KR100684029B1 - Method for generating harmonics using fourier transform and apparatus thereof, method for generating harmonics by down-sampling and apparatus thereof and method for enhancing sound and apparatus thereof - Google Patents

Abstract

A harmonic generation method using the fourier transform, an apparatus therefor, a harmonic generation method by down-sampling, an apparatus therefor, a sound correction method, and an apparatus therefor are provided to have low-complexity by using the fourier transform and sine wave generation instead of a multi-BPF(Band Pass Filter) having high fineness and be easily embodied. An FFT(Fast Fourier Transform) unit(120) performs an FFT for an inputted sound signal, interpolates a Fourier coefficient value for a certain frequency component of the sound signal by using a smoothing function, and outputs the interpolated value. A sine wave generating unit(130) generates a sine wave signal having a size and a frequency corresponding to the Fourier coefficient value. A harmonic generating unit(140) generates a harmonic signal of the sine wave signal.

Description

Method for generating harmonics using Fourier transform and apparatus therefor, method for generating harmonics by downsampling, apparatus and sound correction method for same and apparatus therefor {method for generating harmonics using fourier transform and apparatus approximately, method for generating harmonics by down- sampling and apparatus pretty and method for enhancing sound and apparatus julia}

1 is a block diagram of an apparatus for generating overtones for bass correction according to an exemplary embodiment of the present invention.

2 is a block diagram of an apparatus for generating overtones for bass correction according to another exemplary embodiment of the present invention.

3 is a block diagram of an apparatus for generating overtones for bass correction according to still another exemplary embodiment of the present invention.

4 is a block diagram of a treble correction device for generating treble according to an embodiment of the present invention.

5 is a block diagram of a sound correction apparatus according to an embodiment of the present invention.

6 is a block diagram of an HRTF unit according to an embodiment of the present invention.

7 is an input and output graph of the harmonic generating unit according to an embodiment of the present invention.

8 is a spectrum analysis diagram of a sound signal input and output to a treble correction device for generating treble.

9 is a flowchart illustrating a method for generating overtones for bass correction according to an embodiment of the present invention.

10 is a flowchart illustrating a method for generating overtones for bass correction according to another exemplary embodiment of the present invention.

11 is a flowchart illustrating a method for generating overtones for bass correction according to another embodiment of the present invention.

12 is a procedure of the treble correction method for generating harmonics according to an embodiment of the present invention.

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

100: tone correction device 110,210,310 for bass correction paging unit

120,320: FFT section 125: low pass filter

130: sine wave generation unit 135: aggregation unit

140: harmonic generation unit 150: band pass filter unit

160,270,350: HRTF part 170,360: summation part

200: harmonic generation device for treble correction 220: first band pass filter

230: down sampling unit 240: frame repeating unit

250: smoothing unit 260: second band pass filter

330: low frequency band correction unit 340: high frequency band correction unit

351: low frequency band portion 352: high frequency band portion

353: main signal unit 400: speaker

The present invention relates to a method for generating harmonics using a Fourier transform, and a device therefor, a method for generating harmonics by downsampling, an apparatus and a sound correction method for the same, and a device for the same. Method for generating harmonics using Fourier transform for correcting bass using Fourier transform and correcting treble through subsequent sampling and frame repetition instead of multipass band filter It relates to a sound correction method and an apparatus therefor.

Significant advances in signal processing technology and related devices have led to the development of various sound signal processing methods to achieve more three-dimensional sound. In particular, various low frequency and high frequency sound correction methods have been recently provided to provide listeners with a more realistic and original sound.

In many sound reproducing systems such as low-cost earphones, attenuation of low-frequency band sound signals occurs markedly compared to those of the medium and high frequency bands. In order to solve this problem, conventional bass correction apparatuses use a method of amplifying low frequency band components of a sound signal using an equalizer. However, when directly amplifying low frequency band components, attenuation of the signal must be attenuated to prevent clipping, and a dynamic range of the sound signal is reduced.

Accordingly, recently developed low and high frequency sound correction methods employ a method of generating a harmonic signal by inputting a sound signal having a fundamental frequency component into a nonlinear system. It is based on the known theory of the missing fundamental, which states that even when listening to harmonic sound signals at the fundamental frequency instead of sound signals with a fundamental frequency, The listener does not feel the difference in pitch.

