JP2009237048A - Audio signal interpolation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an audio signal interpolation device capable of reducing low frequency noise to surroundings, when an audio signal is reproduced by emphasizing low sound, by interpolating a high frequency component which is well correlated with a fundamental sound part, for an audio signal which loses the high frequency component due to compression processing. <P>SOLUTION: The audio signal interpolation device comprises: a high region interpolation means for interpolating a high frequency band to the audio signal; a low region emphasizing means for emphasizing a low frequency band of the audio signal by adding a plurality of harmonics of a basic frequency; and a filter means in which the high frequency component is interpolated by the high region interpolation means, and in which a predetermined low frequency component is removed from the audio signal in which the low frequency component is emphasized by the low region emphasizing means. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an audio signal interpolation apparatus that interpolates a high frequency component and a low frequency component in an audio signal.

  In audio data such as MP3, a signal of a high frequency component in a high frequency band (for example, 16 kHz or more) is cut when compression processing is performed. For this reason, audio data that has been subjected to compression processing such as MP3 has poor sound quality as compared to the audio signal before compression processing. In order to improve the audio quality of such audio data, for example, as disclosed in Patent Document 1, there is means for interpolating high frequency components lost due to compression processing into audio data.

  In addition, by using an auditory phenomenon called human missing fundamental, a sound corresponding to a fundamental frequency that does not actually exist can be perceived by humans. For example, when sounds of 100 Hz, 150 Hz, and 200 Hz are emitted, it is possible to perceive sounds around 50 Hz that is their fundamental frequency. By using the missing fundamental phenomenon in this way, the listener can play low-pitched sound when playing compressed audio data such as MP3 or when using a small speaker that is difficult to reproduce the low-frequency range. You can listen to the audio signal emphasized.

JP 2002-175092 A

  The means disclosed in Patent Document 1 copies a part of the high-frequency component of the band-limited audio signal, and adds the copied high-frequency component to the audio signal, so that the high-frequency component lost by the compression processing is added. Is interpolated. In such interpolation means, the correlation between the high frequency component to be added and the fundamental portion of the audio signal is low, and the interpolated audio signal may be unnaturally heard by the listener.

  Moreover, when listening to the reproduced audio signal, noise in the surroundings is not considered in many cases. For this reason, it is required to reduce sound leakage to the surroundings as much as possible. In particular, when emphasizing the low frequency range, the sound wave is more likely to pass through the wall as the frequency becomes lower, so a measure against low frequency noise is required.

  The present invention has been made to solve the above-described problem, and can interpolate a high-frequency component having a good correlation with a fundamental part with respect to an audio signal in which a high-frequency component has been lost by compression processing. An object of the present invention is to provide an audio signal interpolation apparatus. It is another object of the present invention to provide an audio signal interpolating device that can reduce low frequency noise to the surroundings when reproducing an audio signal with emphasis on bass.

  In order to solve the above problems, an audio signal interpolating apparatus according to the present invention emphasizes a low frequency band of an audio signal by adding a high frequency interpolating means for interpolating a high frequency band to the audio signal and a plurality of harmonics of a fundamental frequency. Low-frequency emphasizing means, and filter means for removing a predetermined low-frequency component from an audio signal in which high-frequency components are interpolated by the high-frequency interpolating means and low-frequency components are emphasized by the low-frequency emphasizing means. Features.

  In the audio signal interpolating apparatus, the high frequency interpolating means includes an input means for inputting a first audio signal and a second audio signal, a first audio signal input from the input means, and a second audio signal. Delay means for delaying the audio signal, phase division means for dividing the first audio signal and the second audio signal input from the input means into an in-phase component signal and a difference phase component signal, and the in-phase component Interpolating processing means for performing processing for interpolating a high frequency component on the signal and the difference phase component signal; and synthesizing the same phase component signal and the difference phase component signal interpolated by the interpolation processing means with the high frequency component; And a phase synthesizing unit for outputting the fourth audio signal, and a high-pass filter process on the third audio signal outputted from the phase synthesizing unit First high-pass filter means for applying, second high-pass filter means for applying high-pass filter processing to the fourth audio signal output from the phase synthesizing means, the first audio signal delayed by the delay means, A first addition processing unit for adding signals output from the first high-pass filter means; a second audio signal delayed by the delay means; and a signal output from the second high-pass filter means. And a second addition processing unit.

