JP2009094836A - Digital speech processing apparatus and digital speech processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make it possible to reproduce a full band of music software even if at least either of an audio amplifier and a loudspeaker cannot reproduce a full band of the music software. <P>SOLUTION: An fs conversion circuit 2 upsamples a digital speech signal. An unreproducible band signal addition circuit 3 generates a signal in an unreproducible band higher than the reproducible band of the audio amplifier and/or the speaker from the digital speech signal upsampled by means of the fs conversion circuit and adds the generated signal to the reprpducible band. Then, the fs conversion circuit performs downsampling of the sampling frequency of an output signal of the unreproducile band signal addition circuit to a predetermined sampling frequency and outputs the downsampled output signal of the unreproducible band signal addition circuit to the audio amplifier and the loudspeaker. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention reproduces audio software such as DVD audio, music CD, and television audio signal sampled at a predetermined sampling frequency with an audio amplifier and / or speaker whose reproducible band is lower than 1/2 of the sampling frequency. The present invention relates to a digital audio processing apparatus and a digital audio processing program.

Sampling frequencies fs of DVD audio are 96 kHz and 192 kHz, and fs = 44.1 kHz for a music CD. Therefore, if the band of the audio amplifier and the speaker for reproducing these is up to fs / 2, the entire band of the audio software can be reproduced. Further, since the band of television broadcasting is up to 16 kHz in the case of digital broadcasting (AAC system) and up to 20 kHz in the case of analog broadcasting, a television having an audio amplifier and a speaker that can reproduce this entire band. If it is a receiver, the entire band can be reproduced. As a conventional technique, Patent Document 1 below discloses a technique for converting audio information in a narrow frequency band into audio information in a wide frequency band.
JP 11-1226097 A (Abstract)

  However, if any one of the audio amplifier and the speaker cannot reproduce the entire band of the music software, there is a problem that the entire band of the music software cannot be reproduced as a result. For example, when the reproduction frequency band of the narrower one of the audio amplifier and the speaker is up to 20 kHz, a band of 20 kHz or more (40 kHz or more in terms of sampling frequency fs) of DVD audio cannot be reproduced, and the reproduction frequency band of the CD player is In the case of up to 18 kHz, it is impossible to reproduce a band of 18 kHz or more of CD (36 kHz or more in terms of sampling frequency fs). Furthermore, when the reproduction frequency band of the audio amplifier and the speaker of the television receiver is up to 12 kHz, it is not possible to reproduce a band of 12 kHz or more (24 kHz or more in terms of sampling frequency fs) of the television audio signal.

  In view of the above-described problems of the prior art, the present invention provides a digital audio that can reproduce the entire band of music software even when any one of the audio amplifier and the speaker cannot reproduce the entire band of the music software. It is an object to provide a processing device and a digital audio processing program.

In order to achieve the above object, the present invention provides a digital audio for reproducing a digital audio signal sampled at a predetermined sampling frequency with an audio amplifier and / or speaker whose reproducible band is lower than 1/2 of the sampling frequency. A processing device comprising:
Upsampling means for upsampling the sampling frequency of the digital audio signal;
Non-reproducible band signal adding means for generating a non-reproducible band signal higher than the reproducible band of the audio amplifier and / or speaker from the digital audio signal up-sampled by the up-sampling means, and adding the non-reproducible band signal to the reproducible band; ,
Downsampling means for downsampling the sampling frequency of the output signal of the non-reproducible band signal adding means to the predetermined sampling frequency and outputting it to the audio amplifier and speaker;
Prepared.

In order to achieve the above object, the present invention is for reproducing a digital audio signal sampled at a predetermined sampling frequency with an audio amplifier and / or speaker whose reproducible band is lower than 1/2 of the sampling frequency. A digital audio processing program,
An upsampling step of upsampling a sampling frequency of the digital audio signal;
A non-reproducible band signal adding step of generating a non-reproducible band signal higher than the reproducible band of the audio amplifier and / or the speaker from the digital audio signal up-sampled by the up-sampling step, and adding the non-reproducible band signal to the reproducible band; ,
A downsampling step of downsampling the sampling frequency of the output signal of the non-reproducible band signal adding means to the predetermined sampling frequency and outputting it to the audio amplifier and speaker;
Prepared.

