JP3541333B2 - Sound signal generator - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an acoustic signal generator.
[0002]
[Prior art]
2. Description of the Related Art At present, there is an acoustic signal generating apparatus that generates a sound or a musical sound, and generates an acoustic signal by reading data such as a sound or a musical sound stored in a memory. This type of acoustic signal generator has a configuration as shown in FIG. 5, and stores audio data obtained by sampling voice or musical sound at an appropriate sampling frequency in a ROM 61, which is repeated by an address counter 62. The signal from the ROM 61 is read out, latched by a latch circuit 63, and D / A converted by a D / A converter circuit 64 to generate an acoustic signal. Here, a frequency dividing circuit 65 divides the reference clock signal to generate clock signals for reading and latching, and outputs the signals to the address counter 62 and the latch circuit 63, respectively. When a melody is formed by this type, a melody is generated by generating a sound signal of a target pitch by changing a reading frequency of acoustic data for each target pitch. That is, when obtaining a pitch several octaves higher than the sampled reference audio signal, the frequency read from the ROM 61 is set to several times the sampling frequency of the reference audio signal. In this case, as shown in FIG. 6A, for example, if the frequency band of the signal output from the ROM 61 at the reference read frequency is Fr0, the frequency band of the signal obtained by the number of read cycles several times higher is obtained. Spreads like Fr1 shown in B of the same figure, and an acoustic signal of a target pitch is obtained.
[0003]
[Problems to be solved by the invention]
However, when trying to obtain a pitch higher than a certain level with such a device, the following problems occur.
[0004]
When a signal read from the ROM 61 is latched at an appropriate latch frequency as described above, a frequency band centered on a frequency that is an integral multiple of the latch frequency is generated. For example, if the latch frequency is fl, the frequency nfl (n is an integer,... -2, -1, 1, 2,...) Corresponding to each of the frequency bands Fr0 and Fr1 in FIGS. And the frequency bands Fr0In and Fr1ln centered on. For this reason, when a read frequency several times higher than the read frequency of the reference signal is used, the frequency band of the read signal expands like Fr1, and the frequency band around the latch frequency fl Also spread like Fr1ln. Since the latch frequency is fixed, the frequency band Fr1 and the frequency band Fr11l1 overlap, and the signal output from the latch circuit 63 contains unnecessary components, resulting in deterioration of the acoustic signal.
[0005]
To avoid this, it is necessary to set a high latch frequency corresponding to the upper limit of the musical interval and the upper limit of the read frequency, but this is disadvantageous in cost.
[0006]
[Means for Solving the Problems]
In order to solve the above problems, the present invention relates to a method of comparing a frequency band of a signal obtained by reading acoustic data stored in a ROM at a desired read frequency with a frequency band of a signal generated when the signal is latched. In the case where a specific frequency band that overlaps occurs, a specific frequency band included in the signal read from the ROM is suppressed in advance, and then a latch and D / A conversion are performed to obtain an audio signal. That is, as the pitch increases, the portion on the harmonic side of the frequency component included in the audio data is suppressed by an amount corresponding to the rise of the pitch, so that the deterioration of the audio signal due to the mixing of unnecessary frequency components as described above can be suppressed. As a result, the upper limit of the pitch can be raised while suppressing signal deterioration.
[0007]
BEST MODE FOR CARRYING OUT THE INVENTION
Storage means for storing first sound data obtained by sampling sounds such as voices and musical sounds; reading means for reading out the first sound data from the storage means at a desired first frequency; A first frequency band of a first signal obtained by reading the audio data at a first frequency, and a second frequency of a signal obtained by latching the first signal at a predetermined second frequency Determining means for determining whether or not there is a specific frequency band overlapping with a second frequency band centered on the first frequency band, and when the determining means determines that the specific frequency band exists, it is included in the first signal. A conversion unit that suppresses the specific frequency band, converts the first sound data into the second sound data, and outputs a second signal corresponding to the second sound data; and a first signal or a second signal. The second signal to the second Constituting a latch means for latching and outputting at a frequency, an acoustic signal generating apparatus and a D / A conversion means for generating a sound signal by D / A conversion output from said latch means.
[0008]
【Example】
Next, an acoustic signal generator according to an embodiment of the present invention will be described. FIG. 1 is an explanatory diagram showing the configuration of this example. Reference numeral 1 denotes a ROM in which peak values obtained by sampling sounds such as voices and musical sounds are stored as sound data. Here, at the addresses 0 to 255 of the address of the ROM 1, one cycle of a waveform of a specific sound such as a voice or a musical tone, for example, a waveform of one cycle as shown in FIG. The constituent acoustic data is stored. Here, it is assumed that this fundamental wave includes up to a harmonic component of a frequency that is 40 times the fundamental frequency.
