JP4704872B2 - Audio signal output device - Google Patents

Description

  The present invention relates to an audio signal output device that outputs an audio signal.

  Examples of the audio signal output device include a CD (Compact Disc) player and an AV (Audio Visual) amplifier. In this audio signal output device, a digital audio signal is passed through a digital filter and oversampled (interpolated) before DA (Digital to Analog) conversion, and then passed through a low-pass filter having a high cutoff frequency. As a result, the influence of the low-pass filter on the amplitude and phase characteristics of the analog audio signal after DA conversion is reduced, and the sound quality is improved.

  Generally, when analog (digital to analog) conversion is performed on an analog audio signal, the analog audio signal is passed through an aliasing filter in order to limit the frequency to a frequency equal to or lower than half the sampling frequency fs. For example, the sampling frequency fs of the digital audio signal recorded on the CD is 44.1 kHz. Therefore, when an analog audio signal is recorded on a CD, the analog audio signal is converted into a digital audio signal by AD conversion after a band of 22.05 kHz or higher is removed by an aliasing filter.

  In this way, although the actual audio signal contains frequency components that are higher than the audible band (20 kHz), high frequency components (bands that are 1/2 or more of the sampling frequency fs) are generated by the aliasing filter. It will be removed. For this reason, the reproduced sound reproduced from the CD lacks high-frequency components as compared with the conventional analog reproduced sound, and there are users who are dissatisfied with the reproduced sound of the CD.

  Therefore, there is a method of adding a high frequency component by performing oversampling (so-called zero interpolation processing) using a digital filter on a digital audio signal read from a CD or the like (see, for example, Patent Document 1). Further, there is a method of generating a harmonic component or dither signal from a digital audio signal by a non-linear processing circuit and adding the harmonic component or dither signal to the digital audio signal according to the high frequency spectrum intensity of the digital audio signal. (For example, refer to Patent Document 2).

  Further, when a high frequency component is added by zero interpolation processing using a digital filter, the obtained signal waveform may be insufficiently smooth. For this reason, in order to faithfully reproduce higher quality audio signals, oversampling may be performed using spline interpolation or Lagrange interpolation that smoothly connects sampling points instead of zero interpolation processing.

Japanese Patent Laid-Open No. 9-23127 JP 2002-366178 A

  However, since the sampling point changes in the vicinity of the high frequency (1/2 of the sampling frequency fs), the over-sampling method using spline interpolation or Lagrange interpolation faithfully reproduces the waveform near the high frequency. It becomes difficult to do. In other words, the waveform of the analog audio signal after oversampling and DA conversion is not reproduced as the same waveform as the analog audio signal near the high frequency (hereinafter referred to as the original signal) that was DA converted without oversampling. Therefore, distortion is likely to occur.

  This point will be described with reference to FIGS. FIG. 9 is a diagram showing a waveform of an analog audio signal in a high frequency (frequency near fs / 2). FIG. 10 is a diagram showing a waveform of an analog audio signal in the middle to low range (frequency << fs / 2). 9 and 10, the horizontal axis represents time, and the vertical axis represents amplitude. The solid line indicates the waveform of the original signal. A broken line shows a waveform of an analog audio signal obtained by DA-converting a digital audio signal oversampled by Lagrangian interpolation. A symbol “◯” indicates a sampling point when sampling is performed at the frequency fs.

  As shown in FIG. 10, in the mid-low range analog audio signal, the waveform indicated by the solid line and the waveform indicated by the broken line substantially coincide. On the other hand, as shown in FIG. 9, in the high-frequency analog audio signal, the waveform shown by the solid line and the waveform shown by the broken line are greatly different, and therefore distortion occurs. As the analog audio signal changes from the mid-low range to the high range, the number of sampling points for a waveform of one cycle decreases. For this reason, it becomes difficult to reproduce the waveform of the analog audio signal obtained by Lagrangian interpolation into the waveform of the original signal. Further, when Lagrangian interpolation is used, it is easily affected by image noise and distortion is likely to occur.

  In this way, when using oversampling by spline interpolation or Lagrange interpolation that smoothly connects sampling points near the high frequency range, the waveform of the analog audio signal obtained by spline interpolation or Lagrange interpolation is the same as the waveform of the original signal. It is difficult to reproduce the waveform, and distortion is likely to occur.

  For this reason, when the audio signal changes from the mid-low range to the high range, for example, if switching from oversampling by spline interpolation or Lagrange interpolation to oversampling by zero interpolation processing, the waveform approximates the waveform of the high range of the original signal. It can be reproduced as follows. However, noise may occur when switching from oversampling by spline interpolation or Lagrange interpolation to oversampling by zero interpolation processing.

  This point will be described with reference to FIG. FIG. 11 is a diagram illustrating a waveform of an analog audio signal in a high frequency (frequency near fs / 2). In FIG. 11, the horizontal axis represents time, and the vertical axis represents amplitude. The solid line shows the waveform of the original signal. A symbol “◯” indicates a sampling point when sampling is performed at the frequency fs.

