger than (BW), for example, almost the same value, doubled value, 3
Although it is a multiple value, ..., A value smaller than the frequency bandwidth ωn is also possible. The filters A64 and A65 are realized by the circuit of FIG. 6 or the process of FIG. 7, for example.

In response to this, the value of the bandwidth BW is multiplied by 1 or 2 or 3 by the multiplier (data shifter) A68, the shift amount ωsj is added by the adder A69, and the programmable oscillator ( Alternatively, it is input to a programmable counter) A67. Then, the sampling signal φs2 having a frequency corresponding to this is transmitted to the filter A6.
4 and A65, which determines the cutoff frequency. The cutoff frequency is the filter coefficient data (A0) of the filters A64 and A65, A
, A1, A2, A3, ..., B1, B2, B3 ,.

Also, the multiplier (data shifter) A
63 may be omitted. Further, the filter A6
Similarly to the filter coefficient data A1, A2, A3, ..., The filter coefficient data A4, A65 are also stored in the shift filter table A85 in a multi-layered manner, and similarly input, corrected, combined, changed, etc. be able to.

The sine table A60, the multiplier A62 and the filter A64 in FIG. 11 may be omitted, or the cosine table A61 and the multiplier A may be omitted.
63 and the filter A65 may be omitted. This also allows frequency shifting.

FIG. 12 shows the musical tone waveform data TWj.
(T) is the formant shape signal Ffj (t), Fj
When it is (t), the state of frequency shift by the first frequency shift unit A10 (second frequency shift unit A30) is shown. Formant shape signals Ffj (t) and Fj (t)
And formant carrier signal Gj (t) are combined by multiplier 66 to formant combined signal Wj.
(T) (tone signal) is output. Here, if the formant shapes of the formant shape signals Ffj (t) and Fj (t) are the shapes shown in FIG. 12 (1), the formant shape of the formant synthesis signal Wj (t) is shown in FIG.
It becomes (2).

Here again, the formant shape signal Ffj
When (t) and Fj (t) are frequency-shifted as shown in FIG. 12 (3), the formant synthesized signal Wj (t) is generated.
The formant shape is as shown in FIG. 12 (4). The frequency-shifted formant shape of FIG. 12 (4) and FIG.
The frequency component is different from that of the 2 (2) formant shape, and the overtone ratio of each frequency component of this frequency band is also different, so that the timbre (sound quality) is different. Therefore, the tone color (sound quality) can be changed by such frequency shift.
Since the amount of this frequency shift changes according to the musical factor, the pronunciation elapsed time, the envelope level / phase, etc. as described above, the change in the tone color (sound quality) due to this frequency shift also depends on the musical factor, the pronunciation elapsed time, the envelope. It changes according to the level / phase, etc.

<< 5 >> Shift Filter A0 (Second Embodiment) FIG. 13 shows a second embodiment of the shift filter A0. The shift filter unit A0 is provided between the multiplier 66 in the formant waveform control unit 60 and the formant waveform generation unit 50 (or multiplier 652). The inputted tone waveform data TWj (t) is a signal having a constant frequency regardless of the pitch, such as the formant shape signals Ffj (t), F.
j (t). However, the inputted tone waveform data TWj (t) is also a signal corresponding to the pitch, for example, the formant composite signal Wj (t), the formant carrier signals Gfj (t), Gj (t), and cosωcj (t). Good. In this case, the shift filter unit A0 is provided between the formant waveform control unit 60 and the accumulation unit 70. Except what is stated below,
The description of the filter unit A0 is referred to.

The frequency-shifted tone waveform data Icj (t) and Isj (t) from the filters A64 and A65 of the frequency shifter A10 (A30) are interpolated by the interpolation circuit AB.
3. Each sampling point of the waveform is interpolated by AB3, frequency-shifted by frequency shift circuits A91, A91, added and synthesized by an adder A92, envelope controlled by an envelope control circuit A93, and loudness controlled by a loudness control circuit A94. , Adder A95, and the resultant is output to the multiplier 66 or the accumulator 70 in the formant waveform controller 60. The envelope control circuit A92 or the loudness control circuit A9
3 can be omitted.

The musical tone waveform data Icj (t), Isj
(T) may be stored in the tone waveform memory A05 as described later. The interpolation circuit AB3 is disclosed in
-8924, JP-A-53-50722, JP-A-53
No. 107815, JP-A-63-98699, JP-A-2-126293, JP-A-2-240697, JP-A-3-204696 may be used, and the apparatus shown in the drawings may be used. All descriptions in the specification and drawings are entirely described in the present specification.

<< 6 >> Frequency Shift Unit A10 (A30) FIG. 14 shows the musical tone waveform data Icj (t) and Isj.
A circuit for generating (t) is shown. The circuit of FIG. 14 is the same as the first (second) frequency shift unit A10 (A30) of FIG. 11 described above except the adder A66. Therefore,
The first (second) frequency shift unit A10 (A30) in FIG.
However, the same operation as that shown in FIG. 14 is executed. The circuit of FIG. 14 shows the formant shape of the tone signal of each part. Other than what is stated below, reference is made to the above description of the circuit of FIG.