  In the audio signal interpolating apparatus, the high-frequency interpolation processing means includes a cutoff frequency detecting means for detecting a cutoff frequency of the in-phase component signal and the difference phase component signal, and the cutoff frequency detecting means. Envelope generating means for generating envelope information at the cut-off frequency of the detected in-phase component signal and generating envelope information at the cut-off frequency of the difference phase component signal detected by the cut-off frequency detecting means And defining a frequency band for interpolating a high frequency component from the cutoff frequency of the in-phase component signal according to the envelope information of the in-phase component signal, interpolating the frequency band, and enveloping the difference phase component signal in the envelope information of the difference phase component signal Accordingly, a frequency band for interpolating high frequency components from the cut-off frequency of the differential phase component signal is determined, and the frequency band Characterized in that it comprises a high-frequency component interpolating means for interpolating.

  In the above audio signal interpolating apparatus, the frequency band in which the high frequency component interpolating means interpolates the high frequency component is not more than a predetermined Nyquist frequency.

  In the audio signal interpolating apparatus, the low-frequency emphasizing means includes a plurality of bandpass filter means having harmonics of a fundamental frequency as a center frequency, and amplification for amplifying each audio signal output from the bandpass filter means Means and an adding means for adding the audio signals amplified by the amplifying means.

  ADVANTAGE OF THE INVENTION According to this invention, the audio signal interpolation apparatus which can interpolate the high frequency component with a good correlation with a fundamental part with respect to the audio signal from which the high frequency component was lost by the compression process can be provided. It is another object of the present invention to provide an audio signal interpolating device that can reduce low frequency noise to the surroundings when reproducing an audio signal with emphasis on bass.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing the configuration of an audio signal interpolating apparatus according to an embodiment of the present invention. As shown in FIG. 1, the audio signal interpolating apparatus 10 according to the present embodiment includes an input unit 20, a high frequency interpolating unit 30, a low frequency emphasizing unit 40, a high pass filter 50, and an output unit 60. The audio signal interpolating apparatus according to the present embodiment is provided in a player capable of reproducing compressed audio data such as an AV amplifier and MP3.

  The audio signal interpolating apparatus 10 of this embodiment inputs a left channel (Lch) audio signal and a right channel (Rch) audio signal of a stereo audio signal, which are digital signals, from an input unit 10. The input Lch and Rch audio signals are interpolated with high-frequency components by the high-frequency interpolation unit 30. The Lch and Rch audio signals are interpolated by the bass emphasis unit 40 so that the low frequency components are enhanced. The audio signal in which the high-frequency component and the low-frequency component are interpolated is output from the output unit 60 after the predetermined low-frequency component (for example, 50 Hz or less) is removed by the high-pass filter 50.

  FIG. 2 is a block diagram showing the configuration of the high-frequency interpolation unit 30 of this embodiment. As shown in FIG. 2, the high-frequency interpolation unit 30 includes a phase division unit 31, an interpolation processing unit 32, a phase synthesis unit 33, a filter unit 34, an addition processing unit 35, a delay unit 36, and a delay unit 37.

  As shown in FIG. 2, the Lch audio signal input from the input unit 20 is input to the phase dividing unit 31 and the delay unit 36, and the Rch audio signal from the input unit 20 is input to the phase dividing unit 31 and the delay unit 37. Is done. The phase dividing unit 31 includes a synthesizing unit 311 and a synthesizing unit 312, and divides the Lch and Rch audio signals input from the input unit 20 into signals of the same phase component | L + R | and difference phase component | LR |. . The in-phase component signal is obtained by synthesizing the Lch audio signal and the Rch audio signal by the synthesis unit 311. The differential phase component signal is obtained by inverting the Lch audio signal by the synthesis unit 312 and synthesizing it with the Rch audio signal.