  The dotted line shown in FIG. 11 shows the waveform of an analog audio signal obtained by DA-converting a digital audio signal obtained by oversampling a digital audio signal four times (4 fs) by Lagrange interpolation before the original signal is DA-converted. Show. A symbol “Δ” indicates a sampling point of four times (4 fs) oversampling by Lagrange interpolation. A broken line indicates a waveform of an analog audio signal obtained by DA-converting a digital audio signal obtained by over-sampling the digital audio signal four times by zero interpolation processing before the original signal is DA-converted. A symbol “□” indicates a sampling point of four times oversampling by zero interpolation processing.

  When switching from oversampling by Lagrange interpolation to oversampling by zero interpolation processing at the time point b shown in FIG. 11, the amplitude of the analog audio signal is changed from the sampling point code “Δ” by Lagrange interpolation at time point a to time point b. Changes to the sign “□” of the sampling point by zero interpolation processing in FIG. As described above, the waveform that changes in the direction of the arrow between the time point a and the time point b shown in FIG. 11 is a waveform whose amplitude is attenuated with respect to the waveform by Lagrange interpolation indicated by the dotted line. Therefore, the amplitude of the waveform changes discontinuously at the time of switching, causing noise.

  The present invention has been made to solve the above-described problem, and outputs an audio signal that does not generate noise when switching from oversampling by the first interpolation processing to oversampling by the second interpolation processing. An object is to provide an apparatus.

In order to solve the above-described problem, an invention according to claim 1 of the present application is an audio signal output device that performs output based on a digital audio signal, and input means for inputting the digital audio signal, and the polarity of the digital audio signal. and discriminating means for discriminating the frequency of inversion of a first interpolation processing means for generating a first first interpolated digital audio signal subjected to the interpolation process of the digital audio signal inputted from said input means and input Second interpolation processing means for applying the first interpolation processing or the second interpolation processing to a signal; input to the second interpolation processing means; the digital audio signal input from the input means; Selecting means for selecting between the first interpolated digital audio signal generated by the first interpolation processing means, and the second interpolation processing section; A first coefficient storage unit for storing a first coefficient for performing the first interpolation processing; and a second coefficient for causing the second interpolation processing unit to perform the second interpolation processing. A second coefficient storage unit for storing; a switching unit for switching a coefficient used in the second interpolation processing unit between the first coefficient and the second coefficient; selection of the selection unit; and Control means for controlling switching of the switching means, wherein the first interpolation processing inserts a zero signal between each sample of the digital audio signal, and then performs low-pass filter processing on the digital audio signal. A zero-order interpolation process, wherein the second interpolation process is a Lagrangian interpolation process, and the determination means determines whether or not the frequency of polarity inversion of the digital audio signal is equal to or higher than a predetermined reference frequency. And before The control means determines the second interpolation processing means by the selection means when it is determined that the frequency of polarity inversion of the digital audio signal is not equal to or higher than the predetermined reference frequency according to the determination result by the determination means. The input to the first interpolation digital audio signal generated by the first interpolation processing means is selected, and the switching of the coefficient used in the second interpolation processing section by the switching means is performed in the second And selecting the input to the second interpolation processing means by the selecting means when it is determined that the polarity inversion frequency of the digital audio signal is equal to or higher than the predetermined reference frequency. Is the digital audio signal input from the input means, and the switching means uses the switching of the coefficients used in the second interpolation processing unit. The first coefficient is controlled .

According to a second aspect of the present invention, in an audio signal output device that performs output based on a digital audio signal, reproduction means for reproducing a digital audio signal recorded on a recording medium, and reproduction by the reproduction means Discriminating means for discriminating the frequency of polarity inversion of the digital audio signal, and first interpolation for generating a first interpolated digital audio signal by applying a first interpolation process to the digital audio signal reproduced by the reproducing means Processing means, second interpolation processing means for performing the first interpolation processing or the second interpolation processing on the input signal, and the input to the second interpolation processing means reproduced by the reproduction means A selection selected between a digital audio signal and the first interpolated digital audio signal generated by the first interpolation processing means. Means, a first coefficient storage unit for storing a first coefficient for causing the second interpolation processing unit to perform the first interpolation processing, and the second interpolation processing unit to store the second interpolation. A second coefficient storage unit that stores a second coefficient for performing processing and a coefficient used in the second interpolation processing unit are switched between the first coefficient and the second coefficient. Switching means, and control means for controlling selection of the selection means and switching of the switching means, and the first interpolation processing includes inserting a zero signal between each sample of the digital audio signal, A zero-order interpolation process for performing a low-pass filter process on the audio signal; the second interpolation process is a Lagrangian interpolation process; and the discriminating means has a predetermined frequency inversion of the polarity of the digital audio signal. Standard frequency Determining whether or not, the control means, when it is determined that the frequency of polarity reversal of the digital audio signal is not equal to or higher than the predetermined reference frequency according to the determination result by the determination means, The selection of the input to the second interpolation processing means by the selection means is the first interpolation digital audio signal generated by the first interpolation processing means, and the second interpolation processing by the switching means. Switching the coefficient used in the unit to the second coefficient, and when the frequency of polarity inversion of the digital audio signal is determined to be greater than or equal to the predetermined reference frequency, The selection of the input to the second interpolation processing means is the digital audio signal reproduced by the reproduction means, and the second interpolation processing by the switching means. The switching of the coefficient used in the unit is controlled to be the first coefficient .