[0009]
Reference numeral 2 denotes an address counter, which receives a read clock signal fr from the frequency dividing circuit 3 and thereby sequentially increments the address of the ROM 1 to read acoustic data.
[0010]
Reference numeral 4 denotes a control circuit, for example, as shown in FIG. 3D, a frequency band F0 of a signal read from the ROM 1 and a signal obtained by latching the signal at a predetermined frequency fl by a latch circuit 5. An output is generated by determining whether or not there is a specific frequency band Fs overlapping the frequency band Fl around the frequency fl in the signal. Here, the control circuit 4 is a signal representing the frequency division ratio of the clock signal (frequency fr) output from the frequency divider 3 to the address counter 2 and the clock signal (frequency fl) output to the latch circuit 5. Q1 and Q2 have been received, and the presence or absence of a specific frequency is determined based on these. In this example, the latch frequency fl is set to 40 kHz, which is twice the upper limit of the audible frequency, 20 kHz, and the ROM 1 stores up to 40 times the frequency component of the fundamental wave. The frequency f0 of the fundamental wave realized at the read frequency is calculated, a value 40 times the frequency f0 is set as the upper limit value of the frequency band of the signal from the ROM 1, and it is determined whether or not this value exceeds 10 kHz. Here, the reason for judging whether or not the upper limit exceeds 10 kHz is as follows. Since the frequency bands F0 and Fl begin to overlap at 20 kHz to generate the specific frequency band FS, the specific frequency band FS can be suppressed by cutting off the frequency band F0 of 20 kHz or more by the arithmetic circuit 6 described later. However, since an actual arithmetic circuit cannot realize an ideal low-pass filter with a cutoff characteristic, the cutoff frequency is set to 10 kHz in order to provide a margin, and an upper limit value exceeding the cutoff frequency is determined. .
[0011]
Numeral 6 denotes an arithmetic circuit, which receives audio data from the ROM 1, receives a determination output from the control circuit 4, and outputs a signal in which the specific frequency band is suppressed. In the state where the judgment output is not generated from the control circuit 4, the acoustic data is not subjected to the arithmetic processing, and is output to the latch circuit 5 as it is. Here, the arithmetic circuit 6 includes a digital filter circuit (not shown) and performs an arithmetic process for cutting off 10 kHz or more. For example, the arithmetic circuit 6 performs the following arithmetic operation.
[0012]
D (n) = { ⁷ _{k = -7} [x (n) × {sin (ωck)} / πk]
In this figure, n is the address number of the ROM 1, x (n) is the sound data of the address number n, and D (n) is new sound data corresponding to the address n. Further, ωc is a coefficient for determining the cutoff frequency by this calculation. Here, the read frequency is fr, and ωc = 2π × 10 [kHz] / fr [kHz].
[0013]
Reference numeral 7 denotes a D / A conversion circuit, which D / A converts a signal output from the latch circuit 5 and outputs an audio signal.
[0014]
Next, the operation of this example will be described with reference to the flowchart of FIG.
[0015]
First, when a signal specifying a target pitch is sent from the melody generation control means (not shown) to the apparatus of the present embodiment (START), the frequency dividing circuit 3 responds to the target pitch from the reference clock signal (frequency fb). A frequency division ratio for generating a read clock signal (frequency fr) is designated. The control circuit 4 calculates the frequency of the fundamental wave obtained by reading the ROM 1 with the clock signal (frequency fr) specified here using this frequency division ratio (step a). Next, the upper limit value of the frequency band of the signal obtained by reading from the ROM 1 is calculated from this frequency, and it is determined whether the upper limit value is greater than 10 kHz. Is generated (step b). Here, assuming that the frequency of the reference clock is fb and the frequency division ratio is c, the readout frequency fr is = fb / c. Here, since the fundamental wave to be read is stored at addresses 0 to 255 of the ROM 1, the frequency f0 of the fundamental wave obtained here is f0 = fr / 256. Here, when 40f0 kHz> 10 kHz, a determination output is generated. When this determination output is generated, first, based on the frequency of the fundamental wave (step c), a coefficient ωc for defining the cutoff frequency is calculated as ωc = 2π × 10 [kHz] / fr [kHz. ]. Next, audio data is read from the ROM 1 according to the clock signal of the frequency dividing circuit 3. This acoustic data is stored in the arithmetic circuit 6, for example, in a shift register or the like in the order of address numbers (step d). The arithmetic circuit 6 performs the above-described arithmetic processing using the seven sound data before and after each sound data, and outputs the new sound data to the latch circuit 5 (step e). This new acoustic data is sequentially generated for each acoustic data read from the ROM 1. This new acoustic data is obtained by cutting off the frequency band of the signal output from the ROM 1 at a cutoff frequency of 10 kHz. For example, in the frequency band F0 of the signal from the ROM 1 as shown in FIG. 3A, the frequency component of 10 kHz or more is cut off by the arithmetic processing, and becomes the frequency band F1 as shown in FIG. The latch circuit 5 latches the input signal at the latch frequency fl (step f). For this reason, in the signal output from the latch circuit 5 shown in FIG. 4C, the overlap between the frequency band F12 generated around the latch frequency fl and the frequency band F1 is avoided, and signal deterioration due to the overlap is suppressed. It becomes possible.