  The interpolation processing unit 32 includes a cutoff frequency detection unit 321, an envelope generation unit 322, and an interpolation unit 323 in order to perform processing for interpolating a high frequency range component on the input in-phase component signal. The cutoff frequency detection unit 321 detects the cutoff frequency fc of the in-phase component signal input to the interpolation processing unit 32 using fast Fourier transform or the like. The envelope generation unit 322 performs a cepstrum analysis from a spectrum distribution generated by performing a fast Fourier transform process on the in-phase component signal input to the interpolation processing unit 32, thereby detecting the cutoff detected by the cutoff frequency detection unit 321. Envelope information at the frequency fc is generated. The interpolation unit 323 determines a frequency band for interpolating the high frequency range component from the detected cutoff frequency fc according to the generated envelope information, and sets the frequency band of the in-phase component signal input to the interpolation processing unit 32. Interpolates high frequency components.

  In addition, the interpolation processing unit 32 includes a cutoff frequency detection unit 324, an envelope generation unit 325, and an interpolation unit 326 in order to perform processing for interpolating the high frequency range component on the input differential phase component signal. The cut-off frequency detection unit 324 detects the cut-off frequency fc of the difference phase component signal input to the interpolation processing unit 32 using fast Fourier transform or the like. The envelope generation unit 325 performs a cepstrum analysis from a spectrum distribution generated by performing a fast Fourier transform process on the differential phase component signal input to the interpolation processing unit 32, thereby detecting the cutoff detected by the cutoff frequency detection unit 324. Envelope information at the frequency fc is generated. The interpolation unit 326 determines a frequency band for interpolating the high frequency range component from the detected cutoff frequency fc according to the generated envelope information, and sets the frequency band of the difference phase component signal input to the interpolation processing unit 4 to the frequency band. Interpolates high frequency components.

  The phase synthesizing unit 33 includes a synthesizing unit 331 and a synthesizing unit 332, synthesizes the same phase component signal and the difference phase component signal input from the interpolation processing unit 32, and outputs an Lch audio signal and an Rch audio signal. The combining unit 331 outputs an Lch audio signal obtained by combining the in-phase component signal and the difference phase component signal. The combining unit 332 outputs an Rch audio signal obtained by combining the inverted in-phase component signal and the difference phase component signal.

  The filter unit 34 includes a high pass filter 341 and a high pass filter 342. The high-pass filter 341 removes the low-frequency component of the Lch audio signal output from the synthesis unit 331 at the cutoff frequency fc detected by the cutoff frequency detection unit 321. The high-pass filter 342 removes the low-frequency component of the Rch audio signal output from the synthesis unit 332 at the cutoff frequency fc detected by the cutoff frequency detection unit 324.

  The addition processing unit 35 includes an addition unit 351 and an addition unit 352. The adder 351 adds the Lch audio signal from which the low frequency range component output from the high pass filter 341 is removed and the Lch audio signal output from the delay unit 36. The adding unit 352 adds the Rch audio signal from which the low frequency range component output from the high pass filter 342 is removed and the Rch audio signal output from the delay unit 37.

  The delay unit 36 delays the Lch audio signal input from the input unit 20 by a time corresponding to the phase delay caused by the processing of the phase division unit 31, the interpolation processing unit 32, the phase synthesis unit 33, and the filter unit 34. The delay unit 37 delays the Rch audio signal input from the input unit 20 by a time corresponding to the phase delay caused by the processing of the phase division unit 31, the interpolation processing unit 32, the phase synthesis unit 33, and the filter unit 35.

  Next, an interpolation process by the interpolation processing unit 32 in the audio signal interpolating apparatus 1 of the present embodiment will be described. FIG. 3 is a diagram for explaining high-frequency component interpolation processing.