According to a third aspect of the present invention, in the audio signal output device according to the first or second aspect, the predetermined reference frequency is determined based on a sampling frequency of the digital audio signal. It is characterized by.

  According to the present invention, it is possible to provide an audio signal output device that does not generate noise when switching from oversampling by the first interpolation processing to oversampling by the second interpolation processing.

Embodiments of the present invention will be described below.
FIG. 1 is a block diagram showing a configuration of an audio signal output apparatus according to an embodiment of the present invention. In FIG. 1, 1 is an input unit, 2 is a DSP (Digital Signal Processor), 3 is a filter circuit, 4 is an output unit, 5 is a control unit, and 6 is an operation unit. The audio signal output device shown in FIG. 1 is an AV amplifier or the like that inputs a digital audio signal output from a playback device such as a CD player or a DVD (Digital Versatile Disc) player.

  The input unit 1 inputs a digital audio signal output from the playback device. The DSP 2 performs signal processing such as expansion of the digital audio signal input from the input unit 1 and addition of reverberation sound to the digital audio signal. As will be described later, the filter circuit 3 performs signal processing such as oversampling on the digital audio signal output from the DSP 2.

The output unit 4 includes a DA converter (not shown), and converts the digital audio signal input from the filter circuit 3 into an analog audio signal. The analog audio signal converted by the DA converter is output from a speaker or the like connected to an output terminal (not shown). The control unit 5 controls the input unit 1, DSP 2, filter circuit 3, and output unit 4. The operation unit 6 includes operation buttons for operating the audio signal output device of the present embodiment .

  Here, the present invention is not limited to the configuration shown in FIG. 1, and can also be applied to a playback apparatus such as a CD player or a DVD player. FIG. 2 is a block diagram showing a configuration of an audio signal output apparatus according to another embodiment of the present invention. In FIG. 2, the same components as those of the audio signal output apparatus shown in FIG.

The reproduction unit 7 reads a digital audio signal recorded on a disc such as a CD or a DVD. Digital audio signal read by the reproduction unit 7, a predetermined signal processing is carried out by DSP2 and the filter circuit 3 is converted into an analog audio signal by the output unit 4, it is outputted from a speaker or the like (not shown).

  FIG. 3 is a block diagram showing the configuration of the first embodiment of the filter circuit 3 provided in the audio signal output apparatus of this embodiment. In FIG. 3, 11 is a buffer, 12 is a zero interpolation processing unit, 13 is a buffer, 14 is a selector, 15 is a zero interpolation coefficient storage unit, 16 is a Lagrange interpolation coefficient storage unit, 17 is a selector, 18 is an interpolation processing unit, Reference numeral 19 denotes a high frequency signal detection processing unit, 20 denotes a high frequency signal pattern detection unit, 21 denotes a determination unit, 22 denotes a counter unit, and 23 denotes a selector control unit.

The buffer 11 shown in FIG. 3 temporarily holds the digital audio signal output from the DSP 2. The digital audio signal held in the buffer 11 is input to the zero interpolation processing unit 12 and the selector 14. The zero interpolation processing unit 12 is a first interpolation processing unit, and performs an oversampling process by a zero interpolation process on the input digital audio signal. The zero interpolation processing unit 12 is configured as a general FIR (Finite Impulse Response) filter. In the zero interpolation process, interpolation ( oversampling ) can be performed by adding a zero signal between sampling points.

  The zero interpolation signal oversampled by the zero interpolation processing unit 12 is a first interpolated audio signal and is input to the buffer 13. The buffer 13 temporarily holds the zero interpolation signal (digital audio signal) output from the zero interpolation processing unit 12. The zero interpolation signal held in the buffer 13 is input to the selector 14.

  FIG. 4 is a diagram showing oversampling processing in the zero interpolation processing unit 12. In FIG. 4, the horizontal axis represents time, and the vertical axis represents amplitude. FIG. 4A shows a digital audio signal on which oversampling processing is performed, and white circles indicate sampling points of this digital audio signal.