  According to the present invention, a non-reproducible band signal higher than the reproducible band of the audio amplifier and / or speaker is generated from the upsampled digital audio signal, added to the reproducible band, and down-sampled to a predetermined sampling frequency. Therefore, even if any one of the audio amplifier and the speaker cannot reproduce the entire band of the music software, the entire band of the music software can be reproduced.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a digital audio processing apparatus and a digital audio processing program according to the present invention.

  In FIG. 1, a digital signal input 1 is a signal to be processed. For example, the above-described DVD audio signal (sampling frequency fs = 96 kHz, 192 kHz), music CD signal (fs = 44.1 kHz), audio of digital television broadcasting. Signal (band in case of AAC system = 16 kHz or less). Further, the signal is a digitized signal when an audio signal of analog television broadcasting is processed. The fs conversion circuit 2 includes an oversampling filter (OS filter) and upsamples the digital signal input 1. The non-reproducible band signal adding circuit 3 is reproducible because the reproducible band (several tens of Hz to f1) is lower than half the sampling frequency fs (f1 <fs / 2) and is reproduced by an audio amplifier and / or speaker. A non-reproducible band signal higher than the band is generated and added to the reproducible band.

  The subsequent fs conversion circuit 4 includes a decimation filter (D. filter), down-samples the sampling frequency of the output of the non-reproducible band signal adding circuit 3 to the sampling frequency fs of the digital signal input 1, and outputs the digital signal output 5 (not shown). Output to audio amplifier or speaker. When an audio amplifier (not shown) can process only an analog input, the digital signal output 5 of the fs conversion circuit 4 is converted into an analog signal and output to the audio amplifier.

  An example of the digital signal input 1 will be described. CD is data determined in a format of 16 bit / 44.1 kHz, but actual music data has a margin of amplitude in the format as shown in FIG. Further, the 16-bit / 44.1 kHz compressed sound source is data determined in the same format as the CD, but the actual compressed sound source data also has margins in the amplitude and frequency band in the format as shown in FIG. Therefore, a non-reproducible band signal of f1 or more is added to the marginal band, particularly the reproducible band (several tens of Hz to f1) of the audio amplifier and / or the speaker.

  As an example of upsampling of the fs conversion circuit 2, frequency characteristics when upsampling four times as shown in FIG. 4A and oversampling as shown in FIG. 4B are shown. Even if upsampling is performed, the signal component remains the same as the original input signal. Further, since the sampling frequency is increased, the aliasing noise N is shifted to a higher frequency band.

  FIG. 5 shows a non-reproducible band signal added to a signal upsampled four times as shown in FIG. 4, and the non-reproducible band signal is a marginal band of the digital signal input 1 band (<fs / 2). And (1/2) fs to (3/2) fs. Note that the aliasing component (noise N) remains in the band of (3/2) fs to 3fs.

  FIG. 6 shows a 1/4 down-sampling output (digital signal output 5), and the band (<fs / 2) of the down-sampling output digital signal input 1 is the original digital signal input 1 and the unreproducible band signal. Yes, the folded component (noise N) remains in the band of (1/2) fs to fs. FIG. 7 shows the digital signal output 5, and when it is output to an audio amplifier and / or a speaker whose reproducible band (several tens of Hz to f1) is lower than 1/2 of the sampling frequency fs (f1 <fs / 2), it is less than fs1. Since the non-reproducible band signal is added to this band, the audio amplifier or the speaker can reproduce the non-reproducible band signal. Therefore, high sound quality not obtained with the original signal can be obtained.

