JPH10124062A - Musical sound synthesizer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To avoid the situation in which musical sound data are unintentionally saturated by detecting the level of waveform data that are cycling in a loop means and issuing an attenuation instruction for the input gain of a first attenuator for an input gain adjustment means when the detection result exceeds a threshold value. SOLUTION: A level detector 109 detects the amplitude level of the waveform data being cycled in a loop circuit 108. A gain adjusting circuit 110 adjusts the amount of attenuation, which attenuates an input gain Ign and a loop gain Lgn to be multiplied, in a multiplier 102 that is a first attenuator and a multiplier 105 that is a second attenuator based on the detection result of the detector 109. In other words, the detector 109 detects the amplitude level of the waveform data which cycle in the circuit 108 and when the detection result exceeds the threshold value, a decrement instructing flag is transmitted to the circuit 110 and the circuit 110 attenuates the gains Ign or Lgn.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a musical tone synthesizer for extracting musical tone data used for a sound source of an electronic musical instrument by circulating a waveform through a closed loop circuit.

[0002]

2. Description of the Related Art Conventionally, this type of apparatus has been disclosed in Japanese Patent Publication No.
As disclosed in Japanese Patent No. 9354, for example, by circulating a single-shot signal such as an impulse by a closed-loop circuit (so-called comb filter), a natural musical instrument is used by utilizing the “action of amplifying a periodic component” of the comb filter. A device that synthesizes a tone similar to a sound is disclosed.

According to Japanese Patent Application No. 7-9922, a signal other than a single-shot signal, that is, a steady signal (for example, a continuous triangular wave) is inserted by interposing an amplitude level limiter in a closed loop circuit. (Such as periodic waveforms) have been proposed to prevent overflow.

[0004] Hereinafter, a conventional tone synthesizer as described above will be described with reference to the drawings. FIG. 5 is a block diagram of a conventional tone synthesizer. In FIG. 5, 10
1 is a waveform generating circuit as a waveform generating means for reading waveform data stored in a memory or the like in advance, 102 is a multiplier as a first attenuator for multiplying a predetermined attenuation rate, and 105 is a multiplier for multiplying a predetermined attenuation rate. A multiplier as an attenuator of 2, 1
03 is a delay unit for delaying data by a predetermined time, 104
Is a low-pass filter which is a digital filter, 106 is an adder, 107 is an amplitude level limiter (hereinafter simply referred to as a limiter) for limiting the amplitude level of the addition result of the adder 106, 108 is a delay unit 103, a low-pass filter This is a loop circuit as loop means in which 104, multiplier 105, and adder 106 are put together.

The waveform generation circuit 101 can be easily realized by using a well-known circuit such as a read-only memory in which waveform data is stored in advance and a counter for incrementing a memory address by one according to a sampling clock SCK. The description of the internal configuration is omitted because it is possible.
Further, since the delay unit 103 can be easily realized by forming the read / write memory into a ring memory, the description of the internal configuration is also omitted.

FIG. 6 is a characteristic diagram showing the conversion characteristics of the limiter 107. In FIG. 6, the horizontal axis x and the vertical axis y are axes corresponding to the input and output of the limiter, respectively. Input 19
The linear characteristic of y = x is obtained up to the bit level, and the non-linear characteristic is gradually saturated up to the 19-bit level. The value of y corresponding to the point where x becomes 20 bits is S. The addition output of the adder 106 corresponding to the input to the limiter 107 does not overflow up to 21 bits, the other circuits have a 20-bit operation precision, and the low-pass filter 104 has at least 20 bits inside. Is not amplified, that is, does not overflow internally.

FIG. 7 is a characteristic diagram showing a transfer characteristic of the loop circuit 108. However, the low-pass filter 104 in the loop circuit 108 has a characteristic of 6 dB / 1 octave, and the low-pass frequency fc of the low-pass filter 104 is f
3 (f3 is the third harmonic of the fundamental frequency f1 corresponding to the pitch).

FIG. 8 is a characteristic diagram showing a time axis characteristic of a waveform (triangular wave) generated by the waveform generating circuit 101. FIG. 9 is a characteristic diagram showing a time axis characteristic of musical sound data.