The formant shape of the musical tone waveform data TWj (t) is shown by F1 in FIG. The formant shape of the musical tone waveform data multiplied by cosωsj (t) in the multiplier A63 becomes a shape like F2. F
In the formant shape of 2, the negative frequency component also appears virtually. Also, with multiplier A62, sinωsj
The formant shape of the tone waveform data multiplied by (t) becomes a shape like F3. In the formant shape of F3, the negative frequency component and the negative component level also appear virtually.

In this case, the value of the central angular frequency ωc of the musical tone waveform data TWj (t) and cos ωsj (t) and sin.
It is the same as the value of the angular frequency shift amount ωs of ωsj (t). This causes frequency-shifted musical tone waveform data TW.
The center frequency of the frequency band of j (t) becomes zero. Of course, the value of ωc and the value of ωs may be different. Then, the formant shape of the musical tone waveform data Icj (t) and Isj (t) passed through the filters (low-pass) A65 and A64 becomes a shape like F4 and F5. As a result, one formant having a frequency near zero is selectively extracted from a plurality of formants having the same shape generated by the frequency shift.

Thus, due to the frequency shift, the musical tone waveform data TWj (t) of the formant like F1 becomes
It is converted into formant musical tone waveform data Icj (t) and Isj (t) such as F4 and F5 whose center frequency is "0". Therefore, the musical tone waveform data Icj (t), Isj
The stored sampling frequency of (t) is the tone waveform data T
Since the sampling frequency may be lower than the storage sampling frequency of Wj (t), the tone waveform data TWj (t) is compressed and stored.

This tone waveform data Icj (t), Isj
(T) Each or these two musical tone waveform data Ic
The tone waveform data in which j (t) and Isj (t) are additively synthesized or multiply synthesized is stored in the tone waveform memory A05 of FIG.
Is stored as the tone waveform data TWj (t).

The musical tone waveform data Icj (t), Isj
Part of (t) and TWj (t) is a plurality of partial tone waveforms whose frequency bands do not substantially overlap or partially overlap, and the second frequency shift unit A30 or an adder A9 described later is used.
After passing 2, the sound is synthesized and output as one musical sound. The tone waveform memory A05 may be detachable from the tone generating device, and may be a CD-ROM / RAM or ROM / RAM card or the like.

A large number of stored tone waveform data Ic are stored.
j (t) and Isj (t) are multi-layered, stored for each musical factor such as tone color, range, touch, etc., for each sounding elapsed time, for each envelope level / phase, and for each operator's selection data. Data corresponding to a musical factor or the like is read out.

The musical tone waveform data may be frequency-shifted by the first frequency shift section A10 (second frequency shift section A30) or may be filter-controlled by the formant control filter A20. Thereby, the musical tone waveform data Ic
Since the central angular frequency of j (t) and Isj (t) is “0”, the frequency shift process is simplified. Further, the frequency shift unit A10 (A30) of FIG. 14 may be provided on the input side of the interpolation circuit AB3.

<< 7 >> Frequency Shift Circuit A91 FIG. 15 shows the above frequency shift circuit A91 and adder A92.
Etc. In the multipliers A97 and A96, the read musical tone waveform data Icj (t) and Isj (t).
Is multiplied by cos ωrj (t) and sin ωrj (t), and the frequency shift of the angular frequency ωrj is performed. The formant shape of the musical tone waveform data Icj (t) and Isj (t) passed through the multipliers A97 and A96 is F.
6 and F7, the shape is frequency-shifted according to the angular frequency ωrj. The formant shapes of F6 and F7 also exist on the negative frequency side, but are omitted.

This frequency-shifted musical tone waveform data I
cj (t) and Isj (t) are added and synthesized by the adder A92, reproduced, and output. As a result, the positive frequency component and the negative frequency component of the musical tone waveform data Icj (t) and Isj (t) cancel each other out, and the formant of the synthesized musical tone waveform data becomes a shape as shown in F8, and the original musical tone is reproduced. Formant F of waveform data TWj (t)
1 is frequency-shifted by an angular frequency ωrj.

The addition-combined tone waveform data is envelope-controlled by an envelope control circuit A93 which is a multiplier, and is loudness-controlled by a loudness control circuit A94 which is also a multiplier, and another tone waveform data is added by the adder A95. Is added and combined. The envelope data sent to the envelope control circuit A93 and the loudness data sent to the loudness control circuit A94 are the data Valj (aj (t), cj (t) described in the specification and drawings of Japanese Patent Application No. 6-231817. ,
Either dj (t)) is used.

Further, the frequency shift circuit A9 shown in FIG.
Although not shown in FIG. 1, the accumulator A7 of FIG.
0, cosine table A60 and sine table A61
This cosine table A has the same circuit as
60 and sine table A61 to the multipliers A97 and A96, where cosωrj (t), sinω
rj (t) is input. The frequency shift data ωrj is input to the accumulator A70. The frequency shift data ωrj is the frequency shift data ωs.
It is generated exactly like j.

