JPH0563660A - Synchronous method in voice coding communication - Google Patents

Abstract

(57) [Summary] [Purpose] To increase the amount of coded voice information in multi-mode coded voice coded communication in which parameters that characterize the phoneme of a voice are divided into a plurality of modes and encoded and transmitted. This improves the reproduced voice quality. A vowel or silent mode M0 capable of encoding with a small amount of information when encoding and transmitting voice signals in modes distinguished by phoneme for each frame.
The frame synchronization signal D1 is inserted into the frame No. 2 and the synchronization signal P to be inserted into the frame in the other mode is set to at least 2 bits necessary for detecting the synchronization deviation so that the encoded audio information is increased accordingly. ..

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to adaptive transform coding (AT).
The present invention relates to a synchronization method in voice coding communication by the C) method.

[0002]

2. Description of the Related Art FIG. 4 is an example of a frame structure for explaining a conventional synchronization method. In the figure, F31, F3
2, F33 and F34 indicate frames. Since the number of bits per frame differs depending on the encoding method, it is assumed here to be 100 bits as a specific example. S1, S2, S3
S4 is an 8-bit synchronization bit. In the case of this frame structure, 8 synchronization bits are used for each frame. The receiving side can correctly read the subsequent bit sequence by detecting this synchronization bit. If the synchronization bit cannot be correctly determined, the sequence of bits cannot be correctly recognized, so that the voice cannot be correctly decoded. Therefore, the role played by the sync bit is important.

[0003]

Since it is necessary for the sync bit to have a pattern that does not appear at a position other than the sync bit position, 5 or more sync bits are required. And always 8 sync bits per frame
Inserting bits means that 8% of the originally transmittable information amount is deleted. Further, once synchronization is achieved, it is not necessary to send 8 synchronization bits for each frame, since it is only necessary to detect the synchronization shift in the subsequent frames. Generally, about 2 bits are enough to detect the synchronization deviation.

The object of the present invention is to solve the drawbacks of the prior art by reducing the proportion of synchronization bits occupied in a frame, thereby increasing the information amount of speech coding when the number of bits per coding is constant. The present invention provides a synchronization method in voice coding communication that improves the sound quality and increases the amount of information to be transmitted at the same time when the sound quality remains constant.

[0005]

For example, when a speech signal is coded by an adaptive transform coding (ATC) method, the number of bits required to code it differs depending on the phoneme. Since they are not necessarily the same, different encoding processing methods can be used depending on the phonemes. At this time, if the encoding method is called a mode, encoding of modes 0 to N corresponding to N + 1 kinds of phonemes can be considered. Here, the speech coding in which corresponding modes are selected and coded according to the type of phoneme is called multi-mode speech coding. The synchronization system of the present invention is a synchronization system in multi-mode voice encoding communication in which voice signals in modes distinguished by phoneme type are encoded and transmitted for each frame, and the number of synchronization bits is reduced to provide encoded voice information. In order to increase the number, it is necessary to insert a predetermined sync signal for frame synchronization in the empty information part of the frame in the vowel or silent mode with a small amount of encoded information, and to detect the synchronization deviation in the frame in other modes. It is characterized in that at least a 2-bit sync shift detecting sync signal is inserted and transmitted.

[0006]

1 is a block diagram showing a circuit of an embodiment of the present invention. In this embodiment, vowels and silent parts are divided into three modes, that is, mode 0, consonants are mode 1, and nasal sounds are mode 2. In the figure, 13 is a transmitting side encoding device, and 14 is a receiving side decoding device. 1 is a mode determiner for determining the mode based on the type of phoneme of the input speech, 2 is a switch for switching the mode by receiving information from the mode determiner 1, 3, 4 and 5 are encoders for each mode, and 6 is a mode Reference numeral 20 denotes a multiplexer that multiplexes determination information and encoded data and sends the multiplexed data to a transmission path. Reference numeral 20 denotes a sync signal insertion circuit that inserts a 6-bit sync signal into the mode 0 encoder 3, and 21 denotes each frame. Is a synchronization signal insertion circuit for detecting synchronization deviation, which inserts a 2-bit synchronization signal for detecting synchronization deviation. In the decoding device 14, 15 is a synchronous circuit section, 7 is a separator, 8 is a mode selector for switching the switch 9 by receiving the information of the mode determination from the separator 7, and 9 is a selector for receiving the information from the mode selector 8. Switches for selectively switching the mode decoders 10, 11, 12 are decoders for the respective modes. FIG. 2 is a partial circuit block diagram of the present invention, and shows the details of the synchronization circuit unit 15 in the decoding device 14 of FIG. In FIG. 2, reference numeral 151 is a changeover switch, 152 is a synchronization circuit for establishing frame synchronization by a 6-bit synchronization signal and 2 bits of the next frame, 1
Reference numeral 53 is a sync shift detection circuit for detecting a sync shift by a 2-bit sync signal, and 154 is a switching controller of the switch 151. FIG. 3 is a frame configuration example diagram of the present invention, showing a state on the transmission path. M0 indicates a data string encoded by the mode 0 encoder 3 and mode selection information, and M1
Indicates a data string encoded by the mode 1 encoder 4 and mode selection information, and M2 indicates a data string encoded by the mode 2 encoder 5 and mode selection information. Also, F
21, F22, F23 and F24 are frames. D1
Is a frame synchronization bit, and P1, P2, P3 and P4 are synchronization deviation detecting bits.

