JPS6319918A - Voice decoding system - Google Patents

Abstract

PURPOSE:To compensate the expected value of the data on the bits which are not transferred and to suppress deterioration of signals, by adding '1' to a digit lower by a step than the lowest digit of the code data produced by the semi-instantaneous compression when this data is decoded. CONSTITUTION:The coding data DL is applied to a demultiplexer 11 and head three bits are identified as the scale value SC for each block. This value SC is applied to the scale value input terminal of a semi-instantaneous expanding part 12 and other code data (compression data) are applied to the code data input terminal of the part 12 respectively. The pat 12 divides the received code data every four bits and at at the same time distributes these 4-bit data at the bit positions of the 8-bit data corresponding to the input data SC. Then the contents of the code data are set to the digits higher than the code data together with '1' set to the digit lower by a step and '0' set to the next lower digit respectively. Thus the 8-bit coding voice data is decoded and then undergoes D/A conversion 13.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field] The present invention relates to an audio decoding method for decoding coded data subjected to quasi-instantaneous compression.

[Prior Art] For example, when digitally processing an audio signal, such as transmitting an audio signal using a high-speed digital line or storing and synthesizing the audio signal for a voice response device, the audio signal is converted into a digital signal by some method. need to be converted to .

Basically, audio signals have a frequency band of 0.3 to 3.4KH.
z analog signal, and to convert it to a digital signal, the sampling frequency must be 1, for example, 8 K! Converting can be done using an analog/digital converter with 1z and 8-bit resolution (PCM (Pulse Code Module)
ation) encoding).

Then, to convert this digital signal back to the original audio signal,
The signal can be converted into an analog signal using a digital/analog converter with a sampling frequency of 8 Hz and a resolution of 8 bits, and then passed through a low-pass filter to shape the waveform. At this time, the resolution of the analog/digital converter and digital/analog converter In other words, the larger the bit width of the PCM code, the higher the quality of the reproduced audio.

By the way, such a PCM encoded audio signal has 1
The bit rate (data rate; hereinafter referred to as bit rate) per second is 6 · i Kbps, and transmitting audio signals with such a high bit rate requires an extremely high-speed transmission line, and Storing such audio signals requires a memory with a huge storage capacity. Therefore, various proposals have been made to reduce the bit rate of audio signals.

One of them is a differential PCM encoding method that forms a difference between chronologically adjacent PCM codes. This differential PCM encoding method utilizes redundancy based on the correlation of audio waveforms, and changes in values between samples that are 9I adjacent are often contained within a limited dynamic range. Therefore, the number of bits per sample can be reduced. CCITT (
The adaptive differential PCM method (ADPCMf) is based on the recommendations of the International Telegraph and Telephone Advisory Committee.

It achieves a bit rate of 32Kbps.

In addition, APC-AB (Adaptive Predication), which utilizes the non-stationarity and linear predictability of audio signals,
tion Co-ding with Adaptiv
e Bit A11 location) method, or LSP (Line Spectrum) method using speech analysis and synthesis method.
umPair) method.

However, such ADPCMf method, APC-
The AB method and the SP method have disadvantages in that the encoding and decoding processes are very complicated, and the equipment for implementing them is very expensive.

Furthermore, one of the high-quality PCM audio transmission systems for broadcasting satellites is the quasi-instantaneous companding system. This quasi-instantaneous companding method is
Divide M-encoded audio data into a predetermined number of blocks in time series, identify scale data representing the most significant digit corresponding to the maximum signal absolute value in each block, and include the most significant digit. This formats data of a predetermined number of bits into code data, and the encoding process is relatively simple, and the number of bits of one sample can be easily reduced.

However, when such a quasi-instantaneous companding method is applied to audio data with a low bit rate, the effect of untransmitted bits is large, and the deterioration of the reproduced audio signal may be large, so it is not efficient enough. This caused the inconvenience of not having one.

[Purpose] The present invention was made to solve the above-mentioned disadvantages of the conventional technology, and its purpose is to provide an audio decoding method that can reproduce high-quality audio with a low bit rate and simple processing. It is said that

[Configuration] When decoding coded data formed by quasi-instantaneous compression, the present invention compensates for the expected value of the data of the bits that are not transmitted by adding the expected value of the signal value of the missing bits of the coded data. , thereby suppressing the deterioration of the reproduced audio signal and achieving the above-mentioned purpose.

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

First, the principle of the present invention will be explained.