The operation of the tone synthesizer constructed as described above will be described below. First, a triangular wave shown in FIG. 8 is input to the loop circuit 108 as input waveform data from the waveform generation circuit 101 based on the generation of the tone generation start flag KON and the system clock SCK. In the loop circuit 108, by repeating the cyclic operation,
The tone processing is performed according to the characteristics shown in FIG.

At this time, the amplitude level of the tone data is controlled by the values of the input gain Ig and the loop gain Lg in the multipliers 102 and 105. When the delay amount of the delay unit 103 is set so that f1 of the characteristic shown in FIG. 7 becomes a frequency corresponding to the pitch of the input waveform data, and fc of the low-pass filter 104 is considerably increased, Ig and Lg become (Equation 1) If the relationship of (1) is satisfied, the amplitude level of the musical tone data gradually approaches the S level (a level large enough not to saturate) over time.

However, the frequency corresponding to the pitch of the input waveform data is different from f1 or the low-pass filter 1
When the fc of No. 04 is considerably low, the amplitude level of the musical tone data is considerably lower than the S level even if the relationship of (Equation 1) is satisfied, and the sound volume is insufficient. In other words, a sufficient SN ratio cannot be obtained (Ig + L
g = 1 ... Equation 1).

Therefore, by setting the sum of the input gain Ig and the loop gain Lg to be larger than 1, the amplitude level of the tone data can be made sufficiently large. In this case, the waveform data circulating in the loop circuit 108 may be amplified as time elapses to reach the areas A and B in FIG. 6, but the tone data may be amplified as shown in FIG. The waveform becomes saturated at a certain level. That is, 2 which is the calculation accuracy of the loop circuit 108
Even if the signal is amplified up to 0 bits or more, discontinuous noise due to overflow (sign inversion) can be avoided.

[0013]

However, in the above-described conventional tone synthesizer, even if overflow (discontinuous noise due to sign inversion or the like) can be avoided, tone data is prevented from becoming saturated. I can't. Incidentally, when parameters (such as the input gain Ig and the loop gain Lg) that are set to be in a saturated state are set, the waveform gradually approaches a square wave with the lapse of time regardless of the type of input waveform data. There has been a problem that the tone has changed or the tone has suddenly changed at the moment when the tone has become saturated, so that a sense of incongruity is felt.

Even if the gain is adjusted so that the tone data of the maximum amplitude level is obtained before the sounding is started so as not to be saturated, the fc of the digital filter 104 and the delay amount of the delay unit 103 are not changed during the sounding. If the input waveform data and the resonance state of the loop circuit become larger than the initial state due to the temporal change, the saturation state is caused unintentionally, which hinders intentional tone control. There was a problem that.

The present invention solves the above-mentioned problems, and provides a tone synthesizer capable of avoiding the intentional saturation of tone data.

[0016]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, a musical tone synthesizing apparatus according to the present invention provides a specific periodic component of input waveform data by controlling a cyclic operation of a loop means and a control of a digital filter fc in the loop means. When adjusting the amount and changing the timbre, the level detection means detects the level of the waveform data circulating through the loop means, and when the detection result exceeds the threshold value, the first gain is supplied to the input gain adjustment means. It is characterized in that an input gain is attenuated in the attenuator and the tone data is always kept at an appropriate level.

Further, by adjusting the cyclic operation of the loop means and controlling the fc of the digital filter in the loop means, the level detecting means makes a cyclic movement of the loop means when adjusting the specific periodic component amount of the input waveform data and changing the timbre. The level of the middle waveform data is detected, and when the detection result exceeds the threshold value, the loop gain adjusting means is instructed to attenuate the loop gain in the second attenuator, and the tone data is always kept at an appropriate level. It is characterized by doing.

As described above, it is possible to prevent the tone data from being unintentionally saturated.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A musical sound synthesizer according to a first aspect of the present invention comprises a waveform generating means for generating an arbitrary waveform, and a first means for attenuating an amplitude level of waveform data transmitted from the waveform generating means. , And the waveform data from the first attenuator is circulated by a data circuit in which a delay unit, a digital filter, a second attenuator, and an amplitude level limiter are connected in a closed loop. A tone synthesizer having loop means for extracting waveform data as desired tone data; a level detecting means for detecting a level of waveform data circulating in the loop means; And an input gain adjusting means for adjusting the attenuation rate of the one attenuator.

