JPH0482419A - Audio signal encoder - Google Patents

Abstract

PURPOSE:To perform the compressive encoding of a digitized audio signal with a small scale of circuit by dividing an information source encoder into the high-order and low-order bit groups of the audio signal, and making it function only on each bit group. CONSTITUTION:A digitized audio input signal is inputted from an input circuit 10 to a signal divider 11 as a 16-bit parallel signal, and is divided into a high- order bit signal 21 and a low-order bit signal 22. The signal 21 is changed to a Huffman coder with variable length by the information source encoder 12 with low information loss, and is converted to a high-order bit signal 23. The signal is inputted to an output code comprising circuit 13 with the signal 22, and an output code 24 is which the amount of data per unit time is averaged can be comprised. Thereby, it is possible to reduce the bit processing of the information source encoder with low information loss occupying most part of the circuit of an audio signal encoder from conventional 16 bit processing to 8 bit processing by the half, and to perform the data compression with a simple circuit.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention is used to digitize high-quality audio signals for recording and playback, or to remove signal redundancy and perform data compression when digitizing and transmitting the signals. The present invention relates to an audio signal encoding device to be realized.

BACKGROUND ART In the field of audio, digitalization of recording/playback and signal transmission has now become common knowledge, as seen in compact discs, digital audio tape recorders, etc., and higher sound quality and longer duration are now required. These demands increase the amount of information that must be handled, and therefore, highly efficient audio signal encoding devices that can compress redundant data are important. The core of such audio signal encoding devices is the so-called information source encoder, which includes Huffman codes, Gib-Lembel codes, arithmetic codes, continuous length codes, etc. There are two types, such as those that have no information loss and therefore do not cause distortion to the audio signal, and those that have information loss and therefore do not cause distortion to the audio signal, such as the discrete cosine transform encoder using an adaptive transform coding method. It is well known that there are two types of accompaniments, each with their own advantages and disadvantages.

Generally, when handling audio signals, the above-mentioned information lossless information source encoders, especially Huffman encoders, are often used to minimize distortion.

For example, published patent publication No. 61-500998 describes an audio signal encoding device using a combination of a digital compression filter circuit that generalizes a differential filter and a Huffman encoder, and Japanese Patent Application Laid-open No. 57-20796 describes an audio signal encoding device using a Huffman encoder alone. A structured speech encoding device is disclosed.

An example of the conventional audio signal encoding device as described above will be described below with reference to the drawings. The seventh panel shows the configuration of a conventional audio signal encoding device, in which 70 is an input circuit for inputting a digitized audio signal, 71 is a Huffman encoder, 72 is a buffer circuit for configuring an output code, and 72 is a buffer circuit for configuring an output code. This is an output code configuration circuit.

The operation of the conventional audio signal encoding device configured as described above will be explained. First, the input circuit 70 inputs, for example, 16-bit parallel digitized audio and audio signals. This signal is input to a Huffman encoder 71 and compressed into a Huffman code that minimizes redundancy according to the frequency of appearance of the code. Since the obtained Huffman code is a variable length code, the amount of data per unit time will fluctuate if left as is, so it is unsuitable for output to a general recording/reproducing system or transmission system. Therefore, the amount of data per unit time is averaged by the puff circuit and output code configuration circuit 72, and then output as an output code.

Problems to be Solved by the Invention However, in the conventional audio signal encoding device as described above, since the digitized audio signal is encoded as it is, the standard quantization focus number of current digital audio equipment is In order to handle a digitized audio signal with 16 bits per sample or a digitized audio signal with a greater number of points of focus, the circuit scale becomes large, making it difficult to put it into practical use.

The present invention solves the above-mentioned problems, and aims to compress and encode digitized audio signals of 16 bits per sample, which is the current standard, and digitized audio signals of 20 bits or more per sample, which are expected in the future, with a small number of circuits. It is an object of the present invention to provide an audio signal encoding device that can be easily realized on a large scale.