[0016]
The output of the latch circuit 5 is converted into an audio signal by the D / A conversion circuit 7 (step g).
[0017]
If it is determined in step b that the upper limit of the frequency component of the fundamental wave is less than 10 kHz and no determination output is generated, the acoustic data read from the ROM 1 passes through the arithmetic circuit 6, but is subjected to arithmetic processing here. The value is output to the latch circuit 5 as it is without any change (step h).
[0018]
When the reading of the addresses 0 to 255 in the ROM 1 is completed, the reading is restarted from the address 0. When a signal designating the change of the pitch is received from the control means (not shown), the operation from step a is performed again. When a signal designating the end of the operation is received, the apparatus of the present example ends the operation (END). .
[0019]
As described above, according to the present example, in order to suppress in advance the specific frequency band on the harmonic side of the signal output from the ROM 1, the signal output from the latch circuit 5 has the center of the frequency band of the signal from the ROM 1 and the latch frequency. The frequency band generated as above does not overlap, and unnecessary components due to the overlap can be removed, and signal degradation can be suppressed.
[0020]
【The invention's effect】
According to the present invention, at a desired read frequency, a specific frequency band occurs in which a frequency band of a signal obtained by reading acoustic data stored in a ROM and a frequency band of a signal generated when the signal is latched are generated. In such a case, after a specific frequency band included in the signal read from the ROM is suppressed in advance, a latch and D / A conversion are performed to obtain an audio signal. That is, as the pitch increases, the harmonic component of the frequency component included in the audio data is suppressed by an amount corresponding to the rise of the pitch, so that the deterioration of the audio signal due to mixing of unnecessary frequency components can be suppressed. As a result, the upper limit of the pitch can be raised while suppressing signal deterioration.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram illustrating a configuration of an acoustic signal generating device according to an embodiment of the present invention.
FIG. 2 is an explanatory diagram for explaining a main part of FIG. 1;
FIG. 3 is an explanatory diagram for explaining the operation of FIG. 1;
FIG. 4 is a flowchart for explaining the operation of FIG. 1;
FIG. 5 is an explanatory diagram for explaining a configuration of a conventional technique.
FIG. 6 is an explanatory diagram for explaining the operation in FIG. 5;
[Explanation of symbols]
1 ROM (storage means)
2 Address counter (reading means)
4 control circuit (judgment means)
5. Latch circuit (latch means)
6 arithmetic circuit (conversion means)
7 D / A conversion circuit (D / A conversion means)

Claims (1)

Storage means for storing first sound data obtained by sampling sounds such as voices and musical sounds;
Reading means for reading first acoustic data from the storage means at a desired first frequency;
A first frequency band of a first signal obtained by reading out the first acoustic data at a first frequency, and a second frequency band of a signal obtained by latching the first signal at a predetermined second frequency. Determining means for determining whether or not there is a specific frequency band overlapping the second frequency band centered on the frequency of
When the determination means determines that the specific frequency band exists, the specific frequency band included in the first signal is suppressed, the first audio data is converted to the second audio data, and the second audio data is converted to the second audio data. Conversion means for outputting a second signal corresponding to the acoustic data;
Latch means for latching and outputting the first signal or the second signal at a second frequency;
A sound signal generator comprising: a D / A converter for D / A converting an output from the latch means to generate an audio signal.

Images