According to this structure, the level detecting means is controlled by the loop detecting means for controlling the cyclic operation of the loop means and controlling the fc of the digital filter in the loop means to adjust the specific periodic component amount of the input waveform data to change the timbre. Detecting the level of the waveform data circulating through the means, and when the detection result exceeds the threshold value, instructs the input gain adjusting means to attenuate the input gain in the first attenuator so that the tone data is always properly adjusted. Hold on level.

According to a second aspect of the present invention, there is provided a musical sound synthesizer comprising: a waveform generating means for generating an arbitrary waveform; a first attenuator for attenuating an amplitude level of waveform data transmitted from the waveform generating means; The waveform data from the first attenuator is circulated by a data circulating circuit in which a delay unit, a digital filter, a second attenuator, and an amplitude level limiter are connected in a closed loop, and the circulating waveform data is converted into desired tone data. And a loop means for taking out as
Level detecting means for detecting the level of waveform data circulating in the loop means, and loop gain adjusting means for adjusting an attenuation rate in a second attenuator in the loop means according to a detection result of the level detecting means; The configuration is provided with.

According to this configuration, the level detecting means controls the cyclic operation of the loop means and the control of the fc of the digital filter in the loop means to adjust the specific periodic component amount of the input waveform data to change the timbre. Detecting the level of the waveform data circulating through the means, and when the detection result exceeds the threshold value, instructs the loop gain adjusting means to attenuate the loop gain in the second attenuator so that the tone data is always properly adjusted. Hold on level.

Hereinafter, a tone synthesizer according to an embodiment of the present invention will be specifically described with reference to the drawings.
FIG. 1 is a block diagram of a musical sound synthesizer according to the present embodiment. In FIG. 1, reference numeral 109 denotes a level detector which is means for detecting the amplitude level of waveform data circulating in the loop circuit 108, and reference numeral 110 denotes a multiplication unit which is a first attenuator according to the detection result of the level detector 109. The gain adjusting circuit adjusts the amount of attenuation (gain in a broad sense) for attenuating the input gain Ign and the loop gain Lgn to be multiplied in the multiplier 102 and the multiplier 105 as the second attenuator. Other circuits are the same as those of the conventional tone synthesizer.

FIG. 2 is a circuit diagram of the level detector 109. In FIG. 2, reference numeral 201 denotes an exclusive OR gate. FIG. 3 is a circuit diagram of the gain adjustment circuit 110. In FIG. 3, 301, 302, 304, and 305 are selectors, and 303 is a decrementer.

FIG. 4 is a characteristic diagram showing the time axis characteristics of the tone data and the input gain Ign. The operation of the tone synthesizer configured as described above will be described below. Since the basic synthesizing operation is the same as that of the conventional musical sound synthesizing apparatus, only the differences will be described here.

First, in FIG. 1, the level detector 109
Examines two bits, bit 20 (MSB) and bit 19, of the 21 bits output from the adder 106 in order to detect the amplitude level of the waveform data circulating in the loop circuit 108.

Since the calculation accuracy is the same as that of the conventional tone synthesizer, both bit 20 (MSB) and bit 19 are sign bits. When bit 20 is value 0 and bit 19 is value 1, it is a case where the number exceeds 20 bits in the positive area. When bit 20 is value 1 and bit 19 is value 0, 20 bits in the negative area. Is exceeded. These correspond to the areas A and B in the conversion characteristic diagram of the limiter 107 shown in FIG.

When bit 20 has a value of 0 and bit 19 has a value of 1, or when bit 20 has a value of 1 and bit 19 has a value of 0, the exclusive OR gate 201 of the level detector 109 operates as follows. A decrement instruction flag is sent to gain adjustment circuit 110.