Therefore, this frequency shift amount ωrj
Changes according to a musical factor, a sound generation elapsed time, an envelope level / phase, an operator's setting instruction, and the like.
Further, both the envelope data sent to the envelope control circuit A93 and the loudness data sent to the loudness control circuit A94 change according to the musical factor, the sounding elapsed time, the envelope level / phase, the setting instruction by the operator, and the like.

Further, the frequency shift amount ωrj can be set in accordance with the pitch (key number KN) of the generated musical sound. As a result, a musical tone corresponding to the pitch is generated as shown in FIG.
15 is not the symmetrical formant shape of FIG.
A left-right asymmetric formant shape can be realized. In this case, this frequency shift does not change the density of each frequency component in the frequency band of the musical tone waveform data TWj (t), and the formant width does not change. However, the overtone ratio of each frequency component of the frequency band changes and the timbre (sound quality) changes subtly. Also, the frequency shift amount ωrj during this reproduction
The value of is the same as the value of the frequency shift data ωsj at the time of storage, and may be plus or minus in reverse.

<< 8 >> Shift Filter Section A0 (Third Embodiment) FIG. 16 shows a third embodiment of the shift filter section A0. The shift filter unit A0 can be replaced with the shift filter unit A0 of the second embodiment shown in FIG. Therefore, the description of the same parts as those in the second embodiment will be omitted, but the description of the same parts is entirely described here. Except what is described below, the above description of the shift filter unit A0 is referred to.

Frequency-shifted or read-out musical tone waveform data Icj (t), Isj from the filters A64, A65 of the frequency shifter A10 (A30).
In (t), the sampling points of the waveform are interpolated by the interpolation circuits AB3 and AB3, and the frequency shift circuit AA
1, frequency-shifted by AA2, filter-controlled by the formant control filters A20, A20, frequency-shifted by frequency shift circuits AA3, AA4, additively synthesized by the adder A92, and envelope-controlled by the envelope control circuit A93. The loudness is controlled by the loudness control circuit A94, added and synthesized by the adder A95, and output to the multiplier 66 or the accumulator 70 in the formant waveform controller 60.

The envelope control circuit A92 or the loudness control circuit A93 can be omitted. The circuit for generating the musical tone waveform data Icj (t) and Isj (t) is the circuit shown in FIG. 15 or the first (second) frequency shift unit A10 (A30) of FIG. 11 excluding the adder A66.
Is the same as

<< 9 >> Frequency Shift Circuits AA1 to AA4 FIG. 17 shows the frequency shift circuits AA1 and AA2, the formant control filters A20 and A20, the frequency shift circuits AA3 and AA4, and the adder A92. Except what is described below, the above description of the formant control filter A20 is referred to. Multiplier AA5,
In AA6, cos is added to the musical tone waveform data Icj (t).
ωpj (t) and −sin ωpj (t) are multiplied, and the frequency shift of the angular frequency ωpj is performed. In the multipliers AA7 and AA8, the musical tone waveform data Isj (t) is obtained.
Is multiplied by cosωpj (t) and sinωpj (t), and the frequency shift of the angular frequency ωpj is performed.

The data from the multipliers AA5 and AA8 are added and combined by the adder AA9, and the data from the multipliers AA6 and AA7 are added and combined by the adder AB0. The formant shapes of the musical tone waveform data Icj (t) and Isj (t) that have passed through the adders AA9 and AB0 are frequency-shifted shapes corresponding to the angular frequency ωpj like F9 and FA. The formant frequency components on the negative frequency side cancel each other out.

This frequency-shifted tone waveform data I
cj (t) and Isj (t) are formant control filters A20 and A20, and the above filter control is performed. As a result, the formant frequency component is changed as shown in the formant shapes FB and FC. This change gradually changes overall from the fundamental wave to the harmonic wave or from the harmonic wave to the fundamental wave.

The tone waveform data from the formant control filter A20 is multiplied by cos ωqj (t) in the multiplier AB1 of the frequency shift circuit AA3, and the angular frequency ωqj is frequency-shifted. The tone waveform data from the other formant control filter A20 is the multiplier AB2 of the frequency shift circuit AA4.
Then, sin ωqj (t) is multiplied, and the frequency shift of the angular frequency ωqj is performed. These multiplier AB
The tone waveform data from 1 and AB2 are added and synthesized by an adder A92, reproduced and output.

As a result, the musical tone waveform data Icj is obtained.
Since the positive frequency component and the negative frequency component of (t) and Isj (t) cancel each other out, the formant of the synthesized musical tone waveform data becomes a shape as shown in FD, and the original musical tone waveform data TWj (t) of The formant F1 is filtered and frequency-shifted by the angular frequency ωpj + ωqj.

The frequency shift circuits AA1 to AA of FIG.
Although not shown in A4, the accumulator A7 of FIG.
0, cosine table A60 and sine table A61
This cosine table A has the same circuit as
60 and signature table A61 to the above multipliers AA5 to AA8, AB1 and AB2, to the above cosωpj
(T), ± sinωpj (t), cosωqj (t),
sinωqj (t) is input. The frequency shift data ωpj, ωqj is input to the accumulator A70. The frequency shift data ωpj and ωqj are generated in exactly the same manner as the frequency shift data ωsj.

