KR19990056485A - Precompressor of digital audio - Google Patents

Abstract

The present invention is to provide a digital audio precompressor for transmitting only pure audible signals by applying a temporal masking effect for each frequency band before encoding the digital audio signal. A band division filter for dividing the signal by frequency band, buffering means for storing the audio signal output from the band dividing filter for each frequency band, and forward and backward inaudible regions for each frequency band of the audio signal output from the buffering means Verifying the validity of the audio signal for each frequency band output from the buffering means for each frequency band by referring to the non-audible region information provided by the non-audible region calculating means and the non-audible region calculating means Means Yeokbyeol synthesized frequency band audio signal is transmitted from the configuration parts for outputting synthesized. Therefore, audio coding efficiency can be improved without deteriorating audible sound quality.

Description

Precompressor of digital audio

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital audio transmission system, and more particularly, to a digital audio precompression apparatus that compresses by applying a time-masking effect for each frequency band before encoding a digital audio signal to be transmitted.

In order to provide high quality audio, a technique of converting and transmitting an analog audio signal to a digital signal has been proposed. However, an audio signal converted into a digital signal has a disadvantage in that a high data rate and a large memory are required since the data amount is increased in comparison with an analog audio signal. In order to improve this, encoding techniques for compressing the transmitted audio signal have been proposed.

The proposed audio coding technique generally uses the statistical characteristics of the original signal to efficiently remove the redundancy of the digital signal, thereby saving the characteristics of the transmission path without deteriorating the signal, and using the human auditory characteristics. Therefore, unnecessary or less important information is removed. By using the correlation of the acoustic signal and human hearing characteristics properly, a large amount of digital data can be saved. However, this degree is not satisfactory, and research to reduce the amount of data to be transmitted continues.

The present invention has been made in accordance with the above-described trend, and provides a digital audio precompressor for transmitting a pure audio signal by applying a temporal masking effect for each frequency band before encoding the digital audio according to a conventional method. There is a purpose.

In order to achieve the above object, the digital audio precompressor according to the present invention is an audio system having a function of encoding to compress a digital audio signal, wherein a predetermined digital audio signal is applied to a predetermined number of frequency bands. A band dividing filter for dividing and transmitting; Buffer means for storing the audio signal output for each frequency band from the band split filter for each frequency band; Inaudible region calculating means for calculating anterior and backward inaudible regions for audio signals output from the buffer means for each frequency band; By verifying the validity of the audio signal output from the buffer means for each frequency band by referring to the forward and backward inaudible region information of the corresponding frequency band provided from the inaudible region calculating means, and outputting only the audio signal verified as valid Verification means; And a synthesizing unit for synthesizing the audio signals of the respective frequency bands output from the verifying unit and encoding the audio signals.

1 is a block diagram of an apparatus for precompressing digital audio according to the present invention,

FIG. 2 is an exemplary diagram illustrating a relationship between an inaudible region calculated by the inaudible region calculators shown in FIG. 1 and an audio signal output from the buffer units.

<Description of the symbols for the main parts of the drawings>

100: band split filter 111, 112, 113: first to 32nd buffer unit

121, 122, 123: First to 32nd inaudible region calculating unit 140: Synthesis unit

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of a digital audio precompressor according to the present invention, in which a band split filter 100 for dividing an applied audio signal into 32 subbands and an audio signal for each frequency band output from the band split filter 100 are shown. Frequency forward and backward ratios for the audio signals output from the first to the 32nd buffer units 111 to 113 and the first to the 32nd buffer units 111 to 113 for the respective frequency bands, respectively, for storing in each frequency band. The first to thirty-second inaudible region calculators 121 to 123 for calculating the blue region and the non-audible region information for each frequency band provided from the first to thirty-second inaudible region calculators 121 to 123 may be used. First to thirty-two verification units 131 to 133 and first to thirty-two verification units that perform transmission processing by verifying validity of audio signals for each frequency band output from the first to thirty-second buffer units 111 to 113 ( 131 ~ 133) Synthesizing each frequency band yeokbyeol audio signal consists of the combination unit 140 for transmission to the non-illustrated audio encoder. As can be seen from the above-described configuration, the first to thirty-second buffer units 111 to 113 are buffer means, and the first to thirty-second inaudible region calculators 121 to 123 are inaudible region calculation means. 32 verification units 131 to 133 may be defined as verification means.

The digital audio precompressor configured as described above is operated as follows.

First, the band dividing filter 100 is composed of a digital filter such as a quadrature mirror filter (QMF), and separates and outputs digital audio applied for each frequency band. In this embodiment, since the case of dividing into 32 bands is illustrated, the applied digital audio signal of one PCM type is separated and output for each of 32 frequency bands. With this frequency band separation output, an audio signal may not exist in some frequency bands. The audio signal separated and output for each frequency band is transmitted to the corresponding buffer units 111, 112, and 113.

That is, as shown in FIG. 1, since the buffer units 111, 112, and 113 allocated to each frequency band are configured to be connected to the band split filter 100, the band split filter 100 is divided by frequency band and output. The audio signal to be transmitted is transmitted to the buffer unit 111 or 112 or 113 allocated to process the audio signal of the corresponding frequency band. For example, when the first buffer section 111 is provided for the lowest frequency band, the second buffer section 112 for the next frequency band, and the thirty-second buffer section 113 is provided for the highest frequency band, band division is performed. The audio signal of the lowest frequency band output from the filter 100 is transmitted to the first buffer unit 111, the audio signal of the next frequency band is transmitted to the second buffer unit 112, and the highest frequency band is the audio signal. Is transmitted to the thirty-second buffer unit 113.