Means for Solving the Problems In order to achieve the above object, the present invention has a first configuration in which a digitized audio signal composed of, for example, 16 bits per sample is divided into two groups of upper and lower bits. an information source encoder that performs compression processing only on the divided upper bit group without information loss;
The second configuration includes a buffer circuit and an output code configuration circuit that configure an output code by combining a lower bit group that is not subjected to compression processing and an upper focus group that has been subjected to compression processing. In addition to the above configuration, it is equipped with an information source encoder that performs compression processing on the divided lower bit groups, with information loss, and as a third configuration, a digital encoder consisting of, for example, 16 bits per sample. A bit divider that divides a converted audio signal for each focus, a binary information source encoder that performs compression processing on each divided bit serial signal, a buffer circuit, and an output code configuration circuit, A fourth configuration includes, in addition to the third configuration, a code conversion circuit from two's complement representation to signed binary representation.

Most of the circuitry of the operational audio signal encoding device is occupied by the information source encoder, and in the first and second configurations of the present invention, this information source encoder is Since the operation is applied only to the bit group or the lower bit group, the parallel bit width to be processed can be halved, and the circuit scale of the audio signal encoding device can therefore be significantly reduced.

In addition, in the third and fourth configurations, two very simple circuit configurations that process only 1-bit serial data are used.
Since the original information source encoder is used, an audio signal encoding device can be realized with a smaller circuit scale.

Embodiment Hereinafter, an audio signal encoding apparatus according to an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an audio signal encoding device according to a first embodiment of the present invention. Also, the second
The figure is also a schematic diagram showing the bit structure of a signal processed by the audio signal encoding device according to the first embodiment of the present invention. In the figure, lO is an input circuit for inputting a digitized audio signal, 11 is a signal divider connected to the input circuit 10, and 12 is an information lossless information source code connected to one output of the signal divider 11. 13 is a buffer circuit and an output code configuration circuit connected to the other output of the signal divider 11 and the information lossless information source encoder 12. As the information source encoder 12 without information loss, a Huffman encoder, a Ziv-Lempel encoder, an arithmetic encoder, a combination thereof, etc. can be considered, but here, a Huffman encoder will be used as an example. The operation of the audio signal encoding apparatus according to the first embodiment of the present invention configured as described above will be explained with reference to FIGS. 1 and 2. Here, as a typical example, a 16-bit parallel digitized audio signal is divided into upper 8 bits and lower 8 bits and processed. In FIG. 2, 20 is an input 16-bit parallel signal, 21 is a divided upper 8-bit signal, 22 is a lower 8-bit signal, 23 is a Huffman-encoded upper bit signal, and 24 is an output code. It is. First, a digitized audio signal, which is a human input signal, is input as a 16-pin parallel signal 20 by the input circuit 10 .

This 16-bit parallel signal 20 is input to a signal divider 11 and is divided into an upper 8-bit signal 21 and a lower 8-bit signal 22. Among these, the upper 8-bit signal 21 is manually inputted into an information source encoder 12, that is, a Huffman encoder, without information loss, and converted into an upper bit signal 23 which is Huffman-encoded with a variable length according to the appearance frequency. The Huffman-encoded upper focus signal 23 and the lower 8 focus signals 22 divided by the signal divider 11 are input to a buffer circuit and an output code configuration circuit 13, and the output code is an averaged data amount per unit time. 24 are configured.

The identification bit 25 included in the output code 24 is needed to identify the delimiter between the Huffman code part and the lower 8 bit part. If a reduction in data compression rate is a problem due to the insertion of redundant identification bits 25 for each output code, output codes 26, which are created by combining multiple output codes into broncs, or Huffman codes may be used. The converted upper bit signals are combined into one group,
Similarly, the lower 8-bit signals may be grouped together and the identification bit 25 may be inserted as a delimiter between both groups.

Now, if the configuration of the above embodiment is adopted, the data compression rate will be lower than that of a conventional audio signal encoding device that processes the entire 16-bit, 2-bit parallel digitized audio signal. There are concerns. However, according to the inventor's actual measurements, for a typical music signal quantized at 16 bits, the entropy, which is an index of information compression by an information source encoder without information loss, is calculated for the entire 16 bits and the upper 8 bits. The results shown in Table 1 were obtained. If we use a conventional audio signal encoding device using an ideal source encoder with no information loss from this data, 10.6/16 = 0.6625, that is, about 66% of the data before encoding. This means that it can be compressed.

On the other hand, in the audio signal encoding device according to the first embodiment of the present invention, since it is sufficient to add the lower 8 bits that are not compressed to the entropy for the 8 bits, (2,7+8)/16=0.66875, that is, approximately 67%
It can be seen that there is almost no difference from conventional audio signal encoding devices. In this way, according to this embodiment, the information lossless information source encoder, which occupies most of the circuitry of the audio signal encoding device, can be halved from conventional 16-bit processing to 8-bit processing, which is extremely simple. This makes it possible to realize an excellent audio signal encoding device that can obtain a data compression rate equivalent to that of conventional circuits.

