JP2023067673A - musical tone generator - Google Patents

Abstract

To solve the problem that, in a musical tone generator having sound source waveform storage means, a timbre can be changed by providing filter means when reproducing data from a waveform memory, but simply controlling a cut-off frequency of a low-pass filter or a high-pass filter in a conventional manner results in monotonous timbre changes.SOLUTION: Timbre-changing means arranged in a musical tone generator uses reproduction frequency information selected by pressing a key on a keyboard to control a filter which is means for controlling the volume of harmonic components and inter-harmonic components and extracts the harmonic components and the inter-harmonic components.SELECTED DRAWING: Figure 1

Description

A musical tone generator characterized by generating a variety of timbre changes by controlling the volume of overtone components and interharmonic components.

In general, an electronic piano stores waveform data obtained by recording musical tones of 88 keys of an acoustic piano in corresponding 88 waveform memories. is playing. However, since the recorded waveform is reproduced, the timbre cannot be greatly changed from the timbre at the time of recording.

With a musical tone generator having sound source waveform storage means, it is possible to change the timbre when playing back the data in the waveform memory by providing a filter means. Tone changes become monotonous.

A musical tone generator with sound source waveform storage means. A filter that extracts overtone components and inter-harmonic components is introduced to change the tone color when reproducing data in the waveform memory, and the volume of the extracted overtone components and inter-harmonic components is changed. A variety of timbre changes are generated by
For control of the filter for extracting overtone components and interharmonic components, reproduction frequency information selected by pressing a key on the keyboard is used.

For example, a piano sound is characterized by the striking sound of a hammer striking a string at the time of attack. Therefore, by extracting this striking sound from the recorded waveform data with a filter, for example, by increasing the volume of the striking sound at the time of attack and adding it to the original sound, it is possible to change the tone color to emphasize the feeling of attack.

In general, the struck sound of a piano and the bowed sound of a violin exist between overtones, and pitch information is necessary to extract the component.
On the other hand, in a musical tone generating apparatus having sound source waveform storage means, the reproduction frequency of waveform data selected by pressing a key on the keyboard is the same as the reproduction frequency selected by pressing a key on the keyboard. In other words, the pitch period of the waveform data selected by pressing a key on the keyboard is the same as the pitch period of the reproduction frequency information selected by pressing a key on the keyboard.

The present invention is a musical sound generator characterized in that various timbre changes can be obtained by emphasizing or attenuating striking sounds and bowing sounds of musical sounds.

1 shows a block diagram of an embodiment of the present invention; FIG. The concrete processing of the DSP indicated by the dotted line in the embodiment [FIG. 1] is shown. FIG. 2 shows a waveform diagram for explaining the operation of the DSP indicated by the dotted line in the embodiment [FIG. 1].

FIG. 1 is a block diagram of an embodiment of the present invention, comprising a keyboard 101, a sound source waveform storage unit 102, a sound source readout unit 203, a filter unit 204, a direct signal volume 105, an effect signal volume 106, an addition unit 107, and a DAC 108. .

The keyboard 101 is a piano keyboard having 88 pieces, and outputs the pressed key number Note Number and the key pressing/key release information Note On/Off of the keyboard. These pieces of information are sent to the sound source reading section 203 and the filter section 204 .
The sound source waveform storage unit 102 stores recorded data of piano sounds in areas corresponding to each of 88 piano keyboards.
The sound source reading unit 203 reads the recorded data from the sound source waveform storage unit 102 according to the Note Number information. The recorded data read from the sound source reading section 203 is sent as a Direct signal to the Direct signal volume 105 and also to the filter section 204 .
The filter unit 204 is described here as a comb filter, but various other filters can be used.
The filter unit 204 obtains the pitch period from the note number information from the keyboard 101 , extracts interharmonic components by means of a comb filter controlled by the pitch period, and sends the extracted interharmonic components as an effect signal to the effect signal volume 106 .
After the Direct signal and the Effect signal are added by the adding section 107, the DAC 108 generates an analog signal.

The mixing ratio of the Direct signal and the Effect signal added by the adder 107 may be controlled from the operation panel. You can give it purposefully.