  FIG. 8 is a block diagram showing in detail the configuration of the non-reproducible band signal adding circuit 3. In FIG. 8, the I / O port 10 takes in the audio data upsampled by the fs conversion circuit 2 shown in FIG. 1 and outputs it to the harmonic generation circuit 11 and the delay circuit 12. The harmonic generation circuit 11 forms harmonic data that is a non-reproducible band signal based on the audio data, and the delay circuit 12 performs harmonic generation processing in the harmonic generation circuit 11 on the audio data. Give a delay for the time required. The adder 13 adds the harmonic data from the harmonic generation circuit 11 to the audio data from the delay circuit 12, and the I / O port 14 outputs the output of the adder 13 to the fs conversion circuit 4 shown in FIG. To do.

  FIG. 9 is a block diagram showing in detail the harmonic generation circuit 11 of FIG. The audio data that has been upsampled and taken into the input terminal 21 is supplied to the delay circuit 22, the comparison circuit 23, and the difference detection circuit 28. The delay circuit 22 delays the audio data by one sample, and the comparison circuit 23 delays the one sample by the delay circuit 22 and the audio data of the current sample supplied via the input terminal 21. Compare the levels of the recorded audio data.

  Here, as shown in FIG. 10, the comparison circuit 23 compares the supplied audio data with the audio data of the previous sample for each sample, and the current audio data has a larger sample value than the audio data of the previous sample. Supplies “0” to the peak-to-peak comparison output forming circuit 24 as a comparison result output. When the current audio data and the previous sample audio data have the same sample value, the comparison circuit 23 compares the current audio data with the previous sample audio data. Further, the past sample values are compared retrospectively, for example, compared with the audio data of the sample before and after. In addition, when audio data having the same sample value continues for 9 samples or more, this indicates that it is blank. For this reason, the comparison circuit 23 continues this comparison, and when there is a change, if the sample audio data at the time of change is larger than the current audio data, “0” is indicated; 1 ”is output as the comparison result output.

  The peak-to-peak comparison output forming circuit 24 detects each top peak and under peak of the waveform of the audio data based on the comparison result from the comparison circuit 23, and from the comparison result and the under peak between the top peak and the under peak. The comparison result up to the top peak is output to the pattern detection circuit 25. Here, for example, as shown in FIG. 10A, the voice corresponding to the comparison result “0”, which is the previous sample of the comparison result “1” when the comparison result “0” changes to the comparison result “1”. Data show “top peak”. Similarly, the audio data corresponding to the comparison result “1” that is the previous sample of the comparison result “0” at the time when the comparison result “1” changes to the comparison result “0” indicates “under peak”.

  Therefore, the peak-to-peak comparison output forming circuit 24 detects the top peaks A1, A2, A3... As shown in FIG. 10 (a) based on the change points of the comparison results “0”, “1”. And under peaks B1, B2, B3... Are detected. Then, the comparison result between the continuous top peak and the under peak is supplied to the pattern detection circuit 25 as a peak-to-peak comparison result. That is, since the comparison result between the continuous top peak and the under peak, for example, the top peak A1 and the under peak B1, is “1, 1”, the comparison result “1, 1” is used as the comparison result between peaks. This is supplied to the pattern detection circuit 25. Further, since the comparison result between the under peak B1 and the top peak A2 is “0, 0, 0, 0, 0”, this comparison result “0, 0, 0, 0, 0” is output for comparison between peaks. To the pattern detection circuit 25.

  The pattern detection circuit 25 stores in advance a continuous pattern of the same peak-to-peak comparison result appearing between successive top peaks and under-peaks. When the peak-to-peak comparison result is supplied to the pattern detection circuit 25, the pattern detection circuit 25 Is compared with the comparison result between the peaks, and it is determined which pattern is stored in advance.