However, in order to generate harmonic signals, a nonlinear system must be used. Since a nonlinear system does not follow the principle of superposition, when a sound signal having two or more frequency components is input, inter-modulation components are generated and intermodulated. Inter-modulation distortion occurs.

Conventional sound compensators filter the sound signal through a multi-band filter of fine granularity with high granularity before the sound signal is input to the nonlinear system to prevent the generation of intermodulation components. Produce signals segmented into narrowbands.

However, the conventional sound compensator having a multiband pass filter having a high detail has a high complexity and is difficult to implement.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above problems, and it is easy to implement a harmonic correction harmonic generation method having low complexity by using Fourier transform and sine wave generation instead of a multiband pass filter having a high detail and the same. It is to provide a device for.

Another object of the present invention is to provide a method for generating a harmonic correction harmonic and a device therefor, which is easy to implement with low complexity by using a downsampling and frame repetition method instead of a multiband pass filter having a high detail. There is also in the ship.

Still another object of the present invention is to provide a sound correction device for providing a three-dimensional sound by using the low tone correction harmonic generation device and high tone correction harmonic generation device.

According to an aspect of the present invention for achieving the object, the Fourier transform unit for performing a Fourier transform (fourier transform) on the input sound signal to output a Fourier coefficient value for any frequency component of the sound signal; A sine wave generator for generating a sine wave signal having a magnitude and a frequency corresponding to the Fourier coefficient value; And a harmonic generation unit generating a harmonic signal of the sinusoidal signal. 4.

According to another aspect of the present invention for achieving the above object, performing a Fourier transform on the received sound signal to calculate a Fourier coefficient value for any frequency component of the sound signal; Generating a sinusoidal signal having a magnitude and a frequency corresponding to each of the calculated Fourier coefficients; And a harmonic generation method of generating a harmonic signal of the generated sinusoidal signal. The method of generating harmonics using a Fourier transform includes a sound signal received according to another aspect of the present invention. A downsampling unit for downsampling a predetermined multiple by a predetermined frame size to generate a harmonic signal of the sound signal; And a frame repeater for interpolating the harmonic signal by performing frame repetition processing on the generated harmonic signal.

According to another aspect of the present invention for achieving the above object, the downsampling step of generating a harmonic signal of the sound signal by down-sampling the input sound signal by a preset multiple by a predetermined frame size; And a frame repetition step of interpolating the harmonic signal by performing a frame repetition process on the generated harmonic signal.

According to another aspect of the present invention for achieving the above object, a Fourier transform unit for performing a Fourier transform on the input sound signal and outputs Fourier coefficient values for the sound signal; A low frequency band correcting unit generating a sinusoidal signal having a magnitude and a frequency corresponding to a Fourier coefficient value of an arbitrary frequency component of the sound signal, and generating and outputting a low frequency correction signal that is a harmonic signal of the sinusoidal signal; High-pass filtering the received sound signal, down-sampling to a preset multiple by a preset frame size to generate a harmonic signal of the sound signal, and performing frame repetition processing on the generated harmonic signal to interpolate the harmonic signal A high frequency band correction unit for outputting a high frequency correction signal generated by using; Outputting the summed signal by performing HRTF processing on the low frequency correction signal, the high frequency correction signal, and the Fourier coefficient values for the sound signal to reflect the change of the harmonic signals generated as the sound propagates from the sound source to the listener HRTF unit; characterized in that the sound correction device including a.

According to another aspect of the present invention for achieving the above object, Fourier transform step of performing a Fourier transform on the input sound signal to output the Fourier coefficient values for the sound signal; A first low frequency band correction step of generating a sinusoidal signal having a magnitude and a frequency corresponding to a Fourier coefficient value for any frequency component of the sound signal; A second low frequency band correction step of generating and outputting a low frequency correction signal that is a harmonic signal of the sine wave signal; A first high frequency band correction step of high pass filtering the received sound signal; A second high frequency band correction step of generating a harmonic signal of the sound signal by down sampling a preset multiple by a preset frame size; A third high frequency band correction step of performing a frame repetition process on the generated harmonic signal to output a high frequency correction signal generated by interpolating the harmonic signal; Outputting the summed signal by performing HRTF processing on the low frequency correction signal, the high frequency correction signal, and the Fourier coefficient values for the sound signal to reflect the change of the harmonic signals generated as the sound propagates from the sound source to the listener HRTF processing step; characterized in that the sound correction method comprising a.

