JP2001125582A - Method and device for voice data conversion and voice data recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device which enables a user to experience vocals well with his or her singing voice. SOLUTION: The voice data conversion device is provided with music accompaniment data and a model voice data, and voice data obtained by user's singing in accordance with the accompaniment and the model voice is compared with model voice data in a prescribed unit with respect to voice level and interval, and voice data is inversely corrected by difference data to realize a vocal approximating the model with user's singing voice.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention
In particular, the present invention relates to a voice data conversion device that converts even a person who is relatively poor at singing into a good singing voice by voice data conversion processing.

[0002]

2. Description of the Related Art Conventionally, there is no special audio data processing device that converts even a relatively unskilled person into a slightly better singing voice by voice data processing. One person sings the same song more than once, records them, plays them back, replaces the worse part with the better part, discards the worse part, A method of composing and processing a single song using only good parts was considered.

[0003]

In the above-mentioned song processing method in which only the good parts are taken and connected and finished, the tendency of the poor parts is almost the same, and it is difficult to give an inferiority, and the tendency before and after the connection is the same. It requires a lot of skill to make connections, and unnaturalness tends to be noticeable at the replaced connection parts, and the whole song flow has not been smooth.

[0004]

In order to solve the above-mentioned problems, the invention according to claim 1 comprises accompaniment data 1 of a music piece and sample audio data 2 sung by the accompaniment. Speech data conversion for performing vocal performance by correcting voice data sung along with the synthesized sound through a level difference detection processing means and a pitch correction processing means, and adding the accompaniment data to the corrected voice data. An apparatus, wherein the level difference detection processing means compares the audio data with the sample audio data in a predetermined unit of a sound level, and detects a difference in sound level, and a level difference detection processing circuit 7; A level correction processing circuit 8 that reversely corrects the input audio data by the difference data of the output of the level difference detection processing circuit to match the sound level of the sample data; The pitch correction processing means includes a pitch comparison processing circuit 10 that compares the output of the correction processing circuit with the model data in a predetermined unit, and a pitch of the input voice data based on difference data of the output of the pitch comparison processing circuit. And a pitch correction processing circuit 11 for performing pitch correction so as to match the sample data with the model data.
The invention of the present invention is an audio data conversion method comprising: accompaniment data of a song and sample audio data sung by the accompaniment, and correcting the audio data sung according to the synthesized sound of the accompaniment and the example data. Comparing the sound level of the sound data with the sample sound data in a predetermined unit to detect a difference in the sound level, and inversely correct the input sound data based on the difference data to thereby correct the sound level of the sample data. And comparing the sample data with the pitch in a predetermined unit, and performing the vocal performance by adding the accompaniment data by adjusting the pitch of the input voice data to the sample data based on the difference. Conversion method. According to a third aspect of the present invention, in the audio data conversion apparatus according to the first aspect, the sample data previously stored and the newly created audio data are divided into packets in units of n times a measure having a header. A voice data conversion apparatus characterized in that the voice data is converted by the voice data conversion apparatus according to the first or third aspect of the present invention. An audio data conversion device characterized in that the unit for comparing with the audio data is set to one sound unit of the musical score;
According to a fifth aspect of the present invention, in the audio data conversion apparatus according to the first or third aspect, one of the audio data
A sound data conversion device is provided in which the division of a sound unit is determined based on the length of the note data of the music and the lyrics data, and the invention of claim 6 is described in claim 1. In the audio data recording medium on which the output signal of the audio data conversion device is recorded, together with the output signal of the audio data conversion device, the output signal of the one sound division level difference detection processing circuit 7 and the output signal of the pitch comparison processing circuit 10 And an audio data recording medium characterized by recording at least one of them.

[0005]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An audio data conversion apparatus and method according to the present invention will be described below with reference to FIGS. As shown in FIG. 1 and FIG.
Accompaniment data recording / reproducing device 1, sample audio data recording / reproducing device 2, audio data mixing processing circuit 3, vocal vocalization device 4 sung to the mixed voice, AD conversion / packetization processing circuit 5, buffers 6, 9, 12,
It comprises a one-sound division level difference detection processing circuit 7, a level correction processing circuit 8, a pitch (pitch) comparison processing circuit 10, a pitch correction processing circuit 11, and an audio data mixing processing circuit 13.

First, an embodiment of the present invention will be described below with reference to FIG. The audio data with accompaniment obtained by synthesizing the accompaniment data 1A and the reference audio data 2A stored in the system of the present invention by the mixing processing circuit 3 is output.

While listening to the audio data with accompaniment via headphones (not shown), the vocal is sung to the microphone 4 in accordance with the audio data, and the audio data is converted into digital audio by the AD conversion / packetization processing circuit 5. I do.