Therefore, this frequency shift amount ωpj,
ωqj changes according to a musical factor, a sound generation elapsed time, an envelope level / phase, an operator's setting instruction, and the like. Further, the frequency shift amounts ωpj, ωqj can be set according to the pitch (key number KN) of the generated musical sound. As a result, a musical tone corresponding to the pitch is generated as shown in FIG.
17 is not a symmetrical formant shape, but the FD of FIG.
A left-right asymmetric formant shape can be realized. In this case, this frequency shift does not change the density of each frequency component in the frequency band of the musical tone waveform data TWj (t), and the formant width does not change. However, the overtone ratio of each frequency component of the frequency band changes and the timbre (sound quality) changes subtly. Also, the frequency shift amount ωpj during this reproduction
The value of + ωqj is the frequency shift data ωs at the time of storage.
It is the same as the value of j, and may be plus or minus.

<< 10 >> Formant control filter A20
(Second Embodiment) FIG. 18 shows a second embodiment of the formant control filter A20, the filters A64 and A65. Except what is described below, the above description of the formant control filter A20, the filter A64 or A65 is referred to. This filter is an IIR type digital filter that performs a convolution operation. Each delay device A71 ... Is composed of, for example, a CCD or BBD, and the output of each tap becomes the output of each delay device A71. Input musical tone waveform data TWj (t) passed through adder A76, output H of each delay device A71 ...
1, H2, H3, ... are multipliers A72, respectively
Are multiplied by the multiplication data A0, A1, A2, A3, ... And added and combined by the adder A73 to be output. Also,
The outputs H1, H2, H3, ... Of the delay devices A71 ... are multiplied by the multipliers A75.
.. are multiplied by B3, ... And added to the input musical tone waveform data TWj (t) by an adder A76.

The delay time of each delay device A71 is equal to the period Ts of the sampling frequency fs, and this sampling signal φs1 is output from the timing generator 30, the programmable counter, the programmable oscillator or the like to each delay device A71.
71 ... (CCD). The sampling frequency data fs (Ts) is input to a programmable oscillator (or a programmable counter) A74. Then, the sampling signal φs1 having a frequency corresponding to this is input to each of the delay devices A71 ..., By which the cutoff frequency is determined. The cutoff frequency is the filter coefficient data A0, A1, A2, A3, ...
It is also changed and determined by B1, B2, B3, ....

FIG. 19 shows the formant control filter A.
The flowchart when 20 is implement | achieved by DSP (digital signal processor) or a microcomputer is shown. In this filtering process, the 1st to nth delay data H1 to Hn are multiplied by the filter coefficients B1 to Bn, the sum of products including these multiplication data and the input musical tone waveform data TWj (t) is obtained, and the DSP is used as the current data H0. It is stored in the register of the internal RAM (step 12).

Next, the current data H0 and the 1st to nth delay data H1 to Hn are multiplied by the filter coefficients A0 to Am,
The sum of products of these multiplication data is obtained and output (step 14). Then, the data H0 to Hn in the register of the RAM in the DSP is sequentially shifted from the nth delay data Hn to the delay data of the next higher order (step 16).
~ 20). The above processing is the sampling frequency fs.
It is repeated in the interrupt process in the cycle Ts1 of 1.

The sampling frequency data fs1 (T
s1), filter coefficient data A0, A1, A2, ..., B
1, B2, ... Are the sampling frequency data fs described above.
1 (Ts1), the filter coefficient data A1, A2, ... Are stored in the shift filter table A85 in a multi-layered manner for each musical factor, elapsed time from the start of sounding, envelope level or envelope phase. Further, it may be selectively read out by the operator according to the selection data inputted from the panel switch group of the performance information generating section 10, and further inputted by the operator from the performance information generating section 10. Various calculations (1)
It is also possible to make a correction composition by the above.

<< 11 >> Formant control filter A20
(Third Embodiment) FIG. 20 shows a third embodiment of the formant control filter A20, the filters A64 and A65. Except what is described below, the above description of the formant control filter A20, the filter A64 or A65 is referred to. The tone waveform data TWj (t) is input to any one of filters A78 ... Through a demultiplexer A77, is subjected to filter control, and is output through a multiplexer A79. The filters A78 ... Are those shown in FIG. 6, FIG. 7, FIG. 18 or FIG. The position on the frequency of each transient band or the cutoff frequency of each filter A78 differs depending on the musical range (pitch).

Upper range data of the pitch information (key number data KN) from the adder A01 or higher range data of the frequency number data FN from the frequency number table A03 is selected by the demultiplexer A77 and the multiplexer A79 ( Switching) supplied as data. In this case, the first frequency shifter A10 and the subtractors A49 and A52 are omitted, the output from the filter gain table A55 is arithmetically modified according to the above range data (pitch information), and each transient of each filter A78 ... The correction is performed according to the position of the band on the frequency. Further, the frequency shift in the second frequency shift unit A30 is only the frequency shift according to the lower pitch name data of the pitch information, or the second frequency shift unit A30 is omitted.