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

First, in implementing the present invention, the sound signal input to the devices of the present invention is a discrete signal in which a continuous signal is sampled by a preset frequency.

1 is a block diagram of an apparatus for generating overtones for bass correction according to an exemplary embodiment of the present invention.

Referring to FIG. 1, the apparatus for generating bass correction 100 according to an embodiment of the present invention includes a paging unit 110, a fast fourier transform unit 120, and a sine wave generator 130. ), A harmonic generating unit 140, a band pass filter unit 150, a head related transfer function (HRTF) unit 160 and a summing unit 170.

The paging unit 110 receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, and in order to reduce correlation between sound signals of each channel, the input stereo Paging the sound signal is output to the FFT unit 120 through the left and right channels.

If the correlation between the sound signals of each channel is lowered by paging the stereo sound signal, the speaker may be provided with two speakers through two speakers.

The sound signal on which the paging is performed may be represented by Equation 1 below.

L2 = L1-aR1

R2 = R1-aL1

Here, L1 and R1 are input stereo sound signals of left and right channels, respectively, L2 and R2 are output stereo sound signals of left and right channels, respectively, and a represents a paging coefficient.

In addition to the above-described paging method, another well-known paging method may be employed to implement the present invention, and the paging method is a well-known technology, and thus, the detailed description thereof will be omitted.

The FFT unit 120 receives the paging sound signal from the paging unit 110 through the left and right channels, and performs a Fourier transform on the sound signals of the respective channels, for example. Fast Fourier transform (FFT), which is a discrete fourier transform (DFT), is used to perform Fourier coefficients for any frequency component, preferably low band frequency components of the sound signal. Values are output to the sine wave generator 130.

In general, the music sound signal has a frequency band of 0 KHz to 22 KHz. In order to sample the music sound signal without aliasing, the music sound signal must be sampled at a frequency of 44 KHz or more, which is the Nyquist frequency. When the FFT is performed using 1024 samples sampled at 44 KHz, the frequency interval between the 1024 Fourier coefficients calculated by the FFT is about 40 Hz. The Fourier coefficients for the low band, eg, 0 Hz to 300 Hz band frequency components are the first seven Fourier coefficients.

The Fourier coefficients are calculated by four Fourier coefficient values, i.e., a set of Fourier coefficients, for every 1024 samples of the input sound signal, and thus a Fourier coefficient set for the remaining 1023 sound signal samples among the 1024 sound signal samples. These can be calculated or interpolated using a smoothing function. A linear smoothing function may be used as the smoothing function. For example, one set of Fourier coefficients, calculated using 1024 sound samples, and the next 1024 samples, then Fourier coefficients for the same frequency component among the four sets of Fourier coefficients, are connected in a straight line. A set of Fourier coefficients for can be obtained. However, of course, as the low frequency component of the input sound signal increases, several nonlinear smoothing functions that emphasize the low frequency component can be used to interpolate the Fourier coefficients. Amplitude modulation is performed on the sine wave functions generated by the sine wave generator 130 by the smoothing function.

The FFT unit 120 outputs the seven Fourier coefficient values for each sound signal sample to the sine wave generator 130 through seven different channels.

The FFT unit 120 may be implemented as a multi band filter having a pass band every 40 Hz and a peak value of a stop band of 13 dB. That is, the FFT unit 120 receives seven Fourier coefficients for every 1024 sound signal samples from the multi band pass filter (not shown) that performs the multi band pass filtering and the band pass filter, and the remaining 1023 sound signal samples are received. It may include a smoothing function (not shown) for interpolating the seven Fourier coefficients for the field to output to the sine wave generator 130.

The sine wave generator 130 receives the seven Fourier coefficient values from the FFT unit 120 through the seven channels, and has seven sine waves having magnitudes and frequencies corresponding to the Fourier coefficient values. Signals are generated and output to the harmonic generator 140 through seven separate channels.

The sinusoidal signal generated by the sinusoidal wave generator 130 may be represented by Equation 2 below.

Where k is the frequency index, n is the time index, N is the number of samples, a (k, n) is the real part of the k-th FFT coefficient generated by the smoothing function, and b (k, n) is generated by the smoothing function The imaginary part of the k th FFT coefficient, lf (k, n), represents the output of the sine wave generator 130 with respect to the k th frequency index at the n th time index.