  In the graph showing the spectrum of the in-phase component signal shown in FIG. 3A, fc is the cutoff frequency of the in-phase component signal detected by the cutoff frequency detector 321 and fn is the Nyquist frequency. In the graph representing the spectrum of the differential phase component signal shown in FIG. 3B, fc is the cutoff frequency of the differential phase component signal detected by the cutoff frequency detector 324, and fn is the Nyquist frequency. Since the audio signal input to the audio signal interpolating apparatus 1 is a stereo signal, the cut-off frequency fc shown in FIG. 3A and the cut-off frequency fc shown in FIG. 3B are substantially the same frequency. The Nyquist frequency fn shown in FIG. 3A and the Nyquist frequency shown in FIG. 3B are substantially the same frequency. When the stereo audio signal is audio data subjected to compression processing such as MP3, the cut-off frequency fc is 16 kHz. The Nyquist frequency fn is, for example, 22.05 kHz.

  The envelope shown in FIG. 3A is an envelope at the cut-off frequency fc generated by the envelope generator 322, and the slope thereof is expressed as COMM. The envelope shown in FIG. 3B is an envelope at the cutoff frequency fc generated by the envelope generator 45, and the slope thereof is expressed as DIFF. In this embodiment, the slope COMM of the envelope of the in-phase component signal is steeper than the slope DIFF of the envelope of the difference phase component signal. In general, a stereo audio signal includes many harmonic components such as echo components, reverberation components, and reverb components up to the high frequency range of the differential phase component signal. In addition, harmonic components such as vocal sounds and musical instrument fundamentals are often included in the in-phase component signal, and are attenuated as the pitch increases.

  Usually, an audio signal has a spectrum component that decreases as it goes up. For this reason, as described above, the in-phase component signal and the differential phase component signal also decrease in spectrum components as they become higher, but there is a difference in how they are reduced. In the present embodiment, a signal closer to the original sound is obtained by interpolating the high frequency component along the envelope at the cutoff frequency fc of the in-phase component signal and the difference phase component signal by using the difference in decrease of the spectrum component. It is possible to interpolate to

  In the interpolation processing unit 32, the interpolation unit 323 performs a fast Fourier transform analysis on the input in-phase component signal and performs frequency shift processing or the like, so that the interpolation processing unit 323 can perform the interval from the cutoff frequency fc to the Nyquist frequency along the slope COMM envelope. The high frequency component is interpolated in the frequency band. As shown in FIG. 3A, when the frequency f at the intersection of the envelope and the frequency axis is lower than the Nyquist frequency fn (that is, when fc <f <fn), the interpolation unit 323 calculates the frequency from the cutoff frequency fc. High frequency components are interpolated in the frequency band up to the frequency f of the intersection. Thus, the high frequency component interpolated by the interpolating unit 323 into the in-phase component signal becomes a hatched region shown in FIG.

  Further, the interpolation unit 326 performs a fast Fourier transform analysis on the input differential phase component signal and performs frequency shift processing or the like, so that the frequency band between the cutoff frequency fc and the Nyquist frequency along the slope DIFF envelope is obtained. Interpolates high frequency components. As shown in FIG. 3B, since the frequency f at the intersection of the envelope and the frequency axis is higher than the Nyquist frequency fn, the interpolation unit 326 generates a high frequency component in the frequency band from the cutoff frequency fc to the Nyquist frequency fn. Interpolate. In this way, the high frequency component interpolated into the differential phase component signal by the interpolating unit 326 becomes a shaded area shown in FIG.

  As shown in FIG. 3, the in-phase component signal and the difference phase component signal interpolated with the high-frequency component are combined by the phase combining unit 33 into an Lch audio signal and an Rch audio signal. From the Lch and Rch audio signals, the low frequency range component is removed at the cutoff frequency fc by the filter unit 34, and the high frequency components of the Lch and Rch interpolated by the interpolation processing unit 32 are extracted.

  The Lch and Rch high frequency components extracted by the filter unit 34 are added by the addition processing unit 35 to the Lch and Rch audio signals output from the delay unit 36 and the delay unit 37, respectively. Here, the Lch and Rch audio signals input to the addition processing unit 35 are delayed by the delay unit 36 and the delay unit 37 so as to be the same audio signal as the audio signal interpolated by the interpolation processing unit 32. Is.