(Margin below) Table 1 A second embodiment of the present invention will be described below with reference to the drawings.

The third example is a block diagram showing the configuration of an audio signal encoding device according to a second embodiment of the present invention. This embodiment has a configuration that improves the compression ratio by allowing some distortion to occur. In FIG. 3, 30 is an input circuit, 31 is a signal divider, 32 is an information source encoder without information loss, and 33 is a buffer circuit and an output code configuration circuit, which are the same as those in the first embodiment. Input circuit 10. Signal splitter 11 Information lossless source encoder 12. This is equivalent to the buffer circuit and output code configuration circuit 13.

Further, 34 is an information source encoder with information loss.

The configuration of FIG. 3 is almost the same as the configuration of FIG. 1, but the lower bit signal output from the signal splitter 31 is converted to an information source encoder 34 with information loss, for example, by discrete cosine transform coding using an adaptive transform encoding method. It is configured such that the converted encoded output is inputted to the buffer circuit and output code configuration circuit 33. With this configuration, the upper bit signal can be compressed without distortion as in the first embodiment, and at the same time, the lower bit signal, which was not subjected to any compression processing in the first embodiment, can be compressed. is inputted to the information source encoder 34 with information loss and converted into a compressed lower focus signal, and then sent to the buffer circuit and the output code configuration circuit 33.
The input code is input to form the output code. According to this embodiment, although the compression ratio can be significantly improved, information loss occurs during compression processing of the lower focus signal, so distortion is unavoidable. However, since no distortion occurs in the upper focus signal, the overall distortion is slight.

Next, a third embodiment of the present invention will be described with reference to the drawings.

FIG. 4 is a block diagram showing the configuration of an audio signal encoding device according to a third embodiment of the present invention. This embodiment shows an example of a configuration using a dual information source encoder that processes only 1-pin serial data.

In FIG. 4, 40 is an input circuit, 41 is a bit divider connected to the input circuit, 42 is a binary information source encoder group connected to each bit output of the bit divider, 43 is a buffer circuit and an output code. This is a component circuit. As a binary information source encoder, a continuous length encoder, which is the simplest, or a Ziv-Lempel encoder, which can achieve a high compression rate but is somewhat complicated, can be used. The operation of the third embodiment of the present invention configured as described above will be explained. Here again, the explanation will be made assuming 16 pins in parallel as the input signal. First, a digitized audio signal as an input signal is input to the device by an input circuit 40 as a 16-bit parallel signal. This 16-bit parallel signal is divided into 16 bits by a bit divider 41.
After being divided into 1-bit serial signals of the system, 16 2-bit serial signals are
Binary information source encoder group 42 composed of original information source encoders
is input. The 16 compressed 1-pinto serial signals outputted from the binary information source encoder group 42 each have a different amount of data per unit time, so they are input to the buffer circuit and output code configuration circuit 43 and are outputted from After averaging the amount of data, it is constructed and output as an output code.

According to this embodiment, it is possible to realize an audio signal encoding device with a very simple circuit configuration because it uses a dual information source encoder that processes only 1-bit serial data.

Note that in the configuration of FIG. 4, a binary information source encoder group 42 consisting of 16 binary information source encoders is used, but a modification of the configuration of FIG. 4 as shown in FIG. With this configuration, an equivalent audio signal encoding device can be realized with a single binary information source encoder.

In FIG. 5, 50 is an input circuit, 51 is a focus divider, and 54
is a buffer circuit group consisting of 16 buffer circuits, 55
is a selection circuit, 52 is a single dual source encoder, and 53 is a buffer circuit and output code configuration circuit. In the configuration shown in FIG. 5, instead of compressing the output from the focus splitter in parallel using a plurality of binary information source encoders, the bit serial signal from the bit splitter 51 is sent to the buffer circuit group 54. is stored, and the stored bit serial signals are sequentially read out using a selection circuit 55 and inputted to a single binary information source encoder 52 for compression processing. An output code will be constructed.

A fourth embodiment of the present invention will be described below with reference to the mouth surface.