FIG. 2 is a flow chart showing specific processing of the DSP indicated by the dotted line in the embodiment [FIG. 1], and this routine is called every sampling period of this system. In the following explanation, it is assumed that "Middle-C" is pressed on the keyboard.

When a key is pressed on the keyboard 101, the Note On information becomes "Y", and the sound source reading section 203 reads the recorded data from the sound source waveform storage section 102 according to the Note Number information. In the example, "Middle-C" is pressed and Note Number=60, so the waveform memory 60 is selected as the area to be read.

この計算式は音階の周波数が平均律で定義され、Ｎｏｔｅ Ｎｕｍｂｅｒ＝６９の鍵盤の再生周波数が４４０．０Ｈｚである事を利用している。例では”Ｍｉｄｄｌｅ－Ｃ”が押鍵された場合でありＮｏｔｅ Ｎｕｍｂｅｒ＝６０なので、本システムのサンプリング周波数が３２ｋＨｚの場合のピッチ周期は１２２．３１２０サンプルとなる。
なお、この計算を行わずに予め求めた値をテーブルとして用意して置き、これを参照する様にしても良い。Next, filter section 204 obtains a pitch period (sample) from the Note Number information. A pitch period is the number of samples in one period of a waveform. The pitch period can be obtained by the following formula, where fs (Hz) is the sampling frequency of this system.

This calculation formula utilizes the fact that the frequency of the scale is defined by equal temperament, and the reproduction frequency of the note number=69 keyboard is 440.0 Hz. In the example, when "Middle-C" is pressed and Note Number=60, the pitch period is 122.3120 samples when the sampling frequency of this system is 32 kHz.
Instead of performing this calculation, it is also possible to prepare a table of values obtained in advance and refer to this table.

The subsequent routine is repeated during the Note On period until the waveform memory is read out to the final address.
Since the comb filter uses data read out from two locations in the waveform memory, ReadAddress1 is set to "0" and ReadAddress2 is set to the pitch cycle value "122.3120" obtained earlier as an initial value.
In the next step, the waveform data read from the waveform memory 60 using ReadAddress1 as an address is stored in WaveData1, and the waveform data read using ReadAddress2 as an address is stored in WaveData2.
Then, the value obtained by subtracting WaveData2 from WaveData1 is stored as the output EffectData of the comb filter.
WaveData1 is sent as a Direct component to the Direct signal volume 105 in FIG. 1, and EffectData is sent to the Effect signal volume 106 as an Effect component.
ReadAddress1 and ReadAddress2 are incremented each time this program routine is called.

FIG. 3 shows the operating waveforms of the DSP indicated by the dotted line in the embodiment [FIG. 1].
FIG. 3(3-1) is a waveform read from the waveform memory 60 at ReadAddress1 and corresponds to WaveData1.
FIG. 3(3-2) shows a waveform read from the waveform memory 60 at ReadAddress2, which corresponds to WaveData2. It can be seen that the WaveData1 waveform and the WaveData2 waveform are shifted by one cycle.
FIG. 3(3-3) shows a waveform obtained by subtracting WaveData2 from WaveData1, which corresponds to EffectData.
FIG. 3(3-4) shows the frequency characteristics of WaveData1, and overtones can be confirmed.
FIG. 3(3-5) shows the frequency characteristics of EffectData from which overtone components have been removed.

The present invention can be applied not only to piano sounds but also to various musical sounds. For example, the sound of a violin is characterized by a bowed string sound, and this bowed string sound is taken out in the filter section. A soft tone can be obtained.

Also, in the embodiment, the comb filter for extracting interharmonic noise components is used as the filter unit, but if a comb filter for extracting overtone components is added, for example, musical tones with few percussive sound components and bowing sound components can be generated. (The comb filter for extracting overtone components can be realized by adding WaveData1 and WaveData2, as explained using symbols in the embodiment.)

101... keyboard, 102... sound source waveform storage section, 203... sound source reading section, 204... filter section, 105... direct signal volume, 106... effect signal volume, 107. addition unit 108 DAC

Claims (1)

The timbre changing means provided in the device is means for controlling the volume of overtone components and inter-harmonic components, and the reproduction frequency selected by pressing a key on the keyboard is used to control the filter for extracting the overtone components and inter-harmonic components. A musical tone generator that uses information.

Images