The harmonic generator 140 receives the seven sinusoidal signals from the sinusoidal generator 130 through the seven channels, generates a harmonic signal for the seven sinusoidal signals, and generates the bandpass filter. The output unit 150 outputs seven independent channels.

The harmonic generator 140 may be a nonlinear system, for example, an asymmetrical soft compression system having input / output characteristics illustrated in FIG. 7. The asymmetric soft compression system generates harmonic signals in a relatively high frequency band and generates relatively less intermodulation distortion than other nonlinear systems.

The band pass filter unit 150 receives the harmonic signals through the seven channels from the harmonic generation unit 140, and preset multiples of the harmonic signals of the respective channels, for example, the first and second channels. Bandpass filtering is performed on the harmonic signals of the respective channels so that only the first and second harmonic signals are outputted to the HRTF unit 160 through seven separate channels.

The HRTF unit 160 receives the harmonic signals subjected to the bandpass filtering through the seven channels, and the harmonic signals generated by sound propagating from the sound source to the listener for the harmonic signals of the respective channels. A head related transfer function (HRTF) process reflecting the change of the data is performed and output to the adder 170.

In detail, the HRTF unit 160, as shown in FIG. 6, time for reflecting the time delay and attenuation of the signal generated as the frequency components of the harmonic signals for each channel propagated through space It consists of a delay part and a damping part. In particular, the HRTF unit 160 may have various factors that vary depending on the location of the sound source in addition to the inter-aural intensity difference (IID) or the inter-aural time difference (ITD) of the sound between the two ears of the listener. Reflect them. Since the characteristics vary according to the frequency of the signal, the HTRF unit 160 varies the delay coefficient N and the attenuation coefficients S1 and S2 according to the frequency components of the harmonic signals for each channel.

Referring to FIG. 6, the left input channel of the stereo overtone signal, that is, L1 represents a overtone signal output from a left sound output unit, for example, a left speaker, and the right input channel of the stereo overtone signal, that is, R1 represents a right side. Means a harmonic signal output from the sound output unit, for example, the right speaker. S1 denotes attenuation of a call generated when the signal propagates to the listener in a straight line at the sound output unit, and S2 denotes attenuation of a signal which occurs as the signal propagates diagonally to the sound output unit by the listener. That is, the S1 and S2 reflect the IID. The delay coefficient N also reflects the time delay that occurs when the signal propagates diagonally as compared to when it propagates straight to the listener, i.e. the ITD.

The average gain of the HRTF for the frequency band to which each harmonic signal belongs may be set to S1 and S2, and the delay time of the HRTF for the frequency band to which each harmonic signal belongs may be set to N. .

The adder 170 may add and output the harmonic signals of the HRTF-processed channels input from the HRTF 160.

In general, harmonics are generated by inputting a low pass filtered signal into a nonlinear system. Since the law of superposition does not apply to nonlinear systems, when a signal with multiple frequency components is input, an inter-modulation component is generated, thereby generating a signal having an unwanted frequency in the generated signal. This adds up to inter-modulation distortion (ITD). For example, when a signal containing a DC component is applied to a non-linear system half-wave rectifier, the total energy of the generated harmonics depends on the energy level of the DC, and in the worst case produces no harmonics. It doesn't work.

In order to minimize the generation of the intermodulation components, the signals input to the nonlinear system are segmented into narrow-band signals, and then the respective divided signals are input to the nonlinear system. To this end, a multi-band filter of fine granularity having very fine frequency characteristics can be used, but such a multi-pass filter has a high complexity and is difficult to implement. have.

The bass correction harmonic generation device 100 generates a sine wave signal using the Fourier coefficients and the FFT unit 120 performing Fourier transform, instead of using a multi-band pass filter having the fine frequency characteristics. By employing the sinusoidal wave generator 130, it has a relatively low complexity (low-complexity) and provides a harmonic generation device that is easier to implement.

The low tone correction harmonic generation device 100 may be used to generate a harmonic signal of a high frequency component, but is preferably used to generate a harmonic signal of a low frequency component. This is because the low frequency band for bass correction has a magnitude of 300 Hz, for example from 0 Hz to 300 Hz, while the high frequency band for high frequency correction has a size of 10 KHz, for example from 12 KHz to 22 KHz. When using the calibrating harmonic apparatus 100 to perform sound correction over a band having a size of 10 KHz, it is necessary to use far more Fourier coefficients than the seven Fourier coefficients, and use a separate channel for each Fourier coefficient. . Therefore, in order to implement the harmonic generating device 100 for high-pitched sound correction, a device having a considerably large arithmetic processing capability should be used.

