JPH07254922A - Voice storage device - Google Patents

Abstract

PURPOSE:To attain storage of voice with high quality by discriminating a speech head of a frictional sound such as the sa-series in 50 phonetic sounds as a silent and increasing only speech head voice recording information so as to increase a voice recording information quality thereby preventing interruption of speech head. CONSTITUTION:This device is provided with a frame storage section 5 storing a newest positive integer frame tentatively in addition to a silence detection section 1 and a voice recording section 2. When the decision of the silence detection section 1 indicates a change from a silence into a sound, a newest positive integer frame stored in the frame storage section 5 is added to a head of a frame decided to be a sound frame and the resulting frame is stored in the voice recording section 2, and even when a speech head of a frictional sound is decided to be a silence, interruption of the speech head is prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice storage device which is used for recording a solid-state element such as an answering machine system and which stores digitally encoded input voice data.

[0002]

2. Description of the Related Art In recent years, in an answering machine, etc., a call received when an answering machine is accepted has been recorded using a solid-state element instead of the conventional recording of a magnetic tape. In the recording of a conversation voice by this solid-state element, the voice signal is digitally encoded, the silence of the digitally encoded voice signal is detected, and the voice information excluding the silence portion is recorded. As a result, more audio data can be recorded without requiring a large recording capacity of the recording medium.

FIG. 3 is a block diagram showing the configuration of a conventional voice storage device which detects silence and stores voiced data. In FIG. 3, in this example, a silence detector 1 that detects a silence part from the input audio signal Sa to determine whether there is sound or no sound, and a voice recording unit 2 that records the sound signal Sb after the presence or absence of sound determination.
And an audio reproducing unit 3 for transmitting an output audio signal Sd which reproduces the audio signal Sc recorded by the audio recording unit 2. Further, a control unit 4 is provided for transmitting a voice recording control signal Ca and a voice reproduction control signal Cb to the voice recording unit 2 and the voice reproducing unit 3 from the presence / non-presence judgment information Se from the silence detecting unit 1, respectively. .

Next, the operation of this conventional example will be described. First, in the case of recording, the silence detector 1 calculates the power of each frame of the input audio signal Sa. If the power is large in this power calculation, it is determined to be voiced, and if the power is low, it is determined to be silent. And the silence detector 1
The audio signal Sb to which the presence / absence determination is added is output to the audio recording unit 2. At the same time, the presence / absence determination information Se of the frame is output to the control unit 4.

The control unit 4 counts the number of silent frames,
The audio recording control signal Ca is output to the audio recording unit 2 so as to record the audio data of the number of silent frames and the sound determination frame. The voice recording unit 2 receives the voice recording control signal Ca that is input.
Then, the number of silent frames and the sound determination frame of the audio signal Sb are recorded. In the case of reproduction, the audio reproduction control signal Cb is output from the control unit 4 to the audio reproduction unit 3, and the audio reproduction unit 3 outputs from the audio recording unit 2 the audio signal Sc, which is the recording information, from the number of silent frames to the frame time. Outputs silence and reproduces voiced frames as they are.

[0006]

However, in the sound storage device of the conventional example, the sound detection section 1 detects a sound part from the input sound signal Sa to judge whether there is sound or not, only the power information of the frame. There was a problem in that the beginning of a fricative such as the Japanese syllabary was judged to be silent, and the beginning of the talk was cut off, making it impossible to store high-quality speech.

The present invention solves such a conventional problem, and determines that the head of a fricative sound such as the Japanese syllabary is silent and does not increase the amount of voice recording information. It is an object of the present invention to provide an excellent voice storage device capable of preventing voice loss by increasing only voice recording information at the beginning of a voice and enabling high quality voice storage.

[0008]

In order to achieve the above object, a voice storage device according to a first aspect of the present invention detects the power of digitally encoded voice data and determines whether there is voice or silence. A frame is determined to be voiced by the presence / absence determination unit, the frame storage unit that temporarily stores the latest positive integer frame in the digitally encoded voice data, and the presence / absence determination unit. In this case, a voice recording means for adding and recording a positive integer frame stored in the frame storage means at the beginning of the frame is provided.

In the voice accumulating apparatus according to the present invention, the voiced / non-voiced determination means determines voice data when the power exceeds a preset voiced / non-voiced determination value as a voiced frame, and if not, the voice data. Is a silent frame, and
A positive integer frame in a silent frame before the sound frame when the frame storage means changes from the sound frame to the sound frame and a sound frame after the positive integer frame are recorded as voice data.

[0010]

In the voice accumulating device according to the present invention having this configuration, only the latest positive integer frame of the input voice data is temporarily stored, and when the determination result of the silence detection changes from silence to voice, The latest stored positive integer frame,
Audio data is recorded at the beginning of the voiced judgment frame. In this case, when the power exceeds the voiced / non-voiced determination value, the voice data is determined to be the voiced frame, and when the power does not exceed the voiced / non-voiced determination value, the voice data is determined to be the voiceless frame, and the frame storage means changes the voiceless frame to the voiced frame. A positive integer frame in a silent frame before the voice frame when it changes and a voice frame after the positive integer frame are recorded as voice data. Therefore, the head of a fricative such as the Japanese syllabary is determined to be silent, and only the voice recording information of the head increases without increasing the amount of voice recording information.
You will be prevented from breaking the talk.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a voice storage device of the present invention will be described in detail below with reference to the drawings.