  Specifically, the pattern detection circuit 25 includes a “1Ts pattern” indicating that the same peak-to-peak comparison output between the top peak and the under peak is only one sample of “1” or “0”. "2Ts pattern" indicating that the continuation is 2 samples of "1, 1" or "0, 0", and the continuation is 3 samples of "1, 1, 1" or "0, 0, 0" “3Ts pattern” indicating that “4Ts pattern” indicating that there are four samples of “1, 1, 1, 1” or “0, 0, 0, 0”, and “1, 1, 1, “5, 1, 1” or “0, 0, 0, 0, 0” corresponding to “5Ts pattern”, “1”, “1”, “0”, “0” , “0, 0, 0, 0, 0” indicating “6 samples” "s pattern", "7Ts pattern" indicating that the continuation is 7 samples of "1, 1, 1, 1, 1, 1, 1" or "0, 0, 0, 0, 0, 0, 0" , “8Ts pattern indicating that the sequence is“ 1, 1, 1, 1, 1, 1, 1, 1 ”or“ 0, 0, 0, 0, 0, 0, 0, 0 ” ”And“ special patterns ”indicating the continuation of“ 0 ”or“ 1 ”for nine samples or more are stored in the memories 25a to 25i, respectively. That is, the reciprocals 1 / 1Ts and 1 / 2Ts of these time lengths 1Ts and 2Ts represent the audio frequency. The pattern detection circuit 25 compares each pattern stored in each of the memories 25 a to 25 i with the peak-to-peak comparison result, and supplies a pattern detection signal to the selector 26.

  The selector 26 selects the pattern detection signal supplied from the pattern detection circuit 25 and supplies it to the shift amount control table 27, the difference detection circuit 28 and the addition / subtraction timing control circuit 30. In the shift amount control table 27, the difference detection circuit 28, and the addition / subtraction timing control circuit 30, when the pattern detection signal is supplied from the selector 26, the addition / subtraction amount for forming a harmonic is detected. Then, the adder 13 adds / subtracts a level corresponding to the pattern detection signal from the selector 26 to a predetermined sample value of the audio data (original audio data) from the delay circuit 12, thereby obtaining the original audio data. Extend the frequency band.

  Specifically, when the pattern between the top peak and the under peak is 2Ts pattern or more and 5Ts pattern or less, the sample values of the top peaks A1, A2, A3... Shown in FIG. The sample values C1, C2, C4, C5, C8, C9, C10, C11... Are “added” with data at a level corresponding to the difference between the top peak and the sample values before and after, The sample values of the underpeaks B1, B2, B3, etc. “before and after sample values C2, C3, C6, C7, C9, C10...” Correspond to the difference between the top peak and the sample values before and after. “Subtract” the level data. That is, a predetermined value is added to the sample values before and after the top peak, and a predetermined value is subtracted from the sample values before and after the under peak.

  When the pattern between the top peak and the under peak is 6Ts pattern or more and 8Ts pattern or less, the sample values of the top peaks A5, A6,... Shown in FIG. “C22, C29, C30...” Is “added” with data at a level corresponding to the difference between the top peak and the sample values before and after the top peak, and the sample values “5” of the top peaks A5, A6. For the sample values C23, C24, C31, C32... Before and after, “addition processing” is performed on data at a level corresponding to the difference between the sample values before and after the top peak and the sample values before and after the top peak. That is, a predetermined value is added to the sample values before and after the top peak and the sample values before and after the top peak.

  Further, when the pattern between the top peak and the under peak is 6Ts pattern or more and 8Ts pattern or less, the sample values of the under peaks B4, B5... Shown in FIG. C26... ”Is subtracted from the level data corresponding to the difference between the under peak and the sample values before and after that, and the samples of the under peaks B4, B5. From the values C27, C28..., The data of the level corresponding to the difference between the sample value before and after the under peak and the sample value before and after the under peak is “subtracted”. A predetermined value is subtracted from the sample values before and after the underpeak and the sample values before and after.

  Further, in the case of a 1Ts pattern such as the pattern between the top peak A7 and the under peak B6 and the pattern between the top peak A8 and the under peak B7 shown in FIG. 10C, the sample value is a sufficiently high harmonic. In order to show that there is, there is no addition / subtraction process as described above. Similarly, when the pattern between the top peak and the under peak is a 9Ts pattern or more, such as between the under peak B7 to the top peak A7 and after the top peak A9 shown in FIG. Indicates a blank or the like, and in this case, the above-described addition / subtraction processing is not performed.

