JP2009031377A - Audio data processor, bit width conversion method and bit width conversion device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To transmit more data in data transmission in a transfer width of a bit rate smaller than a quantization bit rate of data. <P>SOLUTION: When data with quantization bit rate of A bits is converted to transmission data with quantization bit rate of N bits to transmit the data in a transmission width of N bits (N<A), a block of bit rate B (N<B≤A) in the data is divided to a first block of bit rate B1 and a second block of bit rate B2, and the second block is compressed to obtain a compressed block of C bits (C=N-B1). The compressed block is connected with the first block which is not compressed to obtain the transfer data. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technique for transmitting data with a transmission width smaller than the number of quantized bits of data, and more particularly to a technique for transmitting encoded audio data when reproducing audio data.

  In the audio field, a PCM (Pulse-Code Modulation) system in which an analog audio signal is converted into digital data and recorded / reproduced is often used (see Patent Document 1). In this method, digital data is recorded / reproduced as a frame structure. FIG. 5 (FIG. 2 in Patent Document 1) shows an example of a frame configuration when the number of quantization bits is 16 bits. 1 is a frame synchronization signal pattern, 2 is PCM data, and 3 is an error detection and correction code.

  The PCM data 2 is composed of 8 samples, and the number of quantization bits of each sample is 16 bits, so the data is a total of 128 bits. Data having such a configuration is generated by an encoder and recorded on a recording medium such as a CD (compact disc). Then, the playback apparatus detects the frame synchronization signal pattern 1 from the data read from the recording medium, performs error detection and correction processing in units of frames, and obtains PCM data 2. The processing until obtaining the PCM data is also called decoding, and is performed by a decoder in the reproducing apparatus.

  Non-Patent Document 1 discloses an example of a playback device that plays back PCM audio data (FIG. 3-169: audio DSP block and its surroundings). FIG. 6 is a diagram in which reference numerals are added to the respective components of the reproducing apparatus disclosed in the non-patent document.

  As shown in FIG. 6, the playback device includes a CD-ROM decoder 10, a FIFO 20, an audio DSP block 30, and a DAC I / F 40. When audio data is recorded on a CD, it is converted and recorded in accordance with the CD recording method, so the CD-ROM Decoder 10 converts the data converted in accordance with the CD recording method into audio data having the frame structure shown in FIG. To do. The FIFO 20 is a FIFO (First In First Out) type buffer, and temporarily stores audio data obtained by the CD-ROM Decoder 10. The audio DSP block 30 reads audio data from the FIFO 20 and decodes it to obtain PCM data. The PCM data is output to a subsequent DAC (D / A converter) via the DAC I / F 40, converted into an analog audio signal by the DAC, and output as audio.

  The audio DSP block 30 has a DSP (Digital Signal Processor) 32 that is a decoder and an output buffer 34 that temporarily stores PCM data obtained by the DSP 32.

  In recent years, DSP functions have been improved, and the DSP 32 can process data with a large number of quantization bits (for example, 32-bit data). By the way, the transmission format of PCM data is normally determined to be 24 bits or less (24 bits in the illustrated example), and the DSP 32 uses this transmission format when outputting the PCM data obtained by decoding to the output buffer 34. In order to comply, when the obtained PCM data is more than 24 bits, such as 32 bits, the lower bits are cut off and only the upper 24 bits are stored in the output buffer 34, and the subsequent stage with a transfer width of 24 bits later. Output to the process.

  Not only in the audio field, but also in other fields, when digital data is transmitted, it may be transmitted after reducing the number of quantization bits of the digital data due to restrictions on the format and transmission circuit related to transmission. May occur.

Japanese Patent Laid-Open No. 7-240071 User's Manual µPD63761, 63761A, PD63762,63762A, PD63763, 63763A, 1-chip LSI for compact disc with built-in compressed audio decoder (Document number S17187JJ1VOUD00 (first edition)

  As in the above example, due to limitations on the transmission format, transmission circuit, etc., if the lower bits of the number of quantization bits of the transmitted data are cut off, the amount of data decreases, and the transmitted data is processed. There is a problem that the quality of the system to be reduced may be lowered.

  For example, in an actual audio playback device, when playing back audio data recorded in the PCM system, a sound that usually adjusts the frequency characteristics of the sound before the PCM data obtained by decoding the audio data is output to the DAC. Field processing is performed. This sound field process is a filter process using, for example, an equalizer. FIG. 7 is a schematic diagram showing the relationship between the encoding processing LSI and the sound field processing LSI in the audio playback device.