In the following texts and drawings, the same constituent elements shown in FIG. 3 are given the same reference numerals.

FIG. 1 is a block diagram showing the configuration of an embodiment of a voice storage device of the present invention. In FIG. 1, in this example, a silence detector 1 that detects a silence part from an input voice signal Sa and determines whether or not there is a voice, a voice recorder 2 that records a voice signal Sb after the voice / sound determination, and this voice An audio reproduction unit 3 is provided which outputs an output audio signal Sd which is an reproduction of the audio signal Sc recorded by the recording unit 2. Furthermore, the silence detector 1
From the presence / absence determination information Se of the presence / absence of sound, the control unit 4 for transmitting the voice recording control signal Ca and the voice reproduction control signal Cb to the voice recording unit 2 and the voice reproducing unit 3, respectively, and the input voice signal Sa from the silence detecting unit 1. A frame storage unit 5 is provided which outputs the frame signal Sf that has been input and has been temporarily stored.

Next, the operation of this embodiment will be described. In the case of recording, the silence detection unit 1 calculates the power of each frame of the input audio signal Sa, and if the power is large in this power calculation, it is determined to be voiced, and if the power is low, it is determined to be silent. At the same time, the presence / absence determination information Se is temporarily and sequentially stored in the frame storage unit 5 only for the latest positive integer frame (which will be referred to as α frame if necessary). The sound detection unit 1 outputs the sound signal Sb to which the sound / sound determination is added to the sound recording unit 2, and the sound / sound determination information Se of the frame is output to the control unit 4. The control unit 4 continues to count the number of silence frames when the silence frames continue, and when the determination changes from silence to sound, the number of silence frames- (minus) α A voice recording control signal for recording the voice data in the voice recording unit 2 by adding the α frame stored in the frame storage unit 5 to the beginning of the The Ca is output to the voice recording unit 2.
In the audio recording unit 2, the audio recording control signal Ca is input, and “the number of silent frames−α”, the α frame of the temporarily stored frame signal Sf, and the sound determination frame of the audio signal Sb are input. To record. Next, in the case of reproduction, the audio reproduction control signal Cb is output from the control unit 4 to the audio reproduction unit 3, and the audio reproduction unit 3 receives the audio signal Sc of the recording information from the audio recording unit 2 to display a “silent frame”. The silence is output for the frame time from "number-α", and the recorded frame is reproduced as it is.

FIG. 2 is a diagram showing voice power transitions in the embodiment. In FIG. 2, a portion whose power exceeds the sound / non-sound determination value K is voice data determined to be a voice frame, and a portion not exceeding the power is voice data determined to be a voice frame Fa. The α frame in the silent frame Fa before the sound frame Fb when the sound frame Fa is changed to the sound frame Fb and the sound frame Fb after the α frame are recorded as sound data in the sound recording unit 2 (recorded sound The recording range is m).

As described above, in this embodiment, the silence detector 1
When the input voice signal Sa changes from silence to voice by the determination of, the latest positive integer frame stored in the frame storage unit 5 is added to the beginning of the frame determined to have voice, When the head of a fricative such as the Japanese syllabary, which is stored in the voice recording unit 2, is judged to be silent, the head break is prevented.

[0017]

As is apparent from the above description, the voice accumulating apparatus according to the first and second aspects temporarily stores only the latest positive integer frame of the input voice data, and the silence detection result is silence. In the case where the voice data is changed from to voice, the latest positive integer frame stored is added to the head of the voice determination frame to record the voice data. This makes it possible to judge that the beginning of a fricative, such as the Japanese syllabary, is silent, and increase only the voice recording information of the voice without increasing the amount of voice recording information, which can prevent breakage of the voice. This has the effect of enabling quality voice storage.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of a voice storage device of the present invention.

FIG. 2 is an explanatory diagram showing power transition of voice in the embodiment.

FIG. 3 is a block diagram showing a configuration of a conventional voice storage device.

[Explanation of symbols]

 1 Silence Detection Section 2 Audio Recording Section 3 Audio Playback Section 4 Control Section 5 Frame Storage Section Sa Input Audio Signal Sb, Sc Audio Signal Sd Output Audio Signal Se Spoken / No Sound Judgment Information Sf Frame Signal Ca Audio Recording Control Signal Cb Audio Playback Control signal

Claims (2)

[Claims] 1. A voiced / non-voiced determination means for detecting the power of digitally encoded voice data and determining whether there is voiced or non-voiced, and always the latest positive data in the digitally encoded voice data. If the frame is judged to be sound by the judgment by the sound / non-sound judgment means, a positive integer frame stored in the frame storage means at the beginning of the frame A voice storage device including a voice recording unit for adding and recording. 2. The sound / silence judging means judges the sound data when the power exceeds a preset sound / sound judgment value as a sound frame, and judges the sound data when the power does not exceed the sound frame as a sound frame. Further, the present invention is characterized in that a positive integer frame in a silent frame before a sound frame when the frame storage means changes from a sound frame to a sound frame and a sound frame after the positive integer frame are recorded as voice data. The audio storage device according to claim 1.