2 is a block diagram of an apparatus for generating overtones for bass correction according to another exemplary embodiment of the present invention.

Since the bass correction harmonic generation device 100 according to the present embodiment is mostly the same as the bass correction harmonic generation device 100 of FIG. 1, the description will be made based on differences, and like reference numerals denote like elements.

Referring to Figure 2, according to another embodiment of the present invention, the low tone correction harmonic generation apparatus 100 according to the paging unit 110, FFT unit 120, sine wave generation unit 130, the aggregation unit 135, The harmonic generating unit 140, the band pass filter unit 150, the HRTF unit 160, and the summation unit 170 are included.

The FFT unit 120 receives the paging sound signal from the paging unit 110 through the left and right channels, as shown in the FFT 120 shown in FIG. 1, to the sound signal of each channel. The FFT is performed to calculate seven Fourier coefficient values for the low-band frequency components of the sound signal, and then outputs the seven Fourier coefficient values to the sine wave generator 130 through seven independent channels. .

The sine wave generator 130 outputs sine wave signals having magnitudes and frequencies corresponding to the Fourier coefficients to the aggregation unit 135 through seven independent channels.

The aggregation unit 135 receives the generated sinusoidal signals from the sinusoidal wave generator 130 through the respective channels, sums the sinusoidal signals of some channels among the respective sinusoidal signals into one signal, and transmits the signal through one channel. The harmonic generator 140 outputs the signals of the remaining channels to the harmonic generator 140 through the respective channels.

The sum of the sine wave signals outputted through two or more channels from the aggregation unit 135 is output to one channel in order to reduce the number of channels to reduce the amount of calculation of the harmonic generating device 100. However, when the number of channels is reduced and input to the harmonic generating device 100, the harmonic signals generated by the summed channels include relatively many intermodulation components, thereby causing relatively large intermodulation distortion. do.

In perceiving a sound signal, a person recognizes the sound signal by dividing it into a critical bandwidth of the ear, and in the specific frequency band, the human signal is not relatively sensitive even if there is an intermodulation component due to the generation of a harmonic signal. . Accordingly, when the harmonic signals are generated by appropriately combining the channels for each specific frequency band, the apparatus for generating a basal correction 100 for basal correction may be implemented in which a relatively small amount of calculation is performed and a listener does not feel a difference due to intermodulation distortion. .

3 is a block diagram of an apparatus for generating overtones for bass correction according to another embodiment of the present invention.

Since the bass correction harmonic generation device 100 according to the present embodiment is mostly the same as the bass correction harmonic generation device 100 of FIG. 1, the description will be made based on differences, and like reference numerals denote like elements.

Referring to Figure 3, according to another embodiment of the present invention, the low tone correction harmonic generation apparatus according to the paging unit 110, FFT unit 120, low pass filter 125, sine wave generation unit 130, The harmonic generating unit 140, the band pass filter unit 150, the HRTF unit 160, and the summation unit 170 are included.

The low pass filter 125 receives the paging sound signal from the paging unit 110 through the left and right channels, and the low low band frequency signal preset for the sound signal of each channel ( Low-pass filtering is performed so that sub-bass frequency signals are output to the harmonic generator 140 through the left and right channels. The low pass filter 125 is preferably an infinite impulse response filter.

The FFT unit 120 receives the paging sound signal from the paging unit 110 through the left and right channels, as shown in the FFT 120 shown in FIG. 1, to the sound signal of each channel. FFT is performed on the circuit to produce seven Fourier coefficient values for the low band frequency components of the sound signal. Thereafter, the FFT unit 120 outputs Fourier coefficient values for mid-bass frequency components not belonging to the low low band to the sine wave generator 130 through independent channels. do.

The sine wave generator 130 receives Fourier coefficient values from the FFT unit 120 through the independent channels, and generates sine wave signals having magnitudes and frequencies corresponding to the Fourier coefficient values, respectively. Output to the harmonic generator 140 through the.