  In the case of double oversampling processing in the zero interpolation processing unit 12, as shown in FIG. 4B, a zero signal (black circle shown in the figure) is inserted in the middle of the sampling point of the digital audio signal. The digital audio signal into which the zero signal is inserted is passed through a low-pass filter, thereby generating a digital audio signal that has been subjected to oversampling twice as shown in FIG.

  At this time, since the digital audio signal shown in FIG. 4C is ½ of the amplitude of the digital audio signal shown in FIG. 4A, in order to restore the original amplitude, the digital audio signal shown in FIG. Double the amplitude of the digital audio signal. In this way, the zero interpolation processing unit 12 obtains a digital audio signal that has been subjected to double oversampling processing as shown in FIG.

  The waveform of the analog audio signal obtained when such zero interpolation processing is performed is not as smooth as the waveform obtained when Lagrange interpolation processing or spline interpolation processing is performed. However, zero interpolation processing that passes through a digital filter having a filter characteristic that can control image noise sufficiently can make the filter characteristic steep enough to cut image noise, and is therefore affected by image noise. Hateful. For this reason, in a high region where the waveform of the analog audio signal is complicated, a waveform with less distortion than the Lagrangian interpolation process or the spline interpolation process can be obtained.

  The selector 14 is selection means for selecting a signal to be output to the interpolation processing unit 18 from a digital audio signal input from the buffer 11 and a zero interpolation signal input from the buffer 13. The selector 14 selects a signal to be output to the interpolation processing unit 15 under the control of the selector control unit 23. The zero interpolation coefficient storage unit 15 is a first storage unit, and stores a zero interpolation coefficient (first coefficient) for causing the interpolation processing unit 18 to function as a zero interpolation processing unit. The Lagrange interpolation coefficient storage unit 16 is a second storage unit, and stores a Lagrange interpolation coefficient (second coefficient) for causing the interpolation processing unit 18 to function as a Lagrange interpolation processing unit.

  The selector 17 is switching means for switching a coefficient used in the interpolation processing unit 18 between a zero interpolation coefficient and a Lagrange interpolation coefficient. The selector 17 switches between the zero interpolation coefficient and the Lagrange interpolation coefficient under the control of the selector control unit 23. The interpolation processing unit 18 performs interpolation processing of the coefficients switched by the selector 17. When the selector 17 switches to the zero interpolation coefficient, the interpolation processing unit 18 performs an oversampling process by a zero interpolation process on the signal input from the selector 14. When the selector 17 switches to the Lagrangian interpolation coefficient, the interpolation processing unit 18 performs oversampling processing by Lagrange interpolation processing on the signal input from the selector 14. The interpolation processing unit is configured as a general FIR filter.

The interpolation processing unit 18 performs an oversampling process by a Lagrangian interpolation process using a Lagrange interpolation formula expressed by the following Equation 1. According to the Lagrangian interpolation formula, the amplitude value at an arbitrary sampling point X can be obtained from the amplitude values at (n + 1) sampling points.

  FIG. 5 is a diagram showing oversampling processing by Lagrange interpolation processing. In FIG. 5, the horizontal axis represents time, and the vertical axis represents amplitude. FIG. 5A shows a digital audio signal on which oversampling processing is performed, and white circles indicate sampling points of the digital audio signal. FIG. 5B shows a digital audio signal subjected to oversampling processing using a Lagrange interpolation formula.

  When the intermediate position of the sampling point of the digital audio signal shown in FIG. 5A is set as a position to be interpolated using the Lagrange interpolation formula, the sample data indicated by the black circle in FIG. 5B is interpolated. In this case, a double oversampled digital audio signal is generated. In this way, as shown in FIG. 5B, the interpolation processing unit 18 obtains a digital audio signal that has been subjected to double oversampling processing by Lagrange interpolation processing.

  The high frequency signal detection processing unit 19 detects a high frequency signal included in the digital audio signal input from the DSP 2. For example, in a digital audio signal having a sampling frequency of 44.1 kHz, the frequency of zero-crossing is high in a high frequency signal near ½ of the sampling frequency of 44.1 kHz. Zero crossing is when the polarity of the digital audio signal is inverted from positive (+) to negative (-), or from negative (-) to positive (+). That is, this is a case where one sign of two consecutive sampling data is positive (+) and the other sign is negative (-).

  FIG. 6 is a block diagram illustrating a configuration of the high frequency signal detection processing unit 19. In FIG. 6, the same components as those of FIG. As shown in FIG. 6, the high frequency signal pattern detection unit 20 includes a shift register 24 and a collation unit 25. The determination unit 21 includes a shift register 26 and a NOR circuit 27. The counter unit 22 includes a counter 28 and a NOT circuit 29.

  The high frequency signal pattern detection unit 20 shown in FIG. 6 detects zero-crossing twice or more in the digital audio signal input from the DSP 2 based on the sign of four consecutive sampling data. Here, since the MSB (Most Significant Bit) indicates a sign in the PCM (Pulse Code Modulation) signal of the audio signal, the high frequency signal pattern detection unit 20 detects a zero cross based on the MSB of the sampling data.