Here, the mode 0 is set for the vowel and the silent part, but since the vowel and the silent part can be encoded with a small amount of information, a blank information part is formed in the rear part of the frame. Therefore, as shown in FIG. 3, a 6-bit sync bit D1 is inserted at the rear of the frame F21. S21 is (D1 + P2) = 8 bits, which is a synchronization bit for actually achieving frame synchronization. The operation of the present invention will be described in detail below with reference to the drawings. In FIG. 1, the mode determiner 1 determines the mode of each frame of the input signal a. b is mode selection information for selecting a mode. Based on this mode selection information b, the encoder 3,
Switch to any one of 4, 5 with switch 2 Input signal a
Is entered. The sync signal is input from the sync signal insertion circuit 20 to the mode 0 encoder 3. Then, the multiplexer 6 multiplexes the encoded audio data, the mode selection information b, and the sync shift detection sync signal from the sync shift detection sync signal insertion circuit 21. c is a data signal that is multiplexed and sent to the transmission path. The data signal c received via the transmission line is divided into audio data d synchronized by the synchronizing circuit unit 15 of the decoding device 14 and encoded by the separator 7 and mode selection information e. Then, the mode selector 8 switches the switch 9 so that the audio data d is decoded by the decoders 10, 1
It is decoded by any one of 1 and 12 and output as a reproduced audio signal f. The operation of the synchronization circuit unit 15 will be described with reference to FIG. First, the switch 151 is the synchronization circuit 15
The input signal c is connected to the second side, and the switching circuit 1 is operated when the synchronizing circuit 152 performs the synchronization verification and the frame synchronization is achieved.
By 54, the connection of the switch 151 is switched to the side of the synchronization deviation detection circuit 153 to detect the synchronization deviation and output the synchronized reception data to the separator 7. Synchronization circuit 152
Data is not output to the separator 7 while the synchronous processing is being performed at. The sync deviation detection circuit 153 continues pattern matching of the 2-bit sync signal, and when a sync deviation is detected, information is given to the switching controller 154 to switch the switch 151 to the sync circuit 1.
Switching synchronization processing is performed on the 52 side. The same operation is continued thereafter. FIG. 3 is a diagram showing a frame configuration example of the data signal c on the transmission path. A synchronization bit D1 is provided at the rear of the frame F21 of mode 0 having a small amount of information. When frame F21 is received, a total of 8 bits of synchronization bits S2 of a random pattern
Synchronize with 1. Since the silent portion is assigned to mode 0, the probability that the frame F21 will appear is high, and the receiving side can immediately synchronize. Then, in the other frames, the two-bit sync shift detection sync signal P1 of "01" or "10" added to the beginning of each frame.
It suffices to detect the synchronization deviation at P2, P3 and P4. As described above, by reducing the ratio of the synchronization bits in the frame, it is possible to increase the total amount of information.

[0008]

As described in detail above, by carrying out the present invention, when the coding speed is constant, a reproduced voice of higher quality can be obtained as compared with the conventional synchronous codec, and the conventional speech can be obtained. In the case of coding / decoding with the same voice quality, more information can be sent at the same time, which is extremely effective.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a block diagram of a partial circuit of the present invention.

FIG. 3 is a diagram showing a frame configuration example of the present invention.

FIG. 4 is a diagram showing a conventional frame configuration example.

[Explanation of symbols]

 1 Mode decision device 2 Mode selection switch 3 Mode 0 encoder 4 Mode 1 encoder 5 Mode 2 encoder 6 Multiplexer 7 Separator 8 Mode selector 9 Mode selection switch 10 Mode 0 decoder 11 Mode 1 decoding Device 12 Mode 2 Decoder 13 Coding Device 14 Decoding Device 15 Synchronous Circuit Section 20 Synchronous Signal Insertion Circuit 21 Synchronous Deviation Synchronous Signal Insertion Circuit 151 Switch 152 Synchronous Circuit 153 Synchronous Deviation Detection Circuit 154 Switching Controller

Claims (1)

[Claims] 1. In a synchronization method in multi-mode voice coding communication for encoding and transmitting a voice signal of a mode distinguished for each frame depending on a phoneme type, a frame synchronization synchronizing signal is added to a frame of a vowel or a silent mode. A synchronization method in voice coding communication, characterized in that at least 2 bits of a synchronization signal for detecting a synchronization deviation, which is necessary for detecting a synchronization deviation, is inserted and transmitted in a frame of another mode.