  As described above, the high-frequency interpolation unit 30 according to the present embodiment can interpolate a high-frequency component having a good correlation with the fundamental part with respect to an audio signal from which a high-frequency component has been lost by the compression process. As a result, an audio signal in which the high frequency component is interpolated is not unnaturally heard by the listener.

  Next, the configuration of the low frequency emphasizing unit 40 provided in the audio signal correcting apparatus 10 of the present embodiment will be described. FIG. 4 is a block diagram showing the configuration of the low frequency emphasizing unit 40 of the present embodiment. As shown in FIG. 4, the low-frequency emphasis unit 40 includes an adder 41, a bandpass filter 42a, a bandpass filter 42b, a bandpass filter 42c, an amplifier 43a, an amplifier 43b, an amplifier 43c, an adder 44, and an adder. Unit 45 and addition unit 46.

  The Lch and Rch audio signals whose high frequency components are interpolated by the high frequency interpolating unit 30 are input to the adding unit 41. The Lch and Rch audio signals added by the adder 41 are input to the bandpass filter 42a, the bandpass filter 42b, and the bandpass filter 42c. The band-pass filter 42a, the band-pass filter 42b, and the band-pass filter 42c are set so that the fundamental frequency is 50 Hz, and the frequency that is a harmonic of 50 Hz is the center frequency. In this embodiment, the bandpass filter 42a has a center frequency set to 100 Hz, the bandpass filter 42b has a center frequency set to 150 Hz, and the bandpass filter 42c has a center frequency set to 200 Hz.

  The amplifying unit 43a amplifies the signal having a center frequency of 100 Hz input from the bandpass filter 42a to a predetermined level. The amplifying unit 43b amplifies the signal having a center frequency of 150 Hz input from the bandpass filter 42b to a predetermined level. The amplifying unit 43c amplifies the signal having a center frequency of 200 Hz input from the bandpass filter 42c to a predetermined level. The adding unit 44 adds the signals output from the amplifying unit 43a, the amplifying unit 43b, and the amplifying unit 43c. The signal output from the adder 44 is added to the Lch audio signal in which the high frequency component is interpolated by the adder 45 and is added to the Rch audio signal in which the high frequency component is interpolated by the adder 46.

  As described above, the low-frequency emphasis unit 40 according to the present embodiment adds the signals of the 100 Hz band, the 150 Hz band, and the 200 Hz band, which are overtones of the basic frequency 50 Hz, to the Lch and Rch audio signals. For this reason, the listener can hear the bass sound around the fundamental frequency of 50 Hz as emphasized by the missing fundamental phenomenon. As described above, in this embodiment, the bass is emphasized by using the missing fundamental phenomenon, so that the emphasized bass is similar to the high frequency component interpolated by the high-frequency interpolation unit 30. Does not sound unnatural.

  The audio signal in which the high-frequency component and the low-frequency component are interpolated by the high-frequency interpolation unit 30 and the low-frequency emphasizing unit 40 are subjected to removal of a predetermined low-frequency component (for example, 50 Hz or less) by the high-pass filter 50. Is output. When the audio signal interpolating apparatus 10 of this embodiment is provided in an AV amplifier, the audio signal output from the output unit 60 is converted into an analog audio signal by a DA converter and then amplified by an amplifying unit. And it outputs from the speaker of Lch and Rch connected with AV amplifier.

  Therefore, even if the listener reproduces audio data from which high-frequency components have been cut during compression processing, the listener can listen to the audio signal in which the high-frequency components are interpolated without any sense of incongruity, and the audio signal in which the bass is emphasized. Can be heard. Further, even when audio data from which low frequency components have been cut during the compression processing is reproduced, the low sound can be emphasized, that is, the low frequency components can be interpolated. Furthermore, since the audio signal of 50 Hz or less is removed by the high-pass filter 50, low frequency noise to the surroundings can be reduced even if the low sound is emphasized.