  The shift register 24 stores MSBs (QA, QB, QC, QD shown in FIG. 6) of four consecutive sampling data from the digital audio signal input from the DSP 2. The MSB of the PCM signal is a sign bit. When the polarity of the digital audio signal is positive (+), the sign of the MSB is “0”, and when the polarity of the digital audio signal is negative (−), the MSB Is “1”. The shift register 24 shifts the MSB to be stored and stores the MSB of the next sampling data every time a pulse of an operation clock with a sampling frequency (44.1 kHz in this embodiment) fs of the digital audio signal is input.

  The collation unit 25 collates 4-bit data consisting of a combination of four MSBs stored in the shift register 24 with the eight high-frequency signal patterns shown. The high frequency signal pattern is 4-bit data representing a combination of four MSBs when four consecutive sampling data zero-cross twice or more. When the combination of MSBs of two consecutive sampling data is (0, 1) or (1, 0), the digital audio signal is zero-crossed once. Therefore, as shown in FIG. 6, when the combination of MSBs is (0010), (0100), (0110), (1001), (1011), (1101), four consecutive sampling data are 2 Zero cross. When the combination of MSBs is (0101) and (1010), four consecutive sampling data zero-cross three times.

  The collation unit 25 stores the 4-bit data stored in the shift register 24 as (0010), (0100), (0101), (0110), (1001), (1010), (1011), (1101) 8 It is determined whether or not it matches any one of the two high frequency signal patterns. When the 4-bit data stored in the shift register 24 matches the high frequency signal pattern, the collation unit 25 detects the detection result data indicating that the high frequency signal pattern has been detected (“1” in this embodiment). Is output to the determination unit 21. On the other hand, when the 4-bit data stored in the shift register 24 does not match the above high frequency signal pattern, detection result data (“0” in this embodiment) indicating that the high frequency signal pattern has not been detected. It outputs to the determination part 21.

  The determination unit 21 stores four consecutive detection result data input from the high frequency signal pattern detection unit 20 and includes detection result data indicating that a high frequency signal pattern has been detected among the detection result data. It is determined whether or not. When the high frequency signal pattern is not detected, the determination unit 21 outputs a signal that causes the counter unit 22 to reset the counter 28 (hereinafter referred to as a reset signal).

The shift register 26 stores four consecutive detection result data (Qa, Qb, Qc, Qd shown in FIG. 6) input from the high frequency signal pattern detection unit 20. The shift register 26 shifts the detection result data to be stored and stores the next detection result data every time a pulse of an operation clock having a sampling frequency (44.1 kHz in this embodiment) fs of the digital audio signal is input. . As described above, the detection result data input from the high frequency signal pattern detection unit 20 becomes “1” when a high frequency signal pattern is detected, and becomes “0” when no high frequency signal pattern is detected.

  The NOR circuit 27 receives four detection result data stored in the shift register 26. The NOR circuit 27 outputs a reset signal to the counter unit 22 when the four detection result data input from the shift register 26, that is, the four detection result data stored in the shift register 26 are all “0”. . Further, the NOR circuit 27 does not output a reset signal to the counter unit 22 when there is “1” in the four detection result data input from the shift register 26.

  As described above, when the four consecutive detection result data input from the high frequency signal pattern detection unit 20 are all “0”, the determination unit 21 determines that the high frequency signal pattern detection unit 20 When “detection result data indicating that a signal pattern has not been detected” is input four times in succession, a reset signal is output to the counter unit 22.

  The counter 28 counts up the count value every time a pulse of an operation clock with a sampling frequency (44.1 kHz in the present embodiment) of the digital audio signal is input. The count value of the counter 28 is represented by a binary number. The counter 28 outputs a signal indicating the value of the bit (Qn) of the (n + 1) -th digit of the count value to the selector control unit 23 and also outputs to a built-in enable signal input unit (EN) via the NOT circuit 29. . Note that n is a natural number, and its value is determined in advance. In addition, when the reset signal is input from the determination unit 21, the counter 28 sets the counted up count value to zero.

  For example, when n is 2, the count values of the counter 28 are Q2, Q1, and Q0. When the counter 28 counts up the count value from (000) four times (square of 2), the count value becomes (100) and the third digit bit (Qn) becomes “1”. That is, when the bit (Qn) of the (n + 1) -th digit of the count value expressed in binary number is “1”, the counter 28 has counted up 2 n times.