  The shift amount control table 27 stores a plurality of addition / subtraction amounts for forming harmonics in advance, and outputs the addition / subtraction amounts corresponding to each pattern detection output from the selector 26 to the bit shifter 29. The difference detection circuit 28 detects a difference between sample values corresponding to each pattern detection signal from the selector 26 based on the audio data supplied via the input terminal 21, and the bit shifter 29 receives the difference from the difference detection circuit 28. The difference data is subjected to bit shift processing corresponding to the addition / subtraction amount from the shift amount control table 27 to form addition / subtraction data. The addition / subtraction timing control circuit 30 outputs the addition / subtraction data from the bit shifter 29 via the output terminal 31 at a timing corresponding to each pattern detection output from the selector 26.

  That is, in the case of the pattern “2Ts pattern or more and 5Ts pattern or less”, the difference detection circuit 28, as shown in FIG. 11A, the difference level A between the sample value of the top peak and the sample values before and after the top peak, Further, as shown in FIG. 11B, a difference level B between the sample value of the under peak and the sample value before and after the under peak is detected and supplied to the bit shifter 29. Further, in the case of the pattern “6Ts pattern or more and 8Ts pattern or less”, the difference detection circuit 28 calculates the sample value before and after the top peak and the sample value before and after the top peak as shown in FIG. The difference level C and the difference level D between the sample value before and after the under peak and the sample value before and after the under peak as shown in FIG. 11D are detected and supplied to the bit shifter 29. In the case of a 1Ts pattern or a special pattern of 9Ts or more, the difference detection circuit 28 does not detect the difference level because the addition / subtraction process is not performed as described with reference to FIG. 10C. .

  In the shift amount control table 27, the shift amounts of the difference levels A to D in the bit shifter 29 for setting the difference levels A to D supplied to the bit shifter 29 as the addition and subtraction amounts corresponding to the respective patterns of 2Ts to 8Ts. For example, the difference levels A to D are stored in advance in four stages. Specifically, in the shift amount control table 27, as shown in FIG. 12, the difference level A or the difference level B in the case of 2Ts pattern or 3Ts pattern is set to 1/2 to 1/16 level depending on the level. Each shift amount for forming the addition / subtraction amount is stored. Further, the shift amount control table 27 forms each of the addition / subtraction amounts with the difference level A or the difference level B in the case of the 4Ts pattern or the 5Ts pattern as a level of 1/4 to 1/32 according to the level. The shift amount is stored. In the shift amount control table 27, each shift amount for forming the addition / subtraction amount with the difference levels A to D in the case of the 6Ts pattern or 7Ts pattern as 1/8 to 1/64 level according to the level. Is remembered. Further, the shift amount control table 27 stores the shift amounts for forming the addition / subtraction amounts with the difference levels A to D in the case of the 8Ts pattern as 1/16 to 1/128 levels according to the levels. ing.

  The addition / subtraction amount corresponding to each shift amount is a value found based on ergonomics after repeated trial listening by the applicant of the present invention over several years. Therefore, as will be described later, it is possible to obtain a high-quality sound output by adding / subtracting the addition / subtraction amount to the original audio data.

  Based on the pattern detection signal supplied from the selector 26, the shift amount control table 27 detects the shift amount (FIG. 12) corresponding to this pattern. Then, the shift data indicating the shift amount is supplied to the bit shifter 29 to which the difference levels A to D are supplied. Based on the shift data from the shift amount control table 27, the bit shifter 29 performs a bit shift process on each difference level A to D from the difference detection circuit 28, thereby changing each difference level A to D between its peaks. The level is set to 1/2 to 1/128 according to the pattern of comparison output, and this is supplied to the addition / subtraction timing control circuit 30 as an addition / subtraction amount (addition / subtraction data).