The harmonic generation unit 140 receives a sound signal from which the low pass filtering is performed through the left and right channels from the low frequency filter unit 135 through one channel, and the frequency of the input sound signal. The harmonic signals for the components are generated and output to the band pass filter unit 150 through one channel. In addition, the harmonic generator 140 receives the sinusoidal signals through the respective channels from the sinusoidal generator 130 and generates harmonic signals for the sinusoidal signals of the respective channels, thereby generating the harmonic signals through the independent channels. Output to the pass filter unit 150.

Unlike the harmonic generation device 100 shown in FIG. 1, the bass correction harmonic generation device 100 does not independently generate harmonic signals by performing Fourier transform on frequency components belonging to a low low band. Through the input to the harmonic generator 140 to generate a harmonic signals. On the other hand, for frequency components belonging to the middle low band, a sinusoidal signal generated by using a Fourier transform is input to the harmonic generator 140 in the same manner as the harmonic generator 100 shown in FIG. 1 to generate harmonic signals. .

The human ear is known to feel rather than hear about frequency components belonging to the low low band. That is, the listener does not perceive the intermodulation distortion as sensitively as the frequency components belonging to the low low band of the sound signal. Therefore, after performing low-band filtering on the sound signal to obtain low low-band frequency components, it generates harmonic signals through one channel to compensate for bass that the listeners do not feel the difference due to the intermodulation distortion even though the amount of computation is relatively low. The overtone generating apparatus 100 may be implemented.

4 is a block diagram of a treble correction device for generating treble according to an embodiment of the present invention.

Referring to FIG. 4, the apparatus for generating treble corrections according to an exemplary embodiment of the present invention 200 includes a paging unit 210, a first band pass filter 220, and a down sampling part 230. The frame repeater 240 includes a smoothing part 250, a second band pass filter 260, and an HRTF part 270.

The paging unit 210 receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, and reduces the correlation between the sound signals of each channel. Paging the sound signal is output to the first band pass filter 220 through the left and right channels. Since the paging unit 210 performs the same functions and operations as the paging unit 110 illustrated in FIGS. 1 to 3, detailed descriptions thereof will be omitted.

The first bandpass filter 220 receives a sound signal subjected to the paging from the paging unit 210, and a frequency component of a pass band preset for the sound signal, for example, a 6 KHz to 11 KHz band. Band pass filtering is performed so that only the signals are outputted to the down sampling unit 230. The first bandpass filter 220 has values obtained by dividing the minimum and maximum frequency values of the frequency band to which the harmonic signals to be output are divided by the down sampling multiple of the down sampling unit 230 as the minimum and maximum frequencies of the pass band. Preferred and the reasons are detailed below.

The downsampling unit 230 downsamples the sound signal received from the first bandpass filter 220 by a preset frame size, for example, 1024 samples, by a preset multiple, for example, a multiple of 2 The harmonic signal of the input sound signal is generated and then output to the frame repeater 240.

When the input sound signal of the down sampling unit 230 is X (z) and the down sampled sound signal is Y (z), the down sampled sound signal may be defined by Equation 3 below.

Here, X (z) is an input sound signal, Y (z) is a down sampled sound signal, z is a z region variable, n is the number of samples, and M is a down sampling multiple.

That is, the down sampling unit 230 performs a task of sampling the first 1024 samples of every 2048 samples of the input sound signal and discarding the remaining 1024 samples. The sound signal on which the down sampling is performed includes harmonic signals having a frequency of the down sampling multiple, for example, two times higher than a frequency component included in the input sound signal.

Therefore, when the harmonic signals of 12 KHz to 22 KHz are to be generated through the treble correcting harmonic generation device 200, the first bandpass filter 220 has a pass band of 6 KHz to 11 KHz. It is preferable.

When downsampling a sound signal to generate a harmonic signal, unlike when generating a harmonic signal by applying another nonlinear system, it is possible to generate desirable harmonic signals without intermodulation distortion because no intermodulation component is generated. There is this.

The frame repeater 240 receives the down-sampled harmonic signal from the down-sampler 230 and performs frame repetition processing on the generated harmonic signal to interpolate the harmonic signal. Output to the smoothing unit 250.

That is, the frame repeater 240 fills the empty portion corresponding to the frame size generated by the downsampling of the input harmonic signal with sound signal samples belonging to the previous sampled frame. For example, the first 1024 samples sampled among every 2048 samples of the input sound signal are filled with the remaining blank portion discarded by the down sampling unit 230.