The block diagram which shows the structure of the audio signal interpolation apparatus of a present Example. The block diagram which shows the structure of the high frequency interpolation part 30 of a present Example. The figure explaining the interpolation process of a high frequency component. The block diagram which shows the structure of the low frequency emphasis part 40 of a present Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Audio signal interpolation apparatus, 20 ... Input part, 30 ... High-frequency interpolation part, 40 ... Low-frequency emphasis part, 50 ... High pass filter, 60 ... Output part, 31 ... Phase division part, 32 ... Interpolation processing part, 33 ... Phase synthesis unit 34... Filter unit 35. Addition processing unit 36. Delay unit 37 37 Delay unit 311 Synthesis unit 312 Synthesis unit 321 Cut-off frequency detection unit 322 Envelope generation unit 323: Interpolation unit, 324 ... Cutoff frequency detection unit, 325 ... Envelope generation unit, 326 ... Interpolation unit, 331 ... Synthesis unit, 332 ... Synthesis unit, 341 ... High pass filter unit, 342 ... High pass filter unit, 351 ... Synthesis 352... Synthesizer 41... Adder 42 a. Bandpass filter 42 b Bandpass filter 42 c Bandpass filter 43 a Amplifier 43 b Amplifier 43 c Width portion, 44 ... adding unit, 45 ... adding unit, 46 ... adding unit

Claims (5)

High-frequency interpolation means for interpolating a high-frequency band into an audio signal;
Low-frequency emphasis means for emphasizing the low-frequency band of the audio signal by adding multiple harmonics of the fundamental frequency;
An audio signal interpolation comprising: a filter means for removing a predetermined low frequency component from an audio signal in which a high frequency component is interpolated by the high frequency interpolation means and a low frequency component is emphasized by the low frequency emphasis means apparatus. The audio signal interpolating device according to claim 1,
The high frequency interpolation means includes
Input means for inputting a first audio signal and a second audio signal;
Delay means for delaying the first audio signal and the second audio signal input from the input means;
Phase dividing means for dividing the first audio signal and the second audio signal input from the input means into an in-phase component signal and a differential phase component signal;
Interpolation processing means for performing processing for interpolating a high-frequency component to the in-phase component signal and the differential phase component signal;
Phase synthesizing means for synthesizing the same phase component signal and the difference phase component signal interpolated by the interpolation processing means with the high frequency component, and outputting a third audio signal and a fourth audio signal;
First high-pass filter means for performing high-pass filter processing on the third audio signal output from the phase synthesis means;
Second high-pass filter means for applying high-pass filter processing to the fourth audio signal output from the phase synthesis means;
A first addition processing unit that adds the first audio signal delayed by the delay unit and the signal output from the first high-pass filter unit;
An audio signal interpolating apparatus comprising: a second addition signal processing unit that adds the second audio signal delayed by the delay unit and the signal output from the second high-pass filter unit. The audio signal interpolating device according to claim 1 or 2,
The high frequency interpolation processing means includes
A cutoff frequency detecting means for detecting a cutoff frequency of the in-phase component signal and the differential phase component signal;
Envelope information at the cutoff frequency of the in-phase component signal detected by the cutoff frequency detection means is generated, and envelope information at the cutoff frequency of the difference phase component signal detected by the cutoff frequency detection means Envelope generating means for generating
A frequency band for interpolating a high frequency component from the cutoff frequency of the in-phase component signal is determined according to the envelope information of the in-phase component signal, and the frequency band is interpolated to correspond to the envelope information of the difference phase component signal. An audio signal interpolating apparatus comprising: high frequency component interpolation means for determining a frequency band for interpolating a high frequency component from a cutoff frequency of the differential phase component signal and interpolating the frequency band. The audio signal interpolating device according to claim 3,
The audio signal interpolating apparatus according to claim 1, wherein a frequency band in which the high frequency component interpolating means interpolates the high frequency component is equal to or less than a predetermined Nyquist frequency. The audio signal interpolating device according to any one of claims 1 to 4,
The low frequency emphasis means is
A plurality of band-pass filter means having a harmonic of a fundamental frequency as a center frequency;
Amplifying means for amplifying each audio signal output from the bandpass filter means;
An audio signal interpolating apparatus comprising: an adding unit that adds the audio signals amplified by the amplifying unit.

Images