For example, an audio signal with a waveform as shown by the solid line in Figure 1(a) is sampled at a sampling frequency of 8 kHz, and each sample is converted into 8-bit digital data (PC
Let's consider converting the digital data to M code) and then semi-instantaneous companding of this digital data. In addition, the conditions for quasi-instantaneous companding are 1
It is assumed that the number of samples constituting a block is 8, each sample is compressed to 4 bits, and the scale value is 3 bits. Note that the 8-bit digital data is digital data written as a two's complement number, and the most significant digit is a sign bit (sign bit for identifying positive/negative).

Here, eight samples #1 forming one block
It is assumed that the digital data of .about.#8 are obtained as shown in FIG. 1(b).

In this block of 8 samples.

Since the data of sample 7 has the maximum absolute value, the most significant digit position of that bit pattern, that is, bit 4, is set to the quasi-instantaneous companding scale position g pos.

Then, for each sample #1 to #8, the lower four digits from bit 5, which is one digit higher than this scale position, are extracted as transmission bits, that is, code data. Therefore, at the MSB (most significant digit) of the code data, there is a sign bit (sign bit) that indicates whether the code data is positive or negative.

As a result, encoded data obtained by quasi-instantaneous companding is obtained, as shown in FIG. 3(C). That is, data at scale position pos is placed in the first 3 bits, followed by 4 bits of code data extracted in sample all-18.

When decoding this encoded data to reproduce an audio signal, first, the scale position pos is identified from the contents of the first three bits.

Next, the following data is sequentially divided into code data of 4 bits each, and arranged in the 8-bit data so that the MSB of the code data is located one digit higher than the previously identified scale position PO8. However, by placing the contents of the MSB of the code data (in other words, the contents of the sign bit) in the upper digits of the code data, and placing 0 in the lower digits of the code data, 8-bit audio can be generated. The data is expanded (see (d) in the same figure).

Then, by converting this audio data from digital to analog and shaping the waveform with a low-pass filter.

An audio signal is played.

Now, according to this kind of quasi-instantaneous companding, O is inserted into the lower digit bits than the transmission bits and decoded.
In this case, the decoded audio data (hereinafter referred to as decoded audio data) has a quantization width larger than the original audio data by the amount of the lower 2 bits that are missing from the original 8-bit audio data (hereinafter referred to as original audio data). Become.

In other words, while the quantization characteristics of the original audio data during encoding are as shown by the solid line in Figure 2(a), the quantization characteristics of the decoded audio data are as shown by the broken line in the same figure. become.

As a result, the data in the part of the original audio data that cannot be expressed using the quantization width of the decoded audio data is expressed as a lower level in the decoded audio data as shown by the arrow in the figure. The data is shifted in the negative direction as a whole compared to the audio data.

For this reason, the waveform of the audio signal reproduced based on the decoded audio data moves in the negative direction compared to the original audio, as shown by the -dotted chain line in Figure 1(a), and the sound quality deteriorates.

Therefore, in the present invention, in order to prevent such deterioration of sound quality, decoded audio data is formed by adding the value of the least significant digit 172 of the encoded data, that is, the expected value of the signal value of the missing bit.

That is, according to the above-mentioned quasi-instantaneous companding, the data of the lower digits of the transmitted bits are missing, but the value of the data of the missing bits is, for example, when the missing bits are 2 bits (
It takes any value from 00)z to (11)2.

Therefore, the expected value of this missing bit is (10)2, and by adding this expected value, it is possible to make the decoded audio data closer to the original audio data compared to when they are not added, which improves the sound quality of the reproduced audio signal. can be improved. That is, the quantization characteristics of the decoded audio data at this time are improved as shown by the broken line in FIG. 2(b).

When the encoded data is decoded in this way, the result is shown in Figure 1 (
Decoded audio data as shown in Figure 1(e) can be obtained from the encoded data in c), and the reproduced audio signal based on this decoded audio data is as shown by the broken line in Figure 1(a). The result is a sound that is closer to the original sound.

FIG. 3 shows a speech encoding device according to an embodiment of the present invention. 1g in this audio encoding device! The conditions for instant companding are as described above.

In the figure, an input audio signal SS is band-limited by a low-pass filter 1, and then applied to an analog/digital transformer 2 with a sampling frequency of 8 and converted into an 8-bit digital signal DS. The digital signal DS is applied to a quasi-instantaneous compression section 4 via a buffer memory 3 having a storage capacity of 8 samples, and a scale value setting section 5 for setting a scale value for quasi-instantaneous compression. has been added to.

The scale value setting unit 5 sets the digital signal O5 to 8 consecutive times.
Among the samples, the one with the largest absolute value is identified, the most significant digit of its bit pattern is determined, and the bit position is output as 3-bit scale data DK.