Next, the gain adjustment circuit 110 shown in FIG. 3 will be described. In FIG. 3, when the parameter selection flag PRM has a value of 0, the input gain Ign adjusted by the detection result of the level detector 109 is set in the multiplier 102, and the loop gain initial value Lg0 is set in the multiplier 105 as it is. Mode. At this time, the gain adjustment circuit operates as input gain adjustment means. On the other hand, when the parameter selection flag PRM has the value 1, the mode in which the loop gain Lgn adjusted by the detection result of the level detector 109 is set in the multiplier 105 and the input gain initial value Ig0 is set in the multiplier 102 as it is. Becomes At this time, the gain adjustment circuit operates as loop gain adjustment means.

First, when the value of PRM is 0, the selector 301
Then, the input gain initial value Ig0 is sent to the selector 302. In the selector 302, the output of the selector 301 is sent to the decrementer 303 at the first time of the start of sound generation, that is, when KON occurs. In the decrementer 303,
When the decrement instruction flag from the level detector 109 has the value 0, the output value of the selector 302 is sent to the selectors 304 and 305 as it is, and when the decrement instruction flag has the value 1, the output value of the selector 302 is decremented. To the selectors 304 and 305. Selector 3
At 04, the output of the decrementer 303, that is, Ign, is selected and sent to the multiplier 102. At the selector 305, the loop gain initial value Lg0 is selected and sent to the multiplier 105.

Next, when the value of PRM is 1, the selector 301 sends the loop gain initial value Lg0 to the selector 302. In the selector 302, the output of the selector 301 is sent to the decrementer 303 at the first time of the start of sound generation, that is, when KON occurs. In the decrementer 303,
When the decrement instruction flag from the level detector 109 has the value 0, the output value of the selector 302 is sent to the selectors 304 and 305 as it is, and when the decrement instruction flag has the value 1, the output value of the selector 302 is decremented. To the selectors 304 and 305. Selector 3
04, the input gain initial value Ig0 is selected and the multiplier 10
2 is sent to the selector 305 and the decrementer 303
, Ie, Lgn, is selected and sent to the multiplier 105.

FIG. 4 shows the time axis characteristics of the tone data and the input gain Ign when the value of PRM is 0. In FIG. 4, at time 0, that is, at the time of KON, the input gain initial value I
g0 and the loop gain initial value Lg0 have a value of 1, and the data held in the delay unit 103 in the loop circuit 108 has a value of 0.
And For example, the input waveform data (continuous triangular wave) shown in FIG. 8 passes through the multiplier 102 as it is, and is subjected to the nonlinear conversion shown in FIG. 6 by the limiter 107 via the adder 106, so that the first wave has the maximum amplitude value. Since the level is S and the level is limited as the amplitude level increases, the waveform has a larger amplitude than the first wave.

The vicinity of the maximum amplitude level covers the areas A and B shown in FIG. 6, that is, since the effective bit length exceeds 20 bits, the level detector 109 sends a decrement instruction flag to the level adjustment circuit 110. Send out.
As a result, the input gain Ign is decremented, and until the decrement instruction flag is no longer generated as shown in FIG.
gn attenuates, and the amplitude level of the input waveform data passing through the multiplier 102 attenuates.

By repeating this operation, the amplitude level of the waveform data circulating in the loop circuit 109 is maintained at a sufficiently large amplitude level (around the S level) so as not to cover the regions A and B in FIG. As a result, the tone data has a waveform that does not saturate, as shown in the third and subsequent waves in FIG.

The characteristics of the tone data and the input gain Ign when the value of the PRM is 0, that is, the case where the input gain is adjusted have been described above. The case where the value of the PRM is 1, that is, the case where the loop gain is adjusted is described. Similarly, non-saturated musical tone data can be obtained.

Note that the decrement width in the decrementer 303 can be set arbitrarily. Also, it is possible to adjust both the input gain and the loop gain at the same time.

The waveform generating circuit 101 may be a circuit for converting an analog audio signal taken in from outside into a digital signal in real time. Also, in the above description, the input waveform data is a repetitive waveform as shown in FIG. 8, but a musical tone waveform (a waveform from the rising edge to a sufficient attenuation) such as a piano sound or a musical tone waveform of the loop circuit 108 is used. A residual waveform obtained by convolution with the inverse characteristic may be used.