The present invention is not limited to the above embodiments, and various modifications can be made without departing from the spirit of the present invention. For example, the musical tone waveform data Icj (t), Isj (t) from the filters A64, A65 of the frequency shift unit A10 (A30) of FIGS. 2, 11, and 14 to 17, the first frequency shift unit A10 or Second frequency shift unit A30
Tone waveform data TWj (t) from A, or multipliers A62, A63, A96, A97, AA5, AA
6, AA7, AA8, AB1, AB2, adder A92,
A95, AA9, AB0, formant control filter A
The musical tone waveform data generated from the respective locations of 20 may be temporarily stored in the musical tone waveform memory A05 for each musical factor, elapsed time from the start of sound generation, envelope level, envelope phase or setting instruction by the operator. Then, each of the generated and stored data is input to the circuit after the generation point.

In this case, the musical tone waveform data Icj (t), Isj (t), TWj (t) read out for each musical factor, elapsed time from the start of sounding, envelope level, envelope phase or setting instruction by the operator. Are switched or changed. Then, the musical factor may be combined with the formant control parameter Valj, which changes according to the envelope information or changes with time, the time count data, and the like by various calculations (1) described later. Good. The read tone waveform data Icj
(T), Isj (t), and TWj (t) are input to the circuits after the occurrence point.

The storage for each elapsed time from the start of sound generation or each envelope level is omitted, and each musical tone waveform data Icj (t), Isj (t), TWj (t).
On the other hand, the elapsed time from the start of sounding or the envelope level may be corrected and combined. This correction synthesis is based on various calculations (1) described later and the like, and an arithmetic unit for correcting synthesis of the sound generation elapsed time or the envelope level is provided at the output end of the tone waveform memory A05.

The "various operations (1) to be described later" described above or below are the addition or subtraction of each data in the adder, the multiplication or the division of each data in the multiplier, the addition or subtraction in the adder and the multiplier. Combination operation of these, other additive operation of each data, other synergistic operation of each data, bit shift operation of other data by certain data in the data shifter, data of one data becomes the upper rank and data of other data Is a lower-order combination calculation, calculation of each data based on a calculation formula in a calculation circuit, etc., calculation data of each data is stored in a memory, and each data is used as read address data. is there.

The tone waveform data Icj (t), Isj
(T) is the component waveform data obtained by extracting only the cosine component from one musical tone waveform data TWj (t), and the component obtained by extracting only the sine component from the other musical tone waveform data TWj (t). It may be waveform data. Such extraction is performed by the even transversal filter 13 and the odd transversal filter 14 described in the specification of Japanese Patent Application No. 56-37342 (Japanese Patent Publication No. 6-95280) and the drawings. The frequencies of the sampling signals (sampling signals) supplied to these transversal filters 13 and 14 are switched in a wide range,
As a result, the cosine component and the sine component are separated and extracted in the entire frequency band.

Other inventions disclosed in the specification and drawings of the present application and the inventors thereof are as follows.

[0116]

[Second Invention-Inventor Seiji Okamoto, Yutaka Washiyama] In the present invention, the frequency band of a musical tone waveform is shifted in frequency without changing the density of each frequency component of this frequency band. As a result, the frequency is shifted without changing the density of each frequency component of the musical sound wave frequency band, so that the overtone ratio of each frequency component of the frequency band changes, resulting in a tone color (sound quality).
Changes subtly, and it is possible to perform musical tone control that has never been seen before. Further, according to the present invention, the density of each frequency component of the frequency band of the musical sound is not changed, the frequency is shifted, and at least one formant generated by this shift is selected from a plurality of the same formants generated by the shift. Is selected and stored. Further, the stored musical tone waveform is shifted in frequency and output without changing the density of each frequency component in the frequency band. As a result, the density of each frequency component in the frequency band of the musical tone waveform does not change, and the frequency is shifted and stored / reproduced, so that the formant width can be made constant at any pitch. Furthermore, if the frequency of the musical tone waveform is lowered by the frequency shift, the stored sampling frequency of the musical tone waveform can be lowered, so that the musical tone waveform to be stored is compressed and stored.

[0117]

[1] For a musical tone waveform generating means for generating a musical tone waveform and a frequency band of the musical tone waveform generated by the musical tone waveform generating means, a frequency that shifts in frequency without changing the density of each frequency component of this frequency band. The frequency shift data generating means for generating the frequency shift data for determining the amount of frequency shift by the frequency shift means and transmitting the frequency shift data to the frequency shift means, and the musical tone waveform of which the frequency is shifted by the frequency shift means And a musical sound output means for outputting as a musical sound control device.

[0118]

2. The musical tone output means converts the musical tone waveform into a musical tone corresponding to a pitch and outputs the musical tone, or the frequency shift means shifts the musical tone waveform to a frequency corresponding to the musical pitch. The musical sound control device according to item 1.

[0119]

(3) The frequency shifting means multiplies and synthesizes the musical tone waveform generated by the musical tone waveform generating means with a sine wave signal and a cosine wave signal of the same frequency, and adds and synthesizes them through a low-pass filter means or a high-pass filter means. The musical sound control device according to claim 1 or 2.