  As shown in FIG. 7, the decode processing LSI 50 has a decode DSP 54 and an output buffer 58, and the sound field processing LSI 60 has a receiving unit 64 and a sound field processing DSP 68.

  The decode processing LSI 50 corresponds to the audio DSP block 30 in the audio playback apparatus shown in FIG. 6, and the sound field processing LSI 60 is located between the audio DSP block 30 and the DAC I / F 40 in FIG.

  The decode DSP 54 decodes the PCM audio data for each frame to obtain PCM data, and temporarily stores the PCM data in the output buffer 58. The output buffer 58 has a width corresponding to the data transmission format between the decode processing LSI 50 and the sound field processing LSI 60. Here, it is assumed that the data transmission format between 150 and the sound field processing LSI 60 is set to a transmission width of 24 bits, and the output buffer 58 has a corresponding 24-bit width.

  FIG. 8 is a flowchart showing the decoding process for one frame by the decoding DSP 54. The decoding DSP 54 first performs bit stream decoding processing on this frame to obtain a plurality of samples included in the audio data (S10). And the process of step S24-S34 is performed with respect to each sample. Specifically, first, sub-band synthesis processing and filter bank processing are performed to obtain one sample of PCM data (S24). Note that the subband synthesizing process and the filter bank process are specifically filter processes that are inversely quantized from the compression information divided into frequency bands and synthesized into PCM data by the synthesis F filter bank.

  In recent years, DSPs of 32 bits or more have become widespread due to improved DSP performance. Here, as an example, the decoding DSP 54 is a 32-bit DSP, and the number of quantization bits of the PCM data obtained thereby is 32 bits.

  In order to transmit this 32-bit PCM to the sound field processing LSI 60, the decode DSP 54 cuts off the lower 8 bits of the PCM data in accordance with the transmission format to produce quantized 24-bit transmission data (S28).

  The 24-bit transmission data is temporarily stored in the 24-bit output buffer 58 and later output to the sound field processing LSI 60 (S34).

  The processing from step S24 to S34 is performed for each sample in the frame (S44: No, S48, S24 to S34). When the process for the last sample is completed (S44: Yes), the decoding process for the frame ends.

  FIG. 9 is a flowchart showing processing by the sound field processing LSI 60. The sound field processing LSI 60 is an equalizer that performs filter processing for correcting the frequency characteristics for the data from the decoding processing LSI 50. As shown in FIG. 9, the sound field processing LSI 60 is decoded by the receiving unit 64. The 24-bit PCM data transmitted from is received (S50), and the received 24-bit PCM data is filtered by the sound field processing DSP 68 (S60).

  The data filtered by the sound field processing DSP 68 is converted into an analog signal by a DAC or the like and output to a speaker.

  As described above, the decoding DSP 54 in the decoding processing LSI 50 outputs the 32-bit PCM data by cutting it down to 24 bits due to the limitation of the transmission format, so that the amount of data that can be used by the sound field processing LSI 60 is cut off. Less than before.

  In the background of the progress of DSP performance improvement, the sound field processing DSP 68 is often used with a 32-bit or higher one for speeding up the processing. That is, the sound field processing DSP 68 outputs 24-bit PCM data from the decode processing LSI 50 even though it can handle data with a quantization bit number greater than 24 bits. The range is suppressed to about 144 dB, and the effect of improving the sound quality by the processing of the sound field processing LSI 60 is reduced.

  A first aspect of the present invention is an audio data processing device for converting audio data having a quantization bit number of A bits into transmission data having a quantization bit number of N bits (N <A). A width converter is provided. The bit width conversion unit includes a dividing unit, a compression unit, and a transmission data combining unit.

  The dividing unit divides a block having the bit number B (N <B ≦ A) in the audio data into a first block having the bit number B1 and a second block having the bit number B2. The compression unit compresses the second block to generate a compressed block having the number of bits C (C = N−B1). The transmission data combining unit combines the first block and the compressed block to obtain transmission data.

  A second aspect of the present invention is a bit width conversion device. This bit width conversion device converts data having a quantization bit number of A bits into transmission data having a quantization bit number of N bits in order to transmit the data with a transmission width of N bits (N <A). A dividing unit that divides a block of bit number B (N <B ≦ A) in the data into a first block of bit number B1 and a second block of bit number B2, and a second block Are compressed to obtain a compressed block of C bits (C = N−B1), and a combining unit that combines the compressed block and the first block to obtain transmission data.

  It should be noted that any combination of the above-described constituent elements and those expressing the present invention as a method, system or program are also effective as an aspect of the present invention.