  The counter 28 outputs a signal representing “0” to the selector control unit 23 when the bit (Qn) of the (n + 1) -th digit of the count value is “0”. When a signal representing “0” is input from the counter 28, the selector control unit 23 controls the selector 14 to select the zero interpolation signal input from the buffer 13 and switches the selector 17 to the Lagrange interpolation coefficient. To control. Therefore, the interpolation processing unit 18 functions as a Lagrangian interpolation processing unit, and a zero interpolation signal input from the buffer 13 via the selector 14 (that is, a zero interpolation signal subjected to zero interpolation processing by the zero interpolation processing unit 12). Lagrangian interpolation is applied to The signal representing “0” output from the counter 28 is inverted by the NOT circuit 29, and the signal representing “1” is input to the enable signal input section (EN).

  On the other hand, before the reset signal is input from the determination unit 21, when the counter 28 counts up to the nth power of 2, that is, when the bit (Qn) of the (n + 1) -th digit of the count value becomes “1”, the counter 28 outputs a signal representing “1” to the selector control unit 23. When a signal representing “1” is input from the counter 28, the selector control unit 23 controls the selector 14 to select the digital audio signal input from the buffer 11 and causes the selector 17 to switch to the zero interpolation coefficient. Take control. Therefore, the interpolation processing unit 18 functions as a zero interpolation processing unit, and performs zero interpolation processing on the digital audio signal input from the buffer 11 via the selector 14. The signal representing “1” output from the counter 28 is inverted by the NOT circuit 25, and the signal representing “0” is input to the enable signal input section (EN). When the signal input from the enable signal input unit (EN) becomes “0”, the counter 28 stops counting up the count value.

  In the high frequency portion near 1/2 of the sampling frequency, the high frequency signal pattern is constantly detected by the high frequency signal pattern detection unit 20, and the counter 28 of the counter unit 22 is not reset, and the counter 28 counts up the count value. Keep doing. When the bit (Qn) of the (n + 1) -th digit of the count value becomes “1”, the signal input to the output unit 4 is switched from the Lagrange interpolation signal to the zero interpolation signal under the control of the selector control unit 23. .

  When a reset signal is input from the determination unit 21, the counter 28 initializes the count value to (000). At this time, a signal representing “1” is input to the enable signal input unit via the NOT circuit 29, and the counter 28 starts counting up from (000). At this time, since a signal representing “0” is input to the selector control unit 23, the signal input to the output unit 4 is switched from the zero interpolation signal to the Lagrange interpolation signal.

  As described above, the high frequency signal detection processing unit 19 according to the present embodiment outputs a signal representing “0” or “1” to the selector control unit 23 based on the detection result by the high frequency signal pattern detection unit 20. . When a signal representing “1” is input to the selector control unit 23, that is, when the high frequency signal pattern detection unit 20 detects a high frequency signal pattern, it is input to the output unit 4 under the control of the selector control unit 23. The signal becomes a zero interpolation signal. The zero interpolation signal is a signal obtained by performing zero interpolation processing on the digital audio signal input from the buffer 11 by the interpolation processing unit 18.

  On the other hand, when a signal representing “0” is input to the selector control unit 23, that is, when the high-frequency signal pattern detection unit 20 does not detect a high-frequency signal pattern, it is input to the output unit 4 under the control of the selector control unit 23. This signal becomes a Lagrangian interpolation signal. This Lagrangian interpolation signal is a signal obtained by performing Lagrangian interpolation processing on the zero interpolation signal input from the buffer 13 by the interpolation processing unit 18.

  In this embodiment, even if a high frequency signal pattern is detected by the high frequency signal pattern detection unit 20, the Lagrangian interpolation signal is not immediately switched to the zero interpolation signal by the control of the selector control unit 23. Whether the state continues for a predetermined time is determined by the determination unit 21 and the counter unit 22. Based on the determination result, the signal input to the output unit 4 is switched from the Lagrange interpolation signal to the zero interpolation signal under the control of the selector control unit 23. For this reason, the frequency | count of switching the signal input into the output part 4 between a zero interpolation signal and a Lagrange interpolation signal can be decreased.

  Next, the analog audio signal waveform obtained by DA conversion of the zero interpolation signal output from the interpolation processing unit 15 and the analog signal obtained by DA conversion of the Lagrangian interpolation signal output from the interpolation processing unit 15 The waveform of the audio signal will be described.

  FIG. 7 is a diagram illustrating waveforms of analog audio signals of a high-frequency zero interpolation signal and a Lagrange interpolation signal. In FIG. 7, the horizontal axis represents time, and the vertical axis represents amplitude. A solid line indicates a waveform of an analog audio signal obtained by DA converting the digital audio signal input from the DSP 2. A symbol “◯” indicates a sampling point when sampling is performed at the frequency fs.

  A broken line shown in FIG. 7 indicates the waveform of an analog audio signal obtained by DA-converting a zero interpolation signal obtained by performing zero interpolation processing on the digital audio signal input from the buffer 11 via the selector 14 by the interpolation processing unit 15. Indicates. A symbol “□” indicates a sampling point interpolated by the zero interpolation process. A dotted line indicates a waveform of an analog audio signal obtained by DA-converting a Lagrangian interpolation signal obtained by performing a Lagrange interpolation process on the zero interpolation signal input from the buffer 13 via the selector 14 by the interpolation processing unit 15. Indicates. A symbol “Δ” indicates a sampling point interpolated by Lagrange interpolation.