  As described above, this addition / subtraction data is added to the sample value before and after the top peak, or the sample value before and after the top peak, and the sample value before and after the under peak or the sample value before and after the under peak. In the case of subtracting this addition / subtraction data, the bit shifter 29 sets the most significant bit (MSB) of the addition / subtraction data as a sign bit to “1 (addition)” or “0 (subtraction). ) ".

  Next, when the addition / subtraction data is formed in this way, the output timing of the formed addition / subtraction data is controlled. That is, as described above, the timing of adding / subtracting this addition / subtraction data to the original audio data is the addition / subtraction data to the sample values before and after the top peak or under peak when the peak-to-peak comparison output pattern is 2Ts pattern to 5Ts pattern In the case of the 6Ts pattern to the 8Ts pattern, the addition / subtraction process is performed on the sample values before and after the top peak or the under peak and the sample values before and after the top peak or the under peak. . For this reason, the addition / subtraction timing control circuit 30 measures the output timing of the addition / subtraction data from the bit shifter 29 in accordance with the pattern detection signal from the selector 26, and adds the addition data shown in FIG. Supply to the vessel 13.

  Thus, in the case of 2Ts pattern to 5Ts pattern, the addition / subtraction data is added to the adder 13 at the timing when the sample values before and after the top peak or under peak are supplied to the adder 13 as shown by the dotted arrows in FIG. 13 is supplied. In the case of the 6Ts pattern to 8Ts pattern, as shown by the dotted arrows in FIG. 10B, the sample values before and after the top peak or under peak and the sample values before and after the top peak or under peak are added. The addition / subtraction data is supplied to the adder 13 at the timing supplied to the adder 13.

  Next, when addition / subtraction data is supplied to the adder 13, the adder 13 adds / subtracts the addition / subtraction data (harmonic) formed by the harmonic generation circuit 11 to the original audio data supplied from the delay circuit 12. Process. Specific examples of waveform changes before and after this addition / subtraction process are shown in FIGS. 13 (a) to 13 (c) and FIGS. 14 (a) to 14 (c). FIGS. 13A to 13C show addition processing in the case of the 2Ts pattern. As can be seen from FIGS. 13A to 13C, when the peak-to-peak comparison output pattern is a 2Ts pattern, the difference level between the sample value of the top peak of the original audio data and the sample values before and after the top peak. The addition / subtraction data calculated according to the above is added to the sample values before and after the top peak as shown by the oblique lines in FIGS. As a result, the waveform of the original audio data indicated by the dotted line in FIGS. 13A to 13C is converted into a waveform in which the frequency band is expanded as indicated by the solid line in FIGS. 13A to 13C. Can be shaped.

  Similarly, FIGS. 14A to 14C show the addition / subtraction processing when the peak-to-peak comparison output pattern is a 3Ts pattern. As can be seen from FIGS. 14A to 14C, when the peak-to-peak comparison output pattern is a 3Ts pattern, the difference level between the sample value of the top peak of the original audio data and the sample values before and after the top peak. The addition / subtraction data calculated according to the above is added to the sample values before and after the top peak as shown by the right oblique lines in FIGS. Also, the addition / subtraction data calculated according to the difference level between the sample value before and after the under peak of the original audio data and the sample value before and after the under peak is the left diagonal line in FIGS. 14 (a) to 14 (c). As shown in Fig. 8, the sample value is added to the sample value before and after the top peak. As a result, the waveform of the original audio data indicated by the dotted line in FIGS. 14A to 14C is converted into a waveform whose frequency band has been expanded as indicated by the solid line in FIGS. 14A to 14C. Can be shaped.

  The audio data that has been subjected to the frequency shaping of the frequency band extension (addition of harmonics) in this way is supplied to the fs conversion circuit 4 shown in FIG. 1 via the I / O port 14 to obtain the original sampling frequency fs. Downsampled.