  According to the data transmission technique of the present invention, more data can be transmitted when data is transmitted with a transmission width having a number of bits smaller than the number of quantized bits of data.

  Before describing specific embodiments of the present invention, first, the principle of converting the bit width of data according to the present invention will be described.

  FIG. 1 is a schematic diagram for explaining the principle of the present invention. In FIG. 1, A, B, B1, B2, C, and N indicate numbers, and these numbers satisfy the relationship represented by the following formula (1).

N <B = (B1 + B2) ≦ A
0 <B1 <N (1)
B2> C = (N−B1)

  As shown in the figure, the technique of the present invention is for converting data having a quantization bit number of A bits into transmission data having a quantization bit number of N bits in order to transmit the data with an N-bit transmission width. Is. First, the upper B bits of the data are divided into a first block of B1 bits and a second block of B2 bits. The B2 bit of the second block is compressed to obtain a C-bit compressed block. Then, the uncompressed first block and the compressed block are combined to obtain N-bit transmission data.

  As shown in equation (1) and FIG. 1, (B1 + B2) is larger than N bits of the transmission width, so it is more than the method in which the lower N bits of the data are cut off and only the upper N bits are transmitted. The amount of data can be transmitted.

  Further, since only a part of the bits is compressed, the quality of the transmission data can be maintained.

  Further, the value of B is matched with the maximum value M of the number of quantization bits that can be processed by the lower-order processing unit at the transmission destination of the transmission data (B = M), and the lower-level processing unit decompresses the compressed block to obtain the first If combined with the block, the lower processing unit can maximize its processing capability.

  In the description here, the upper B1 bits of the B bits are divided and compressed as the first block and the lower B2 bits are divided into the second blocks. May be a block composed of the higher order bits, or may be a block composed of the lower order bits of the B bits. That is, the lower B1 bits of the B bits may be divided and compressed as the first block and the upper B2 block as the second block.

  Hereinafter, embodiments embodying the above principle of the present invention will be described.

  FIG. 2 shows an audio playback apparatus 100 according to an embodiment of the present invention. The audio reproduction device 100 includes a decoding processing LSI 110 that performs decoding processing on compressed audio data such as MP3, a sound field processing LSI 160 that performs sound field processing on data obtained by processing lower than decoding, for example, the decoding processing LSI 110, A transmission circuit 150 that transmits data obtained by the decoding processing LSI 110 to the sound field processing LSI 160 is provided. Here, in order to make the gist of the present invention easier to understand, a device for reading data from a recording disk such as a CD, a device for decoding data read from the recording disk into PCM audio data, and a sound field processing LSI 150 are used. Description and illustration of a device having the same function or configuration as that of a normal audio reproduction device, such as a device for outputting analog audio data, is omitted.

  The decode processing LSI 110 includes a decode DSP 120, a bit width conversion unit 130, and an output buffer 140.

  The decoding DSP 120 is a processor of 32 bits or more (here, 32 bits as an example), and decodes compressed audio data such as MP3 for each frame to obtain PCM data having a quantization bit number of 32 bits.

  Data for transmission is obtained by converting the number of quantization bits of the PCM data obtained by the bit width conversion unit 130 and the decoding DSP 120 from 32 bits to 24 bits. As shown in FIG. 2, the bit width conversion unit 130 includes a division unit 132, a compression unit 134, and a combining unit 136.

  The output buffer 140 has a 24-bit width, temporarily stores 24-bit transmission data obtained by the bit width conversion unit 130, and this transmission data is later transmitted to the sound field processing LSI 160 by the transmission circuit 150. Is transmitted.

  FIG. 3 is a flowchart showing a flow of processing in which the decode processing LSI 110 decodes one frame. With reference to this, each component of the decode processing LSI 110, in particular, the operation of the bit width conversion unit 130 will be described in detail.

  The decoding DSP 120 in the decoding processing LSI 110 performs a bit stream decoding process on the target frame to obtain a plurality of samples included in the audio data (S110). And the process of step S114-S138 is performed with respect to each sample. Specifically, first, sub-band synthesis processing and filter processing are performed to obtain one sample of PCM data (S114). The number of quantization bits of this PCM data is 32 bits.

  This PCM data quantization bit number 32 corresponds to the numerical value “A” used to explain the principle of the present invention. In addition to “A”, in the following description, each numerical value used as an example is “B: 32, B1: 16, B2: 16, C: 8, N: 24, M: 32”. It has become.