As described above, the first to thirty-second buffer units 111 to 113 store and output the audio signal of the corresponding frequency band for a predetermined period of time. The number of buffer units 111, 112, and 113 provided may be changed by the number of divided bands of the band split filter 100. That is, in the present embodiment, since the number of divided bands formed by the band split filter 100 is 32, the number of buffer units 111, 112, and 113 provided is 32, but the audio signal to which the band split filter 100 is applied In the case of using a filter for dividing the frequency band into four, the number of buffer sections 111, 112, 113 provided is four.

When the audio signals for each frequency band output from the band split filter 100 are stored in the first to thirty-second buffer units 111 to 113, the first to thirty-third inaudible region calculators 121 to 123 correspond to the corresponding frequency bands. An inaudible region for tetragonal and posteriority of the audio signal to be output from the buffer units 111 to 113, which store the audio signal, is calculated. That is, as shown in FIG. 2, when the audio signal to be output from the first to the 32nd buffer units 111 to 113 is the audio signal at the time f (t), the first to the 32nd inaudible region calculator ( 121 to 123 are omnidirectional by the backward inaudible region A1 and the audio signal f (t + m) after m hours by the audio signal f (tm) preceding m hours in the corresponding frequency band. Each of the inaudible regions A2 is calculated.

Therefore, when the audio signal of the lowest frequency band at the time f (t) is output from the first buffer unit 111, the first inaudible region calculator 121 is connected to the audio signal of the lowest frequency band at the time f (tm). The backward inaudible region A1 and the forward inaudible region A2 by the audio signal of the lowest frequency band at the time f (t + m) are calculated. The second to thirty-second inaudible region calculating units 122 and 123 are also driven in the same manner as the first inaudible region calculating unit 121 to calculate forward and backward inaudible region information of the corresponding frequency band at the corresponding time point, respectively. do.

At this time, each inaudible region information is calculated based on the sound pressure of the audio signal. That is, the backward inaudible region information is calculated based on the sound pressure of the audio signal in the corresponding frequency band at f (tm) time, and based on the sound pressure of the audio signal in the corresponding frequency band at f (t + m) time. Anterior inaudible region information is calculated. To calculate the inaudible region information, the first to thirty-second inaudible region calculators 121 to 123 are configured to include a table storing the inaudible region information corresponding to each sound pressure. Therefore, when the audio signal stored in the buffer unit of the corresponding frequency band is applied, the inaudible region is calculated by referring to the table based on the sound pressure.

As such, when the forward and backward inaudible region information for each frequency band is calculated through the first to thirty-second inaudible region calculators 121 to 123, the information is transmitted to the corresponding frequency band verification units 131, 132, and 133. do. That is, the first non-audible region calculating unit 121 transmits the non-audible region information calculated by the first verifying unit 131, and the second non-audible region calculating unit 122 transmits the information to the second verifying unit 132. The calculated non-audible region information is transmitted, and the third non-audible region calculator 123 transmits the calculated non-audible region information to the third verification unit 133.

When the first to third verification units 131, 132, and 133 transmit audio signals from the corresponding buffer units 111 to 113, the forward and backward ratios provided from the corresponding inaudible region calculating units 121 to 123 are determined. The validity of the transmitted audio signal is determined by referring to the green area information. The validity is determined to be valid when the sound pressure of the audio signal currently transmitted from the buffer units 111 to 113 is not included in the inaudible region provided by the corresponding inaudible region calculating units 121 to 123. If it is included, it determines that it is not valid and processes it so that it is not transmitted.

For example, as shown in FIG. 2, an inaudible region in which the sound pressure of the audio signal at the f (t) point output from the first buffer unit 111 is calculated by the audio signal at the f (tm) point in advance of m hours is provided. When not included in the inaudible region A2 calculated by the audio signal at the time f (t + m) after (A1) and m hours, the first verification unit 131 is the first buffer unit 111. The audio signal output from time f (t) is determined to be valid and is processed to be transmitted.

The audio signal for each frequency band output by the verification of the first to thirty-two verification units 131 to 133 is transmitted to the combining unit 140. At this time, not all audio signals are output from the first to thirtieth verification units 131 to 133.

The synthesizer 140 synthesizes and outputs an audio signal of each frequency band transmitted from the first verifier 131 to the thirty-second verifier 133. The output audio signal is transmitted to an audio encoding apparatus (not shown). The audio encoding apparatus, not shown, is an apparatus for encoding an audio signal applied in a conventionally proposed manner.

As described above, the present invention performs a temporal masking process for each frequency band before compressing an audio signal with a conventional coding scheme, thereby performing pre-compression processing so that only an audio signal corresponding to a pure audio region is compressed with the existing coding scheme. The coding efficiency of the audio signal can be improved without degrading the audio quality.

Claims (1)

In an audio system having a function of encoding to compress a digital audio signal, A band dividing filter for dividing and transmitting a predetermined number of frequency bands when a predetermined digital audio signal is applied; Buffer means for storing the audio signal output for each frequency band from the band split filter for each frequency band; Inaudible region calculating means for calculating anterior and backward inaudible regions for audio signals output from the buffer means for each frequency band; The validity of the audio signal output from the buffer means for each frequency band is verified by referring to the forward and backward inaudible region information of the corresponding frequency band provided from the inaudible region calculating means. Verification means for outputting only; And a synthesizer for synthesizing the audio signals of the respective frequency bands outputted from the verification means and outputting the encoded signals.