The smoothing unit 250 receives the interpolated sound signal from the frame repeater 240, and smoothes a portion of the harmonic signal in which the frame is repeated, ie, a frame edge boundary. A smoothing operation) is performed on the interpolated harmonic signal and output.

This is to allow a more natural sound signal to be output by preventing a sudden change in the signal in a portion where each frame is repeated. The smoothing unit 250 may set a value obtained by averaging the last sample of each frame and the first and second samples of the next frame as the sound signal value of the first sample of the next frame.

When performing the frame repetition process, unwanted frequency components may be generated by a frame edge effect. The frame edge effect can be prevented by using a frame whose frame size is much larger than a period of a fundamental frequency and performing the smoothing process.

The second band pass filter 260 receives the interpolated harmonic signal from the smoothing unit 250, and preset harmonic signals of multiples of the interpolated harmonic signal, for example, a first harmonic signal (first). The bandpass filtering is performed on the interpolated signal to output only the harmonic signal to the HRTF unit 270. This is to remove frequency components other than the desired harmonic signal.

When the harmonic signals of 12 KHz to 22 KHz are to be generated through the treble correction harmonic generator 200, the second band pass filter 220 preferably has a pass band of 12 KHz to 22 KHz. Do.

The HRTF unit 270 receives the harmonic signals from the second passband filter 260, and reflects the change of the harmonic signals generated as the sound propagates from the sound source to the listener with respect to the input harmonic signals. A head related transfer function (HRTF) process is performed to output the result.

The HRTF unit 270 includes a time delay unit and attenuator for reflecting time delay and attenuation of a signal generated as frequency components of the generated harmonic signals propagate through space. Since the HRTF unit 270 performs the same functions and operations as the paging unit 110 illustrated in FIGS. 1 to 3, detailed descriptions thereof will be omitted.

In general, when decompressing and playing back a music file compressed by a compression algorithm such as MP3, frequency components belonging to a high frequency band, for example, about 12 KHz or about 15 KHz or more, are lost, so that the original sound is properly reproduced. You won't be able to play it. Since there is no information on frequency components belonging to the lost high frequency band, there is no way to reproduce them as they are. However, the harmonic signals generated using the low frequency components of the reconstructed sound signal are likely to be quite similar to the frequency component signals belonging to the lost high frequency band. Therefore, when the sound signal is corrected using the treble correcting harmonic generating device 200, the sound signal faithful to the original sound can be reproduced in the high frequency band.

8 is a spectrum analysis diagram of a sound signal input and output to the treble correction device 200 for generating treble correction. First, the vertical axis represents magnitude and the horizontal axis represents frequency. Referring to FIG. 8, a harmonic signal having a frequency component of 16 KHz and 18 KHz is inputted to the treble correcting harmonic generating device 200 performing down sampling of a signal having a frequency component of 8 KHz and 9 KHz by a multiple of 2. You can see that the output. The size of the other intermodulation components is negligibly small. When a harmonic signal is generated using a nonlinear system other than the harmonic generating apparatus 200 for correcting the treble, a significant amount of intermodulation components appears at 17 KHz.

The harmonic generating device 200 is more suitable for correcting a frequency component of a high frequency band than a low frequency band frequency component of a sound signal. This is because, in the case of the low frequency signal, the amount of information included in the same time interval is much smaller than that of the high frequency signal, and the size of the frame for performing the down sampling should be much larger for the low frequency signal. For example, for a signal with a frequency of 20 Hz, the size of the frame should be greater than about 10,000 samples. If the size of the frame is too large, not only a large amount of memory is required to process and store the samples, but also an echo effect due to time delay may occur.

5 is a block diagram of a sound correction apparatus according to an embodiment of the present invention.

Referring to FIG. 5, the sound calibrating apparatus 300 according to an exemplary embodiment of the present invention may include a paging unit 310, an FFT unit 320, a low frequency band correction unit 330, and a high frequency band correction unit 340. The HRTF unit 350 and the sum unit 360 are included.

The paging unit 310 receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, and reduces the correlation between the sound signals of each channel. Paging the sound signal is output to the FFT unit 320 and the low frequency band correction unit 330 through the left and right channels. Since the paging unit 310 performs the same functions and operations as the paging unit 110 shown in FIGS. 1 to 4, detailed descriptions thereof will be omitted.