In the above description, the gain is adjusted only in the attenuation direction as shown in FIG. 4. For example, while the decrement instruction flag is 0, the gain is incremented by a width smaller than the decrement width and the gain is increased. It may be possible to adjust the gain at the beginning.

As described above, when the level detector 109 detects that the amplitude level of the waveform data circulating in the loop circuit 109 falls on the areas A and B in FIG.
, And the level adjustment circuit 110 sends the input gain Ign or the loop gain L
By attenuating gn, as shown in FIG. 4, the amplitude level is large enough not to be saturated (level S).
Music data can be obtained.

As a result, it is possible to prevent the musical sound data from being unintentionally saturated. Further, as shown in FIG. 9, when the input gain Ign is automatically adjusted after setting the initial value of the input gain to the maximum level (value 1), the rising characteristic of the musical tone data can be sharpened.
Envelope characteristics of a musical tone having a relatively large level from the initial level to the steady level can be faithfully reproduced.

The number of output bits of the adder 106 has a gain margin (1 bit), and the level detector 109
The level is detected at a stage before the limiter 107 (input portion of the limiter 107), so that a waveform sample (saturated sample) at the time of level detection is not output as musical tone data as it is.

[0042]

As described above, according to the present invention, the cyclic operation of the loop means and the f of the digital filter in the loop means are performed.
When the control of c, when adjusting the specific periodic component amount of the input waveform data and changing the timbre, the level detection means detects the level of the waveform data circulating through the loop means, and the detection result exceeds the threshold. Instructs the input gain adjusting means to attenuate the input gain in the first attenuator,
The tone data can always be held at an appropriate level.

Further, by controlling the cyclic operation of the loop means and the control of the fc of the digital filter in the loop means, the level detecting means makes the cyclic operation of the loop means when the specific periodic component amount of the input waveform data is adjusted to change the timbre. The level of the middle waveform data is detected, and when the detection result exceeds the threshold value, the loop gain adjusting means is instructed to attenuate the loop gain in the second attenuator, and the tone data is always kept at an appropriate level. can do.

As described above, it is possible to prevent the musical sound data from being unintentionally saturated.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a musical sound synthesizer according to an embodiment of the present invention.

FIG. 2 is a circuit diagram of a level detector 109 according to the embodiment;

FIG. 3 is a circuit diagram of a gain adjustment circuit 110 according to the embodiment;

FIG. 4 is a time axis characteristic diagram of musical sound data and input gain according to the embodiment;

FIG. 5 is a block diagram showing a configuration of a conventional tone synthesizer.

FIG. 6 is a characteristic diagram of the limiter 107 of the conventional example.

FIG. 7 is a transfer characteristic diagram of the loop circuit of the conventional example.

FIG. 8 is a time axis characteristic diagram of input waveform data of the conventional example.

FIG. 9 is a time axis characteristic diagram of musical sound data of the conventional example.

[Explanation of symbols]

 109 level detector 110 gain adjustment circuit 201 exclusive OR gate 303 decrementer

Claims (2)

[Claims] 1. A waveform generating means for generating an arbitrary waveform,
A first attenuator for attenuating the amplitude level of the waveform data sent from the waveform generating means; a delay unit, a digital filter, a second attenuator, and an amplitude level limiter for applying the waveform data from the first attenuator to the first attenuator; Are looped by a data loop connected in a closed loop, and the loop means for taking out the waveform data during the loop as desired music data is detected by detecting the level of the waveform data circulating in the loop means. A musical tone synthesizer comprising: a level detecting means for performing an adjustment; and an input gain adjusting means for adjusting an attenuation rate of the first attenuator according to a detection result of the level detecting means. 2. A waveform generating means for generating an arbitrary waveform,
A first attenuator for attenuating the amplitude level of the waveform data sent from the waveform generating means; a delay unit, a digital filter, a second attenuator, and an amplitude level limiter for applying the waveform data from the first attenuator to the first attenuator; Are looped by a data loop connected in a closed loop, and the loop means for taking out the waveform data during the loop as desired music data is detected by detecting the level of the waveform data circulating in the loop means. And a loop gain adjusting means for adjusting an attenuation rate of a second attenuator in the loop means in accordance with a detection result of the level detecting means.