A header is attached to the digital audio in units of measures, packetized, and temporarily stored in a buffer 6 for adjusting the timing to the next processing. The stored voice data is output in packet units, and the reference data stored in the form of a packet of voice data based on the musical score information and lyrics information within the range of the voice data is divided by the voice data in the one-sound division level difference detection circuit 7. Then, the level difference between them is detected.

Using the level difference data 7A, the level correction processing circuit 8 performs level correction on the audio data 6A supplied from the buffer 6, and adjusts the audio data 6A in units of one sound to the level of the reference audio data. , And outputs the audio data 8A.

[0010] Audio data 8A divided and level corrected
Are temporarily stored in the buffer 9 together with the information divided for each sound, and the sound data 9A is output to the pitch comparison processing circuit 10 for each sound.

The pitch comparison processing circuit 10 receives one-packet reference audio data 2A from the reference (Ref) audio data recording / reproducing device 2 and uses the one-sound divided data 9A stored in the buffer 9 to input one packet. The pitch is compared with the reference voice data 2A divided for each sound.

The pitch deviation 10A is corrected by the pitch correction processing circuit 11, temporarily stored in the buffer 12, synthesized with the accompaniment data 1A by the synthesizer 13, and the pitch and sound level are corrected with the accompaniment. This is a device that outputs vocals.

By recording the output with an MD (or cassette tape) recording device and distributing the MD or the like, it can be used as a music data correction (conversion) device for business use.

At the time of recording, together with the output signal (output 16) of the audio data converter, the output signal (output 14) of the one tone division level difference detection processing circuit 7 and the output signal (output) of the pitch (pitch) comparison processing circuit 10 At least one of the outputs 15) is recorded.

In this way, when the same vocal is generated next time, the reproduced level and pitch correction signal, which is the difference between the last sung song and the sample song, is monitored by the monitor T.
Singing while watching with V can be a reference for better singing.

In this block, since the level detection processing circuit 7 and the pitch comparison processing circuit 10 compare the fundamental wave level and the pitch of the fundamental wave, the level detection processing circuit 7 and the pitch comparison processing circuit 10 Filter on 10 inputs
(Not disclosed) can increase the processing accuracy.

The reason why the headphone 4 is used for vocal vocalization while listening not only to the accompaniment data 1A but also to the reference audio data 2A is to make it as close as possible to the reference audio data 2A.

Next, another embodiment of the audio data converter of the present invention will be described with reference to FIG. The embodiment shown in FIG. 2 and the above-described embodiment described with reference to FIG. 1 are different from the level correction-related blocks 7 to 8 and the pitch correction-related blocks 10.
The difference is that the positions of 順序 12 are reversed in order, and other partial configurations are almost the same.

The audio data with accompaniment obtained by synthesizing the accompaniment data 1A and the reference audio data 2A stored in the system of the present invention by the mixing circuit 3 is output.

While listening to the audio data with accompaniment via headphones (not shown), the vocal is sung to the microphone 4 in accordance with the audio data, and the audio data is converted into digital audio by the AD conversion / packetization processing circuit 5. I do.

A header is attached to the digital audio in units of measures, packetized, and temporarily stored in a buffer 6 for adjusting the timing to the processing of the next stage.

The stored audio data is output in packet units, and the pitch comparison processing circuit 10 outputs a reference (Ref).
The one-packet reference audio data 2A from the audio data recording / reproducing device 2 is input, and the pitch is compared with the reference audio data 2A divided into one sound unit using the one-sound divided data 6A stored in the buffer 6.

The pitch correction processing circuit 1 uses the deviation 10A.
1 and temporarily stores the data in the buffer 12 and then sends the sound data 12
A is output.

The stored voice data is output in units of packets, and the one-sound division level difference detection circuit 7 converts the reference data stored in the form of packets into voice data based on musical score information and lyric information within the range of the voice data, and stores the reference data in one voice. Each level is divided and their level difference is detected.

Based on the level difference data 7A, the level correction processing circuit 8 performs level correction on the audio data 12A supplied from the buffer 12, and adjusts the audio data 12A in units of one sound to the level of the reference audio data. , And outputs the audio data 8A.

Audio data 8A divided and level-corrected
Is a device that temporarily stores in the buffer 9 together with the information divided for each sound, synthesizes it with the accompaniment data 1A and the synthesizer 13, and outputs its own vocal with the accompaniment and the pitch and sound level corrected. .

By recording the output of the audio data conversion device together with the correction data by an MD (or cassette tape) recording device and distributing the MD or the like, it can be used as a music data correction (conversion) device for business use.

At the time of recording, together with the output signal (output 16) of the audio data converter, the output signal (output 14) of the one tone division level difference detection processing circuit 7 and the output signal of the pitch (pitch) comparison processing circuit 10 (Output 15).

By doing so, when the same vocal is generated next time, it is possible to sing while watching the reproduced level and pitch correction signal on the monitor TV, and it is possible to use the singing data as reference data.

