JPH0477320B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a musical tone synthesis device using a so-called phoneme piece editing synthesis method, and can be used in electronic musical instruments that require sounds similar to the original instrument.

Configuration of conventional examples and their problems In general, phoneme segment editing synthesis methods extract representative phoneme data in pitch units from the original musical tone signal based on strong similarities between adjacent waveforms of musical tones. The phoneme piece data is edited and a desired musical tone signal is synthesized by sequentially connecting the phoneme piece data while repeating it a plurality of times according to musical tone synthesis control information. Also, in such a method,
Because the waveforms and pitch periods of phonemes change rapidly at the connection points of different phonemes, periodic noises occur due to the repetition of phonemes, making it difficult to obtain a smooth musical tone signal. To solve this problem, attempts have been made to insert an interpolated phoneme segment obtained by interpolation calculation between two phoneme segments.
That is, in the phoneme segment interpolation method that is compatible with the conventional phoneme segment editing and synthesis method, the k-th waveform C _k of the interpolation waveform C that is repeated j times between the two phoneme segments A and B to be interpolated is expressed by the following equation (1 ).

C _ki = (B _i −A _i )k/(j+1)+A _i ...(1) i=1,2,...,N k=1,2,...,j However, each phoneme has N pieces It consists of elemental piece data, and let the respective elemental piece data be A _i , B _i , and C _ki .
However, in such a phoneme segment interpolation method, all operations including addition, subtraction, multiplication, and division are required in order to calculate the interpolated phoneme segment C _ki , so there is a problem in that the time load on the musical tone synthesis device becomes large.

Such a technique is disclosed in, for example, Japanese Patent Application Laid-Open No. 78599/1983.

OBJECTS OF THE INVENTION The present invention solves the above-mentioned conventional drawbacks, and an object of the present invention is to provide a musical tone synthesis device for an electronic musical instrument that has a simple configuration and is more faithful to the original instrument sound.

Configuration of the Invention The present invention is configured such that a part of the first term on the right side of the above equation (1) is stored in a memory as differential phoneme data D _i . According to this configuration, the calculation required to calculate the interpolated phoneme piece data C _ki as shown in the following equation (2) can be performed with only one addition process. Therefore, it has the advantage that the time load on the musical tone synthesizer itself can be greatly reduced.

C _ki = D _i + C _(k-1)i ...(2) where C _pi = A _i D _i = (B _i −A _i )/(j+1) i=1, 2, ..., N k=1 , 2, ..., (J+1) j indicates the number of repetitions as in the conventional example.

DESCRIPTION OF EMBODIMENTS An embodiment of the musical tone synthesis device of the present invention will be described below. FIG. 1 shows an embodiment of the present invention, in which a control section 1 to which a musical tone output instruction signal is inputted, controls a phoneme memory section 3 of synthesis information (phoneme piece data or differential phoneme piece data) corresponding to a musical tone to be output. (the start address, the distinction between phoneme pieces or differential phoneme pieces, and the number of repetitions) are read from the synthesis information memory section 2. If the read synthesis information instructs the use of a phoneme piece, N phoneme piece data is read from the phoneme memory unit 3 from the start address corresponding to the read synthesis information and directly transferred to the temporary storage unit 5 as a digital musical sound. Output as a waveform. If the read synthesis information instructs the use of differential phoneme pieces, N differential phoneme data are read from the phoneme memory section 3 from the start address corresponding to the read synthesis information, and the difference phoneme data and the immediately preceding The N digital musical sound waveforms written in the temporary storage unit 5 that have been output to
After performing this process twice, it is outputted to the temporary storage unit 5 as a digital musical tone waveform. After repeating the above operation a number of times corresponding to the read composite information, the next composite information memory is read and the same process is continued until an output stop instruction signal is input to the control section 1. On the other hand, at the same time as the output instruction signal is input to the control section 1, it generates a timing signal with a cycle corresponding to the pitch to be output, and outputs this to the temporary storage section 5 as a digital musical sound waveform output cycle instruction signal. , the output cycle generating section 4 performs a function of sequentially outputting the N digital musical tone waveforms stored in the temporary storage section 5.

According to this embodiment, the repetition circuit for phoneme units or differential phoneme units directly contributes to the compression rate. That is, if the number of repetitions is 4 on average, a compression ratio of 4 can be obtained. By increasing the number of repetitions, it is possible to greatly compress data regarding a considerable number of types of musical sound waveforms.

In addition, the upper rows a and c of FIG. 2 show the stored waveforms A and B necessary for synthesis by conventional phoneme segment synthesis, and their combined waveforms, and the lower rows b and d show the stored waveforms necessary for synthesis according to this embodiment. Waveform A and differential phoneme D
and their combined waveforms are shown. Note that in order to obtain the composite waveform of B in figure d, waveform B may be stored as a phoneme segment waveform, or the waveform can be obtained by interpolating the number of repetitions j (=2) by one extra time. You may also try to get B.

Effects of the Invention As described above, according to the present invention, since the difference value in units of pitch periods is stored as a difference phoneme, the calculation required for synthesis is only addition, and the calculation load on the control unit is reduced. Since it is extremely small, it is possible to construct a system using a general-purpose microcomputer, and a musical tone synthesizer capable of producing high-quality musical tones can be easily constructed.

[Brief explanation of drawings]

FIG. 1 is a block configuration diagram showing one embodiment of the present invention, and FIG. 2 is an explanatory diagram showing the formation of a composite waveform in the same embodiment and a conventional example. 1...Control unit, 2...Composition information memory unit, 3...
...Phoneme memory section, 4... Output cycle generation section, 5...
Temporary memory.

Claims (1)

[Scope of Claims] 1. A phoneme segment consisting of N pieces of data in one period unit, and a differential phoneme piece consisting of N pieces of data for obtaining an interpolated phoneme piece that smoothly connects the phoneme pieces. A phoneme memory for storing information on synthesis; a synthesis information memory for storing information regarding synthesis; a temporary storage section for storing N pieces of data to be output as a musical sound waveform; and a synthesis information memory for storing synthesis information corresponding to the musical tone to be output. reads a phoneme piece or a differential phoneme piece specified by the synthesis information from the phoneme memory, and reads a phoneme piece or an interpolated phoneme piece composed of N pieces of data according to the synthesis information into the temporary memory. and an output cycle generating section that instructs an output timing to sequentially output the N pieces of data stored in the temporary storage section as musical waveforms, the control section outputting the data to the reading section. Outputting a phoneme segment consisting of N pieces of data specified by the synthesis information to the temporary storage section in order to output it as a musical sound waveform, or outputting the phoneme segment consisting of the N pieces of data specified by the read synthesis information to the temporary storage section. and the corresponding N pieces of data stored in the temporary storage unit, respectively, and output an interpolated phoneme segment composed of N pieces of data to the temporary storage unit in order to output it as a musical sound waveform. A musical tone synthesis device characterized in that the above steps are repeated for each period as many times as specified by the read synthesis information.