  The dividing unit 132 of the bit width converting unit 130 divides the upper 32 bits of the 32-bit PCM data obtained by the decoding DSP 120 into two blocks. Here, as an example, since the decode DSP 120 generates 32-bit PCM data, the upper 32 bits thereof are all bits.

  The division by the dividing unit 132 is shown in steps S118 and S124 in FIG. First, the dividing unit 132 cuts out the upper 16 bits of the PCM data obtained by the decoding DSP 120 to obtain the first block (S118), and then cuts out the 16 bits following the upper 16 bits to obtain the second block. (S124).

  The compression unit 134 compresses 16 bits of the second block of the two blocks obtained by the division unit 132 to 8 bits to obtain a compressed block (S134).

  The combining unit 136 combines the first block of the two blocks obtained by the dividing unit 132 and the compressed block obtained by the compression unit 134 to obtain 24-bit transmission data (S136). It is stored in the output buffer 140 (S138).

  The processing from step S114 to S138 is performed for each sample in the frame (S144: No, S148, S114 to S138). When the process for the last sample is completed (S144: Yes), the decoding process for the frame ends.

  The transmission data stored in the output buffer 140 is later transmitted to the sound field processing LSI 160 by the transmission circuit 150. The transmission circuit 150 has a transmission width of 24 bits.

  Returning to FIG. 2, the sound field processing LSI 160 will be described.

  The sound field processing LSI 160 is an equalizer that performs filter processing for correcting the frequency characteristics of the data from the decoding processing LSI 110, and includes a receiving unit 170, a bit width restoring unit 180, and a sound field processing DSP 190.

  The receiving unit 170 receives data transmitted by the transmission circuit 150. The number of quantization bits of this received data is 24 bits.

  The sound field processing DSP 190 is a 32-bit processor, and performs a filtering process on the PCM data from the bit width restoration unit 180.

  The bit width restoration unit 180 restores the 24-bit received data received by the reception unit 170 to 32 bits and supplies the data to the sound field processing DSP 190. The bit width restoration unit 180 includes a division unit 182, an expansion unit 184, and a combination unit 186. .

  FIG. 4 is a flowchart showing the flow of processing by the sound field processing LSI 160. With reference to this, each component of the sound field processing LSI 160, in particular, the operation of the bit width restoration unit 180 will be described in detail.

  The receiving unit 170 receives the data transmitted from the transmission circuit 150 and outputs 24-bit received data to the bit width restoring unit 180 (S150).

  The dividing unit 182 of the bit width restoring unit 180 divides the 24-bit received data to obtain the first block and the compressed block. The division by the dividing unit 182 is shown in steps S154 and S158 shown in FIG. The dividing unit 182 cuts out the upper 16 bits of the received data to obtain a first block (S154), and then cuts out 8 bits following the upper 16 bits to obtain a compressed block (S158).

  The decompressing unit 184 decompresses the compressed block of the two blocks obtained by the dividing unit 182 to 16 bits to obtain a decompressed block (S164).

  The combining unit 186 combines the first block of the two blocks obtained by the dividing unit 182 and the decompressed block obtained by the decompressing unit 184 to obtain 32-bit PCM data to obtain the sound field processing DSP 190. (S168).

  The sound field processing DSP 190 executes filter processing on the 32-bit PCM data output from the bit width restoration unit 180 (S174).

  The data filtered by the sound field processing DSP 190 is converted into an analog signal by a DAC or the like and output to a speaker.

  As described above, according to the audio reproduction device 100 of the present embodiment, when the PCM data generated by the decoding DSP 120 is transmitted to the sound field processing LSI 160 with the transmission width of 24 bits, the upper 32 bits of the PCM data are set to 2 The block is divided into two blocks, and one block is compressed to reduce the number of quantization bits, and then combined with the other block to obtain 24-bit transmission data for transmission. By doing so, it is possible to transmit lower bits cut off to fit the transmission width of 24 bits.

  In addition, since the sum of the number of bits of the two blocks is the same as the maximum number of bits that can be processed by the sound field processing DSP 190 in the sound field processing LSI 160 that performs the low-order processing, the received 24 bits are received by the sound field processing LSI 160. The PCM data having the same number of quantized bits as the maximum number of bits that can be processed by the sound field processing DSP 190 is provided by decompressing the compressed portion of the data of the data and combining it with the uncompressed portion. be able to. As a result, the processing capability of the sound field processing DSP 190 can be maximized, and the effect of improving the sound quality by the processing of the sound field processing LSI 160 can be increased.

  The present invention has been described above based on the embodiment. The embodiment is an exemplification, and various changes and increases / decreases may be added without departing from the gist of the present invention. It will be understood by those skilled in the art that modifications to which these changes and increases / decreases are also within the scope of the present invention.