  As described above, since the Lagrangian interpolation signal is a signal obtained by performing the Lagrangian interpolation processing on the zero interpolation signal by the interpolation processing unit 18, the sampling point (symbol “Δ”) of this Lagrange interpolation signal is As shown in FIG. 7, it coincides with the sampling point (symbol “□”) of the zero interpolation signal at the same time as the sampling point (symbol “◯”) of the original analog audio signal at time b. That is, at the time point b, the amplitude of the analog waveform of the Lagrangian interpolation signal indicated by the dotted line and the analog waveform of the zero interpolation signal indicated by the broken line match.

  Therefore, in this embodiment, when a signal representing “1” is input from the counter 28, the selector control unit 23 receives a zero signal at the sampling point of the original digital audio signal. By controlling the selector 14 and the selector 17 so as to be an interpolation signal, the waveform changes continuously at the time of switching. For example, when the signal input to the output unit 4 is switched from the Lagrangian interpolation signal to the zero interpolation signal at time b shown in FIG. 7, the amplitude of the analog audio signal is the code “ The symbol “□” changes to the symbol “□” of the sampling point by the zero interpolation process at time point b.

  At this time, since the sampling point by the zero interpolation process at the time point b coincides with the sampling point by the Lagrange interpolation, the waveform from the time point a to the time point b becomes an analog waveform of the Lagrange interpolation signal indicated by the dotted line, and the broken line from the time point b. The analog waveform of the zero interpolation signal indicated by For this reason, since the waveform changes continuously at the time of switching, no noise is generated. Even when a signal representing “0” is input from the counter 28 and the selector control unit 23 is controlled so that the signal input to the output unit 4 becomes a Lagrangian interpolation signal, the waveform continues at the time of switching. Therefore, noise does not occur.

  As described above, the selector control unit 23 switches the signal input to the output unit 4 at the position of the sampling point of the original digital audio signal, so that the waveform continuously changes at the time of switching, and noise is generated. Absent. Therefore, the audio signal output apparatus according to the present embodiment has no noise generated due to the waveform discontinuously changing when the signal output to the output unit 4 is switched.

  Next, a second embodiment of the filter circuit 3 will be described. FIG. 8 is a block diagram showing the configuration of the second embodiment of the filter circuit 3 provided in the audio signal output apparatus of this embodiment. In FIG. 8, the same components as those of FIG. The filter circuit of the second embodiment has a configuration that does not include the high frequency signal detection processing unit 19 provided in the filter circuit 3 of the first embodiment.

  The selector control unit 23 illustrated in FIG. 8 performs control to switch the signal input to the output unit 4 to a Lagrange interpolation signal or a zero interpolation signal in accordance with a control signal input from the control unit 5. The operation unit 6 includes a selection switch that selects a signal to be input to the output unit 4 between a zero interpolation signal and a Lagrange interpolation signal. The control unit 5 controls the output unit 4 according to the operation of the selection switch. A control signal for switching the input signal is output to the selector control unit 23.

  When a control signal is input from the control unit 5, the selector control unit 23 performs control to switch a signal input to the output unit 4 at the position of the sampling point of the original digital audio signal. Therefore, as described above, since the waveform continuously changes when the signal output to the output unit 4 is switched, no noise is generated at the time of switching.

  Further, since the user can select the signal to be input to the output unit 4 by operating the selection switch at any time, the user can select the signal according to the genre of music to be listened to or the user's preference. The selection switch can be operated. For example, when the music genre is rock or pop, the zero interpolation signal is selected, and when the music genre is classical or jazz, the Lagrange interpolation signal is selected. Therefore, the audio signal output device of the present embodiment can output an audio signal according to the music genre or the user's preference.

1 is a block diagram showing the configuration of an audio signal output apparatus that is an embodiment of the present invention. The block diagram which shows the structure of the audio signal output apparatus which is another Example. The block diagram which shows the structure of the filter circuit 3 of a 1st Example. Explanatory drawing of the oversampling process in the zero interpolation process part 12. FIG. Explanatory drawing of the oversampling process of a Lagrange interpolation process. The block diagram which shows the structure of the high frequency signal detection process part 19. FIG. The figure which shows the waveform of the analog audio signal of the zero interpolation signal of a high region, and a Lagrange interpolation signal. The block diagram which shows the structure of the filter circuit of a 2nd Example. The figure which shows the waveform of the analog audio signal of a high region. The figure which shows the waveform of the analog audio signal of a mid-low range. The figure which shows the waveform of the analog audio signal of a high region.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Input part, 2 ... DSP, 3 ... Filter circuit, 4 ... Output part, 5 ... Control part, 6 ... Operation part, 7 ... Reproducing part, 11 ... Buffer, 12 ... Zero interpolation process part, 13 ... Buffer, 14 ... Selector, 15 ... Zero interpolation coefficient storage unit, 16 ... Lagrange interpolation coefficient storage unit, 17 ... Selector, 18 ... Interpolation processing unit, 19 ... High frequency signal detection processing unit, 20 ... High frequency signal pattern detection unit, 21 ... Judgment unit, 22 ... counter unit, 23 ... selector control unit, 24 ... shift register, 25 ... collation unit, 26 ... shift register, 27 ... NOR circuit, 28 ... counter, 29 ... NOT circuit