This scale data DK is applied to one input terminal of the quasi-instantaneous compression section 4 and a multiplexer 6 for shaping one block of data into a predetermined signal format.

The quasi-instantaneous compression unit 4 has 8 bits added from the buffer memory 3.
Regarding the bit digital signal O5, for each sample, 3-bit data with the MSB being one bit higher than the scale position indicated by the scale data DK added from the scale value setting section 5 is extracted, and this is converted into compressed data DC. It is output to the multiplexer 6 as .

As shown in FIG. 4, the multiplexer 6 arranges the scale data D output from the scale value setting unit 5 at the beginning, and then sequentially arranges the compressed data DC of each sample to generate a signal. , and outputs it as one block of encoded data DL to a next-stage device (for example, a data transmission device or a data storage device, etc.).

In this way, the input audio signal SS is quasi-instantaneously compressed and output as encoded data DL.

FIG. 5 shows an example of an audio decoding device according to an embodiment of the present invention. This audio decoding device decodes the encoded data DL encoded by the audio encoding device described above using the decoding process described above, and outputs an audio signal.

In the figure, encoded data DL output from a pre-stage device (shown in the figure) such as a data receiving device or a data storage device is applied to a demultiplexer 11, and the first three bits of each block are converted into a scale value. The other code data (compressed data) is identified as SC and added to the scale value input terminal of the quasi-instantaneous decompression unit 12.
It is applied to the code data input terminal of the quasi-instantaneous decompressor 12.

The quasi-instantaneous decompression unit 12 divides the added code data into 4 bits each and divides the inputted scale data SC into
The 4th bit position in the 8-pit data corresponding to
Decode the 8-bit decoded audio data by placing the code bit data in the upper digit of the code data, placing 1 in the lower digit, and placing 0 in the lower digit. Then, this decoded audio data is output to the digital/analog converter 13.

The digital/analog converter 13 converts the received signal value into 8
It is converted into a corresponding analog signal (level signal) at a conversion frequency of KHz, and this is output to the low-pass filter 14. This analog signal is waveform-shaped by the low-pass filter 14 and then output as a reproduced audio signal to a subsequent device (for example, an audio output device).

In this way, the configuration of the audio decoding device for decoding encoded data according to the present invention is extremely simple. Therefore, for example, this audio decoding device can be realized using a general-purpose 8-bit microprocessor, and the cost can be kept to an extremely low level.

Note that the various constants such as the number of bits in the embodiments described above are merely examples, and appropriate values can be set.

[Effects] As explained above, according to the present invention, when decoding coded data formed by quasi-instantaneous compression, by adding 1 to one digit lower than the least significant digit of the coded data,
It compensates for the expected value of the data of the bits that are not transmitted, thereby suppressing the deterioration of the reproduced audio signal.
With low bit rate and simple processing, you can achieve the excellent effect of reproducing high-quality audio.

[Brief explanation of the drawing]

1(a)-(e) are detailed explanatory diagrams of the present invention, FIGS. 2(a) and (b) are signal arrangement diagrams for detailed explanation of the present invention, and FIG. 3 is a detailed explanatory diagram of the present invention. FIG. 4 is a block diagram showing an audio encoding device according to an embodiment, FIG. 4 is a signal arrangement diagram showing an example of encoded data, and FIG. 5 is a block diagram showing an audio decoding device according to the present invention. . 1.14...Low pass filter, 2...Analog/
Digital converter, 3... Buffer memory, 4... Quasi-instantaneous compression section, 5... Scale value setting section, 6... Multiplexer, 11... Demultiplexer, 12...
Quasi-instantaneous expansion section, 13...digital/analog converter. 7・'1・ Agent Patent Attorney Crest 1) Makoto 1 Figure 1 (a) Figure 1 (b) (c) Figure 1 (e) O5 Jar

Claims (1)

[Claims] Scale data representing the most significant digit corresponding to the maximum absolute value of the signal value in each block formed by dividing PCM encoded audio data into a predetermined number of blocks in time series, and the most significant digit from each audio data. In an audio decoding method that decodes quasi-instantaneous compressed data formed by sequentially arranging compressed code data formed by extracting data of a predetermined number of bits including the upper digits, each code data is placed in the most significant digit represented by the scale data. The code data is arranged in a predetermined number of bits of data after decoding with the most significant digits of An audio decoding method characterized in that a 1 is placed one digit below the digit where the least significant digit of code data is located, and a 0 is placed in the lower digits.