  For example, in the audio playback device 100 of the present embodiment, since the upper bits of the PCM data are normally assigned to a larger amplitude, the bit width conversion unit 130 of the decode processing LSI 110 is the upper of the upper 32 bits of the PCM data. The 16 bits are divided into the first block and the 16 bits following the 16 bits are divided as the second block, and the second block consisting of the lower 16 bits is compressed. On the other hand, PCM data is expressed by a linear value of ±, and in a voice waveform, since the change of upper bit data is usually small, when compressing upper bits, load such as DPCM (Differential PCM) method is compared. There is an advantage that a light compression method can be used. From the viewpoint of reducing the compression load, the lower 16 bits are divided into the first block, the 16 bits (that is, the upper 16 bits) following the lower 16 bits are divided into the second block, and a second block consisting of the upper 16 bits is divided. You may make it compress a block.

It is a figure for demonstrating the principle of this invention. It is a figure which shows the audio reproduction apparatus concerning embodiment of this invention. 3 is a flowchart showing a flow of processing of a decoding processing LSI in the audio reproduction device shown in FIG. 3 is a flowchart showing a processing flow of a sound field processing LSI in the audio reproduction device shown in FIG. 2. It is a figure which shows the example of the audio data of a PCM system. It is a figure which shows the example of the audio reproduction apparatus of a prior art. It is a figure which shows the decoding processing LSI and sound field processing LSI in the conventional audio reproduction apparatus. It is a flowchart which shows the flow of a process of the decoding process LSI shown in FIG. It is a flowchart which shows the flow of a process of the sound field processing LSI shown in FIG.

Explanation of symbols

1 Frame sync signal pattern 2 PCM data 3 Error detection and correction code 10 CD-ROM Decoder
20 FIFO 30 Audio DSP block 32 DSP 34 Output buffer 40 DAC, I / F 50 Decode processing LSI
54 decoding DSP 58 output buffer 60 sound field processing LSI 64 receiving unit 68 sound field processing DSP 100 audio playback device 110 decoding processing LSI 120 decoding DSP
130 bit width conversion unit 132 division unit 134 compression unit 136 combination unit 140 output buffer 150 transmission circuit 160 sound field processing LSI 170 reception unit 180 bit width restoration unit 182 division unit 184 expansion unit 186 combination unit 190 sound field processing DSP

Claims (6)

A bit width conversion unit that converts audio data having a quantization bit number of A bits into transmission data having a quantization bit number of N bits (N <A);
The bit width converter is
A dividing unit that divides a block having a bit number B (N <B ≦ A) in the audio data into a first block having a bit number B1 and a second block having a bit number B2.
A compression unit that compresses the second block to generate a compressed block having the number of bits C (C = N−B1);
An audio data processing apparatus comprising: a transmission data combining unit that combines the first block and the compressed block.   2. The audio data processing device according to claim 1, wherein the first block corresponds to the number of B1 bits from the high order of the audio data when the high order bits of the audio data are assigned to a larger amplitude. . A decoder for generating the A-bit audio data with compressed audio data as an input;
A data restoration unit for restoring the block of B bits from the N-bit transmission data;
The audio data processing apparatus according to claim 1, further comprising: a sound field processing unit that performs sound field processing on the block of B bits restored by the restoration unit. The data decompression unit decompresses the compressed block in the N-bit transmission data to obtain a decompressed block of bit number B2, and
The audio data processing apparatus according to claim 3 , further comprising a combining unit that combines the expansion block obtained by the expansion unit and the first block in the transmission data. When the data is converted into data for transmission having a quantization bit number of N bits in order to transmit data having a quantization bit number of A bits with a transmission width of N bits (N <A),
A block having a bit number B (N <B ≦ A) in the data is divided into a first block having a bit number B1 and a second block having a bit number B2.
Compressing the second block to obtain a compressed block of C bits (C = N−B1);
A bit width conversion method comprising: combining the compressed block and the first block to obtain the transmission data. A bit width conversion device that converts data having a quantization bit number of A bits into transmission data having a quantization bit number of N bits in order to transmit the data with a transmission width of N bits (N <A). And
A dividing unit that divides a block having a bit number B (N <B ≦ A) in the data into a first block having a bit number B1 and a second block having a bit number B2.
A compression unit that compresses the second block to obtain a compressed block of C bits (C = N−B1);
A bit width conversion apparatus comprising: a transmission data combining unit that combines the compressed block and the first block to obtain the transmission data.

Images