Claims (3)

In an audio signal output device that performs output based on a digital audio signal ,
An input means for inputting a digital audio signal;
Discriminating means for discriminating the frequency of polarity inversion of the digital audio signal;
First interpolation processing means for performing a first interpolation process on the digital audio signal input from the input means to generate a first interpolation digital audio signal;
Second interpolation processing means for applying the first interpolation processing or the second interpolation processing to the input signal;
The input to the second interpolation processing means is selected between the digital audio signal input from the input means and the first interpolation digital audio signal generated by the first interpolation processing means. A selection means;
A first coefficient storage section for storing a first coefficient for causing the second interpolation processing section to perform the first interpolation processing;
A second coefficient storage section for storing a second coefficient for causing the second interpolation processing section to perform the second interpolation processing;
Switching means for switching a coefficient used in the second interpolation processing unit between the first coefficient and the second coefficient;
Control means for controlling selection of the selection means and switching of the switching means,
The first interpolation process is a zero-order interpolation process for performing a low-pass filter process on the digital audio signal after inserting a zero signal between each sample of the digital audio signal,
The second interpolation process is a Lagrangian interpolation process,
The determining means determines whether the polarity inversion frequency of the digital audio signal is equal to or higher than a predetermined reference frequency;
The control means determines the second interpolation process by the selection means when it is determined that the frequency of polarity inversion of the digital audio signal is not equal to or higher than the predetermined reference frequency according to the determination result by the determination means. The selection of the input to the means is the first interpolated digital audio signal generated by the first interpolation processing means, and the switching of the coefficients used in the second interpolation processing unit by the switching means is performed in the first way. When the frequency of inversion of the polarity of the digital audio signal is determined to be equal to or higher than the predetermined reference frequency, the input to the second interpolation processing unit by the selection unit is controlled. The selection is the digital audio signal input from the input means, and the coefficient used in the second interpolation processing unit is switched by the switching means. Audio signal output apparatus characterized by controlling so that said first coefficient. In an audio signal output device that performs output based on a digital audio signal ,
Reproducing means for reproducing the digital audio signal recorded on the recording medium;
Discriminating means for discriminating the frequency of polarity inversion of the digital audio signal reproduced by the reproducing means ;
First interpolation processing means for performing a first interpolation process on the digital audio signal reproduced by the reproduction means to generate a first interpolation digital audio signal;
Second interpolation processing means for applying the first interpolation processing or the second interpolation processing to the input signal;
The input to the second interpolation processing means is selected between the digital audio signal reproduced by the reproduction means and the first interpolated digital audio signal generated by the first interpolation processing means. A selection means;
A first coefficient storage section for storing a first coefficient for causing the second interpolation processing section to perform the first interpolation processing;
A second coefficient storage section for storing a second coefficient for causing the second interpolation processing section to perform the second interpolation processing;
Switching means for switching a coefficient used in the second interpolation processing unit between the first coefficient and the second coefficient;
Control means for controlling selection of the selection means and switching of the switching means,
The first interpolation process is a zero-order interpolation process for performing a low-pass filter process on the digital audio signal after inserting a zero signal between each sample of the digital audio signal,
The second interpolation process is a Lagrangian interpolation process,
The determining means determines whether the polarity inversion frequency of the digital audio signal is equal to or higher than a predetermined reference frequency;
The control means determines the second interpolation process by the selection means when it is determined that the frequency of polarity inversion of the digital audio signal is not equal to or higher than the predetermined reference frequency according to the determination result by the determination means. The selection of the input to the means is the first interpolated digital audio signal generated by the first interpolation processing means, and the switching of the coefficients used in the second interpolation processing unit by the switching means is performed in the first way. When the frequency of inversion of the polarity of the digital audio signal is determined to be equal to or higher than the predetermined reference frequency, the input to the second interpolation processing unit by the selection unit is controlled. The selection is the digital audio signal reproduced by the reproducing means, and the coefficient used in the second interpolation processing unit is switched by the switching means. Audio signal output apparatus characterized by controlling so that said first coefficient. The audio signal output device according to claim 1 or 2,
The audio signal output apparatus according to claim 1, wherein the predetermined reference frequency is determined based on a sampling frequency of the digital audio signal.

Images