An embodiment of a method for separating sound data for each sound in the sound data converter of the present invention will be described below with reference to FIG. In the separation of sound data for each sound, a header is added and the sound data is separated for each measure, and the packetized data is divided as one unit for each sound included in the measure. The division process for each sound is performed on the sample data and the music data to be corrected.

The measure is made with the timing and length made at the tempo of the accompaniment, and as shown in FIG.
In the first measure, one sound starts from the rising part of the sound from silence, and the change of the signal is determined until the rising of the next one sound.

In order to minimize judgment errors, score information (pitch and length) and lyric information (words) of the music are prepared, and first, the outline of the separation point is determined according to the length of each note. Know the point of, look for the next sound rise near that point.

Further, by extracting a part of the next rising sound that has been found and recognizing that the word uttered from the lyric information is the word by pattern recognition, it is possible to determine a division point without errors. You may make it.

In this way, while dividing one sound and one sound, the level processing for adjusting the level of the model data and the audio data and the pitch processing for adjusting the pitch (pitch) are performed to correct each sound. . By performing each of the above processes, even a person who is relatively poor at singing can correct and convert the singing voice into a good singing voice by a voice data conversion process that brings the voice data closer to the model voice data.

[0035]

As described above, according to the present invention,
Even a person who is relatively poor at singing can correct and convert the voice data into a good singing voice by voice data conversion processing that brings the voice data closer to the model voice data. Also, MD disks and tapes on which the output of the audio data converter and the correction data are recorded can be handed over for entertainment use for business use. This M
It is also possible to practice singing better by playing a D disk or tape and watching the correction data.

[Brief description of the drawings]

FIG. 1 is a system block diagram showing an embodiment of an audio data converter according to the present invention.

FIG. 2 is a system block diagram showing another embodiment of the audio data converter of the present invention.

FIG. 3 shows audio data 1 of the audio data converter of the present invention.
FIG. 4 is a drawing illustrating a separation for each sound.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Accompaniment data recording / reproducing apparatus 1A Accompaniment data 2 Model audio data recording / reproducing apparatus 2A Model (reference) voice data 3 Voice data mixing processing circuit 4 Microphone (vocal vocal device sung in accordance with mixing voice) 5 AD conversion processing / packet Processing circuit 6, 9, 12 Buffer 6A, 8A, 12A Audio data 7 1-tone division level difference detection processing circuit 7A Level difference data 8 Level correction processing circuit 10 Pitch (pitch) comparison processing circuit 10A Pitch deviation 11 Pitch correction processing Circuit 13 Audio data mixing processing circuit 14 Level correction signal output (terminal) 15 Pitch correction signal output (terminal) 16 Converter output signal (terminal)

Claims (6)

[Claims] An audio data sung in accordance with the synthesized sound of the accompaniment and the sample data is provided with level difference detection processing means and pitch correction. An audio data conversion device for performing level and pitch correction via processing means and adding the accompaniment data to the corrected audio data to perform a vocal performance, wherein the level difference detection processing means converts the audio data By comparing the sound level with the sample audio data in a predetermined unit,
A level difference detection processing circuit that detects a difference in sound level, and difference data of an output of the level difference detection processing circuit,
A level correction processing circuit that reversely corrects the input voice data to match the sound level of the sample data, wherein the pitch correction processing means compares the output of the correction processing circuit with the sample data in a predetermined unit. And a pitch correction processing circuit that corrects the pitch of the input voice data according to the difference data of the output of the pitch comparison processing circuit so that the pitch of the input voice data matches the model data. An audio data conversion apparatus comprising an audio data mixing processing circuit for performing a vocal performance by adding the accompaniment data to the audio data to which the audio data has been applied. 2. A voice data conversion method comprising: accompaniment data of a music piece; and sample voice data sung by the accompaniment, and correcting the voice data sung in accordance with the synthesized sound of the accompaniment and the sample data. The audio data and the sample audio data are compared in sound level in a predetermined unit to detect a difference in the sound level, and the input data is inversely corrected based on the difference data to obtain the sound of the sample data. The sound is characterized in that the sample data and the pitch are compared in a predetermined unit in accordance with a level, and the pitch of the input voice data is matched with the sample data based on the difference, and vocal performance is performed by adding accompaniment data. Data conversion method. 3. The audio data conversion apparatus according to claim 1, wherein the pre-stored sample data and newly created audio data are divided into packets in units of n times a bar having a header and converted. An audio data conversion device characterized by processing. 4. The audio data conversion device according to claim 1, wherein the unit for comparing the audio data with the model audio data is a unit of one note of a musical score. Conversion device. 5. The audio data conversion device according to claim 1, wherein the division of the audio data in units of one sound is determined based on the length of the note data of the music and the lyrics data. An audio data conversion device characterized in that: 6. An audio data recording medium on which an output signal of the audio data conversion device according to claim 1 is recorded, together with an output signal of the audio data conversion device and an output signal of the one sound division level difference detection processing circuit. And an output signal of the pitch comparison processing circuit.