FIG. 6 is a block diagram showing the configuration of an audio signal encoding device according to a fourth embodiment of the present invention. In this embodiment, the data compression rate is improved by adding some circuits to the third embodiment. The structure aims to improve the In FIG. 6, 60 is an input circuit, 64 is a code conversion circuit connected to the input circuit, 61 is a bit divider, 62 is a binary information source encoder group, and 63 is a buffer circuit and an output code configuration circuit. be.

Among these, the input circuit 60. Focus splitter 612 Original information source encoder group 62, buffer circuit and output code configuration circuit 6
3 is a circuit completely equivalent to the input circuit 40, the bit divider 4l, the binary information source encoder group 42, the buffer circuit, and the output code configuration circuit 43 in the third embodiment, and the input circuit 60 and the bit divider The configuration is such that a code conversion circuit 64 is newly inserted between 61 and 61. In general, the most standard code representation for digitized audio signals is two's complement representation, and the compact discs and digital audio recorders that are currently widely used also use two's complement representation. This two's complement input signal inputted by the input circuit 60 is inputted to the code conversion circuit 64 and converted into a binary number representation with positive and negative signs. Table 2 shows an example of the correspondence between two's complement representation and signed binary representation for some numbers. The digitized audio signal subjected to such conversion is input to the bit divider 61, undergoes the same processing as in the third embodiment, and is output as an output code. By performing the conversion as in this example, it is possible to significantly reduce the appearance frequency of 1, especially in the upper bits, and for example, it is possible to significantly reduce the appearance frequency of 1 in the upper bits. A high compression ratio can be achieved even in the case of

(Blank below) Effects of the Invention As is clear from the above embodiments, the present invention can accommodate digitized audio signals of 16 bits per sample, which is the current standard, and 20 bits per sample, which is expected in the future. It becomes possible to easily provide an excellent audio signal encoding device that performs highly efficient compression encoding of a bit-digitized audio signal with a small circuit scale.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an audio signal encoding device according to the first embodiment of the present invention, and FIG. 2 is a schematic diagram showing the bit structure of a signal processed in the first embodiment of the present invention. FIG. 3 is a block diagram showing the configuration of an audio signal encoding device according to a second embodiment of the present invention, and FIG. 4 is a block diagram showing the configuration of an audio signal encoding device according to a third embodiment of the present invention. FIG. 5 is a block diagram showing a second configuration of an audio signal encoding device according to a third embodiment of the present invention, and FIG. 6 is an audio signal encoding-dio signal encoding according to a fourth embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of the device. 10... Input circuit, 11... Signal divider, 12... Information source encoder without information loss, 1
3...Hanwha circuit and output code configuration circuit, 3
4... Information source encoder with information loss, 40...
...Input circuit, 41...Bit divider, 4
2... Binary information source encoder group, 43...
Hasofa circuit and output sign configuration circuit, 64...
Signal conversion circuit.

Claims (4)

[Claims] (1) An input circuit that inputs a digitized audio signal sampled at a predetermined frequency, quantized with a predetermined number of bits N, and converted into a digital signal; a signal divider that divides the configured digitized audio signal into upper m bits and lower N-m bits; and information whose input is the upper m bits of the digitized audio signal divided by the signal divider. a lossless information source encoder; and a buffer configured to receive an output of the lossless information source encoder and the lower N-m bits of the digitized audio signal divided by the signal divider to form an output code. An audio signal encoding device comprising: a circuit and an output code configuration circuit. (2) comprising an information source encoder with information loss that receives as input the lower N-m bits of the digitized audio signal divided by the signal divider, and the output of the information source encoder with information loss and the information loss; 2. The audio signal encoding apparatus according to claim 1, wherein the output of the information source encoder without the information source encoder is arranged to be input to the buffer circuit and the output code configuring circuit. (3) an input circuit that inputs a digitized audio signal sampled at a predetermined frequency, quantized with a predetermined number of bits N, and converted into a digital signal; a bit divider that divides the configured digitized audio signal bit by bit to obtain a total of N series of bit series signals, and inputs the N series of bit series signals output from the bit divider. An audio signal encoding device comprising: N binary information source encoders, and a buffer circuit and an output code configuration circuit that input the outputs of the N binary information source encoders and configure an output code. . (4) A code converter for inputting the output of the input circuit from two's complement representation to signed binary representation indicating positive or negative, and arranged so that the output of the code converter is input to the bit divider. The audio signal encoding device according to claim 3.