[0120]

4. The frequency shift data generated by the frequency shift data generating means is data for selecting which band of the frequency characteristic of the filter is used. The tone control device described.

[0121]

[5] The musical tone waveform generating means is a partial musical tone waveform generating means for generating a plurality of partial musical tone waveforms having different frequency bands, or the musical tone outputting means synthesizes a plurality of partial musical tone waveforms to form one musical tone. 5. The musical sound control device according to claim 1, wherein the musical sound control device is a synthetic output means for outputting.

[0122]

6. The frequency shift data generated by the frequency shift data generating means is responsive to a musical factor, an elapsed time from the start of sounding, an envelope level, an envelope phase or an operator's setting instruction. The musical tone control device according to item 1, 2, 3, 4 or 5.

[0123]

[7] A musical tone waveform generating step of generating a musical tone waveform and a frequency band of the musical tone waveform generated in the musical tone waveform generating step are shifted in frequency without changing the density of each frequency component of this frequency band. 1 frequency shift process,
A selective extraction step of selectively extracting at least one formant from a plurality of formants of the same shape generated by the first frequency shifting step and a musical tone waveform selectively extracted by the selective extraction step are stored. And a first frequency shift data generating step for generating and transmitting first frequency shift data for determining the amount of frequency shift in the first frequency shifting step. Music waveform storage method

[0124]

8. A musical tone waveform reading step for reading a musical tone waveform stored by the musical tone waveform storing method according to claim 7, and for each frequency band of the musical tone waveform read in this musical tone waveform reading step, each frequency component of this frequency band A second frequency shifting step of shifting the frequency without changing the density, and a musical tone outputting step of outputting the musical tone waveform frequency-shifted by the second frequency shifting step as a musical tone,
The second frequency shift data generating step of generating and transmitting second frequency shift data for determining the amount of frequency shift in the second frequency shifting step. Waveform playback method.

[0125]

9. The musical tone waveform storing method or the musical tone waveform reproducing method according to claim 7, wherein the center frequency of the frequency band of the musical tone waveform frequency-shifted by the first frequency shifting step is zero.

[0126]

10. The musical tone waveform reproducing method according to claim 8, wherein the frequency shift in the second frequency shifting step is a frequency shift according to a pitch.

[0127]

11. The frequency shift in the first frequency shift step and the frequency shift in the second frequency shift step have opposite shift directions and the same shift amount. Alternatively, the method for reproducing a tone waveform as described in 10 above.

[0128]

12. The first frequency shifting step or the second frequency shifting step is characterized in that the sine wave signal and the cosine wave signal of the same frequency are multiplied and synthesized with the tone-shifted tone waveform. , 8, 9, 10 or 11, the musical tone waveform storing method or the musical tone waveform reproducing method.

[0129]

The musical tone waveform generating step is a partial musical tone waveform generating step of generating a plurality of partial musical tone waveforms having different frequency bands, or the musical tone outputting step synthesizes a plurality of partial musical tone waveforms into one musical tone. 13. The musical tone waveform storing method or the musical tone waveform reproducing method according to claim 7, 8, 9, 10, 11 or 12, which is a combined output step of outputting.

[0130]

14. A musical tone waveform according to claim 7, wherein the musical tone waveform storing step stores the musical tone waveform separately into a sine wave component and a cosine wave component. Memorization method or musical sound waveform reproduction method.

[0131]

15. The musical tone waveform storing step stores each musical tone waveform in accordance with a musical factor, an elapsed time from the start of sounding, an envelope level, an envelope phase or an operator's setting instruction. 8, 9, 1
0, 11, 12 or 13, the musical tone waveform storing method or the musical tone waveform reproducing method.

[0132]

The first frequency shift data generated in the first frequency shift data generating step or the second frequency shift data generated in the second frequency shift data generating step is a musical factor, a sound start. 16. The musical tone waveform storing method or musical tone according to claim 7, 8, 9, 10, 12, 13, 14 or 15, wherein the musical tone waveform storing method or musical tone is in accordance with an elapsed time from, a envelope level, an envelope phase or an operator's setting instruction. Waveform playback method.

[0133]

[Third Invention-Inventor Katsushi Ishii, Yutaka Washiyama] In the present invention, the gains of the frequency bands of a certain partial tone waveform are matched by matching the gains at the boundary portions of the respective frequency bands of the plurality of filtered partial tone waveforms. And the gains of the frequency bands of other partial tone waveforms are almost matched. As a result, the gains of the respective frequency band boundary portions of the respective partial tone waveforms are matched, and a balanced synthetic tone can be output.

[0134]

[1] Partial tone waveform generating means for generating a plurality of partial tone waveforms having different frequency bands, filter means for filtering a plurality of partial tone waveforms generated by the partial tone waveform generating means, and a filter by this filter means There is a synthesis output means for synthesizing a plurality of processed partial musical tone waveforms and outputting as one musical tone, and a gain matching at a boundary portion of each frequency band of the plurality of partial musical tone waveforms synthesized by this synthesizing means. A musical tone control apparatus comprising: a gain matching means for substantially matching a gain of a frequency band of a partial musical tone waveform with a gain of a frequency band of another partial musical tone waveform.