  In the non-reproducible band signal addition circuit 3, waveform shaping of the original audio data (formation of the harmonic component and synthesis of the harmonic component and the original audio data) can be performed using only addition / subtraction processing. Conventionally, waveform shaping can be performed without using a conversion table for nonlinear processing, a differentiation circuit, a cube circuit, or the like, which has been necessary for such waveform shaping, and the harmonic generation circuit 11 and the adder 13 can be used. The need to provide a high-pass filter (HPF) for extracting harmonic components from a non-linear signal can be eliminated. In particular, a filter system such as a high-pass filter, as can be seen from the configuration of so-called IIR filters and FIR filters, causes a complicated and large circuit scale when trying to obtain good filter characteristics. Although the use of a filter causes a distortion of the waveform, the non-reproducible band signal adding circuit 3 can eliminate this factor from the root. Accordingly, the circuit scale of the non-reproducible band signal addition circuit 3 can be reduced to reduce the chip size, and the cost can be reduced, the productivity can be improved, and the performance can be improved. Since small and high-performance products can be provided at low cost, it is possible to sufficiently cope with today's price destruction and downsizing.

It is a block diagram which shows embodiment of the digital audio processing apparatus and digital audio processing program which concern on this invention. It is explanatory drawing which shows the zone | band of CD. It is explanatory drawing which shows the zone | band of a compression sound source. It is explanatory drawing which shows the upsampling in FIG. It is explanatory drawing which shows the reproduction | regeneration impossible band signal addition in FIG. It is explanatory drawing which shows the downsampling in FIG. It is explanatory drawing which shows the digital signal output in FIG. FIG. 2 is a block diagram illustrating in detail a non-reproducible band signal adding circuit in FIG. 1. It is a block diagram which shows the harmonic generation circuit in FIG. 8 in detail. It is explanatory drawing which shows the pattern detection process in FIG. It is explanatory drawing which shows the addition / subtraction process in FIG. It is explanatory drawing which shows the addition / subtraction data in FIG. It is explanatory drawing which shows the addition / subtraction process in FIG. It is explanatory drawing which shows the addition / subtraction process in FIG.

Explanation of symbols

1 Digital signal input 2 fs conversion circuit (upsampling means)
3 Non-reproducible band signal addition circuit 4 fs conversion circuit (down sampling means)
5 Digital signal output 10, 14 I / O port 11 Harmonic generation circuit 12, 22 Delay circuit 13 Adder 21 Input terminal 23 Comparison circuit 24 Inter-peak comparison output formation circuit 25 Pattern detection circuit 26 Selector 27 Shift amount control table 28 Difference Detection circuit 29 Bit shifter 30 Addition / subtraction timing control circuit 31 Output terminal

Claims (2)

A digital audio processing apparatus for reproducing a digital audio signal sampled at a predetermined sampling frequency with an audio amplifier and / or a speaker whose reproducible band is lower than 1/2 of the sampling frequency,
Upsampling means for upsampling the sampling frequency of the digital audio signal;
Non-reproducible band signal adding means for generating a non-reproducible band signal higher than the reproducible band of the audio amplifier and / or speaker from the digital audio signal up-sampled by the up-sampling means, and adding the non-reproducible band signal to the reproducible band; ,
Downsampling means for downsampling the sampling frequency of the output signal of the non-reproducible band signal adding means to the predetermined sampling frequency and outputting it to the audio amplifier and speaker;
Digital audio processing device provided. A digital audio processing program for reproducing a digital audio signal sampled at a predetermined sampling frequency with an audio amplifier and / or a speaker whose reproducible band is lower than 1/2 of the sampling frequency,
An upsampling step of upsampling a sampling frequency of the digital audio signal;
A non-reproducible band signal adding step of generating a non-reproducible band signal higher than the reproducible band of the audio amplifier and / or the speaker from the digital audio signal up-sampled by the up-sampling step, and adding the non-reproducible band signal to the reproducible band; ,
A downsampling step of downsampling the sampling frequency of the output signal of the non-reproducible band signal adding means to the predetermined sampling frequency and outputting it to the audio amplifier and speaker;
Digital voice processing program provided.

Images