[0135]

The gain matching means generates a gain characteristic according to the filter characteristic of the filter means, and the gain characteristic generating means determines a boundary portion between frequency bands of the plurality of partial tone waveforms. Based on the gain acquired by the gain acquisition means and the gain acquired by this gain acquisition means, the gain of the boundary portion is matched, and the gain of the frequency band of a certain partial musical tone waveform and the frequency band of another partial musical tone waveform are obtained. And a boundary gain matching means for substantially matching the gain of the above-mentioned.
A musical tone control device as described.

[0136]

The gain acquisition means acquires the gain of the frequency band of one partial musical tone waveform and the gain of the frequency band of the other partial musical tone waveform in the boundary portion, and the boundary gain matching means obtains one gain. 3. The tone control apparatus according to claim 2, wherein the difference between the other gain and the other gain is calculated, and the correction calculation is performed according to this difference.

[0137]

4. The tone control apparatus according to claim 1, 2 or 3, wherein the frequency of the partial tone waveform changes according to a change in generation speed, and the gain at the boundary changes accordingly. .

[0138]

5. The partial musical tone waveform generating means comprises a partial musical tone waveform storing means for storing a partial musical tone waveform and a reading means for reading out the partial musical tone waveform from the partial musical tone waveform storing means. The musical sound control device described in 2, 3, or 4.

[0139]

(6) The filtering means performs filtering on almost all frequency bands of the tone waveform in the transition band between the pass band and the stop band, or the tone waveform from the pass band to the stop band. 6. The musical tone control apparatus according to claim 1, wherein the filter processing is performed for each frequency band.

[0140]

[7] The filter characteristic of the above-mentioned filter means includes a musical factor, an elapsed time from the start of sounding, an envelope level,
7. The musical tone control apparatus according to claim 1, wherein the musical tone control apparatus changes according to an envelope phase or an operator's setting instruction.

[0141]

As described above in detail, according to the present invention, filtering is performed for almost all frequency bands of the tone waveform only in the transition band between the pass band and the stop band of the filter. Therefore, in the transient band of the filter, the attenuation characteristic gradually changes, so that the change in the frequency characteristic of the musical tone to be filtered is generally changed from the fundamental wave to the harmonic wave or from the harmonic wave to the fundamental wave. Will change to. For this reason, only a part of the frequency characteristic of the musical tone does not change, and it is possible to realize an overall change in frequency, and it is possible to perform an unprecedented control over the musical tone. Play. Further, according to the present invention, the frequency band of the musical tone waveform is shifted on the frequency without changing the density of each frequency component of this frequency band, is filtered, and is further shifted to the frequency according to the pitch. It was done. Therefore, it is possible to select a specific portion of the filter characteristic and perform the filter control only at this portion, and it is possible to achieve the filter characteristic in a stable manner. Further, since the frequency shift according to the pitch is performed after the filter processing, it is possible to perform the filter processing regardless of the pitch.

[Brief description of drawings]

FIG. 1 is an overall circuit diagram of a musical sound generation device and a musical sound control device.

FIG. 2 is a circuit diagram showing a shift filter unit A0.

FIG. 3 is a diagram showing frequency characteristics of a band control filter A06.

FIG. 4 is a diagram showing frequency shifts of a first frequency shift unit A10 and a second frequency shift unit A30.

FIG. 5 is a diagram showing a gain matching state at the boundary of the frequency band of each partial sound of the musical tone waveform data TWj (t) in the transient band of the formant control filter A20.

FIG. 6: Formant control filter A20, filter A
It is a circuit diagram which shows 64 and A65.

FIG. 7: Formant control filter A20, filter A
It is a figure which shows the flowchart of the filter process of 64 and A65.

FIG. 8 is a diagram showing frequency characteristics of a formant control filter A20.

FIG. 9 is a diagram showing frequency characteristics of another example of the formant control filter A20.

FIG. 10 is a diagram showing a shift filter table A90.

FIG. 11 is a circuit diagram showing a first frequency shift unit A10 or a second frequency shift unit A30.

FIG. 12 is a diagram showing a difference in tone color (sound quality) of a musical tone due to frequency shift.

FIG. 13 is a circuit diagram showing a shift filter unit A0 (second embodiment).

FIG. 14 is a diagram showing an operation of a frequency shift unit A10 (A30).

FIG. 15 is a circuit diagram showing a frequency shift circuit A91 and the like.

FIG. 16 is a circuit diagram showing a shift filter unit A0 (third embodiment).

FIG. 17 is a circuit diagram showing frequency shift circuits AA1 to AA4 and the like.

FIG. 18 is a circuit diagram showing a formant control filter A20, filters A64 and A65 (second embodiment).

FIG. 19 is a diagram showing a flowchart of a filter process of a formant control filter A20, filters A64 and A65 (second embodiment).

FIG. 20 is a circuit diagram showing a formant control filter A20, filters A64 and A65 (third embodiment).

[Explanation of symbols]

A0 ... shift filter section, A05 ... musical tone waveform memory,
A06 ... band control filter, A10 ... first frequency shift unit, A20 ... formant control filter, A30 ... second frequency shift unit, A55 ... filter gain table, A
60 ... Cosine table, A 61 ... Sine table, A
67, A74 ... Programmable oscillator, A70 ... Accumulator, A77 ... Demultiplexer, A79 ... Multiplexer, A90 ... Shift filter table, AB
3 ... Interpolation circuit, A91, AA1, AA2, AA3, AA
4 ... Frequency shift circuit, A93 ... Envelope control circuit, A94 ... Loudness control circuit.

Claims (12)

[Claims] 1. A musical tone waveform generating means for generating a musical tone waveform, and a filter means for filtering the musical tone waveform generated by the musical tone waveform generating means, the filter means being located between its pass band and stop band. Only in the transient band of
Filtering is performed on almost all frequency bands of the musical tone waveform, and the frequency of the musical tone waveform is generated so that the frequency band of the musical tone waveform generated by the musical tone waveform generating means falls within the transient band of the filter means. A musical tone control apparatus comprising: limiting means for limiting a band; and musical tone outputting means for outputting a musical tone waveform filter-controlled by the filtering means as a musical tone. 2. Musical tone control data generating means for generating musical tone control data, and based on the musical tone control data generated by this musical tone control data generating means, filter control is performed at any position within the transient band of the filter means. 2. The musical tone control apparatus according to claim 1, further comprising: a transition band selecting means for selecting the or. 3. The frequency band of the tone waveform is frequency-shifted to select where in the transient band of the filter means the filter control is performed, or the transient band of the filter means is frequency-shifted. 3. The musical tone control apparatus according to claim 1 or 2, characterized in that a position in the transition band of the filter means to perform the filter control is selected. 4. The musical tone waveform generating means is a partial musical tone waveform generating means for generating a plurality of partial musical tone waveforms having different frequency bands, or the musical tone output means synthesizes a plurality of partial musical tone waveforms to form one musical tone waveform. 4. The musical tone control device according to claim 1, wherein the musical tone control device is a synthetic output means for outputting as a musical tone. 5. The frequency characteristic of a portion of the filter means in which the filter control is performed in the transient band is linear, and the plurality of partial tone waveforms are subjected to the filter control only in the linear portion, and the filter means is provided. 5. The musical tone control apparatus according to claim 1, 2, 3 or 4, wherein even if the actual frequency characteristic of the transient band is non-linear, the result is linear. 6. The musical tone control data generated from the musical tone control data generating means corresponds to a musical factor, an elapsed time from the start of sound generation, an envelope level, an envelope phase, or an operator's setting instruction. The musical sound control device according to claim 2 or 3. 7. A musical tone waveform generating means for generating a musical tone waveform and frequency bands of the musical tone waveform generated by this musical tone waveform generating means are shifted in frequency without changing the density of each frequency component of this frequency band. First frequency shifting means, filter means for filtering the musical tone waveform frequency-shifted by the first frequency shifting means, and the musical tone waveform filtered by this filtering means is shifted to a frequency according to the pitch. A second frequency shift means, a tone output means for outputting the tone waveform whose frequency is shifted by the second frequency shift means as a tone, and a frequency band of the tone waveform generated by the tone waveform generating means. A first means for determining the amount of frequency shift in the first frequency shift means so as to fall within the transient band of the filter means. First frequency shift data generation means for generating wave number shift data and sending it to the first frequency shift means, and first frequency shift data from the first frequency shift means are generated according to the pitch. And a second frequency shift data generating means for correcting the frequency shift data and sending it to the second frequency shifting means. 8. The filtering means performs filtering for almost all frequency bands of the tone waveform in a transition band between its pass band and stop band, or
8. The musical tone control apparatus according to claim 7, wherein the musical tone frequency band is filtered from the pass band to the stop band. 9. The frequency shift means multiplies and synthesizes the musical tone waveform generated by the musical tone waveform generating means with a sine wave signal and a cosine wave signal having the same frequency, and adds and synthesizes them through low-pass filter means or high-pass filter means. 9. The musical sound control device according to claim 7, wherein: 10. The first frequency shift data generated from the first frequency shift data generating means is data for selecting which band of the frequency characteristic of the filter means is to be used. Claims 7 and 8 characterized
Alternatively, the musical tone control device described in 9. 11. The musical tone waveform generating means is a partial musical tone waveform generating means for generating a plurality of partial musical tone waveforms having different frequency bands, or the musical tone output means synthesizes a plurality of partial musical tone waveforms to form one musical tone waveform. 11. The synthesis output means for outputting as a musical sound, claim 7, 8, 9 or 10.
A musical tone control device as described. 12. The first frequency shift data generated from the first frequency shift data generating means is in accordance with a musical factor, an elapsed time from the start of sounding, an envelope level, an envelope phase or an operator's setting instruction. The musical sound control device according to claim 7, 8, 9, 10 or 11.