The FFT unit 320 receives the paging sound signal from the paging unit 310 through the left and right channels, performs an FFT on the sound signals of the respective channels, and performs a low band of the sound signal. Fourier coefficient values for frequency components are output to the sine wave generator 130 through independent channels, and all Fourier coefficient values calculated by the FFT are respectively output to the HRTF unit 350 through independent channels. .

The FFT unit 320 performs the same functions and operations as the FFT unit 120 shown in FIGS. 1 to 4 except for outputting all Fourier coefficient values calculated by the FFT to the HRTF unit 350. Detailed description will be omitted below.

The low frequency band correction unit 330 receives Fourier coefficient values for low band frequency components of the sound signal from the FFT unit 320 and generates and outputs a low frequency correction signal. , A sine wave generator (not shown), a harmonic generator (not shown), and a band pass filter (not shown).

In detail, the sine wave generator of the low frequency band corrector 330 generates sine wave signals having a magnitude and a frequency corresponding to the Fourier coefficients input through the respective channels, and outputs the sine wave signals to the harmonic generator through independent channels. .

The harmonic generator generates a harmonic signal of the input sinusoidal signal for each channel and outputs the harmonic signal to the band pass filter.

The band pass filter unit receives the input harmonic signals from the harmonic generation unit through the respective harmonic generators, and generates harmonics of a preset multiple of the harmonic signals of the respective channels, for example, first and second multiples. Band pass filtering is performed on the harmonic signals of the respective channels so that only the signals are output, and then output to the low frequency band unit 351 of the HRTF unit 350 through the separated channels.

The sine wave generator, the harmonic generator, and the band pass filter of the low frequency band corrector 330 include the sine wave generator 130, the harmonic generator 140, and the band pass filter 150 shown in FIGS. 1 to 3. The same function and operation as described above will be omitted.

The high frequency band corrector 340 receives the paging sound signal from the paging unit 210 and generates and outputs a high frequency correction signal. A first band pass filter (not shown) And a sampling unit (not shown), a frame repeater (not shown), a smoothing unit (not shown), and a second band pass filter (not shown).

In detail, the first band pass filter of the high frequency band corrector 340 receives the sound signal from which the paging is performed from the paging unit 310 and passes through a preset pass band, for example, the sound signal. Bandpass filtering is performed to output only frequency components in a 6 KHz to 11 KHz band and output to the down sampling unit.

The down sampling unit down-samples the input sound signal by a preset multiple by a preset frame size to generate a harmonic signal of the sound signal.

The frame repeater performs a frame repetition process on the generated harmonic signal to interpolate the harmonic signal, and outputs the interpolated harmonic signal to the second band pass filter.

The smoothing unit receives the interpolated sound signal from the frame repeater, and performs a smoothing process on the interpolated harmonic signal to smooth the frame edge boundary among the harmonic signals.

The second band pass filter receives the interpolated harmonic signal from the smoothing unit, and outputs a predetermined multiple of harmonic signals, for example, a first harmonic signal, among the interpolated harmonic signals, to the interpolated signal. Band pass filtering is performed to generate the high frequency correction signal and then outputs the high frequency correction signal to the high frequency band unit 353 of the HRTF unit 350.

The first band pass filter, the down sampler, the frame repeater, the smoother, and the second band pass filter of the high frequency band corrector 340 may include a first band pass filter 220 and a down sampler (shown in FIG. 4). 230, the frame repeater 240, the smoother 250, and the second band pass filter 260 perform the same functions and operations, and thus detailed description thereof will be omitted.

The HRTF unit 350 performs HRTF processing that reflects the change of the harmonic signals generated as the sound propagates from the sound source to the listener with respect to the Fourier coefficient values for the low frequency correction signal, the high frequency correction signal, and the sound signal. A low frequency band unit 351, a main signal unit 352, and a high frequency band unit 353 are performed.

In detail, the low frequency band unit 351 of the HRTF unit 350 receives the band pass immersed harmonic signals from the low frequency band correction unit 330 through the respective channels and performs HRTF processing. Each channel is composed of a time delay unit and attenuator for reflecting the time delay and attenuation of the signal generated as the frequency components of the harmonic signals propagate through the space. The high frequency band unit 353 receives the band pass peeled harmonic signals from the high frequency band corrector 330 through one channel to perform HRTF processing, and frequency components of the harmonic signals are propagated through the space. It consists of a time delay and attenuator for reflecting the time delay and attenuation of the signal generated according to. The low frequency band part 351 and the high frequency band part 353 perform the same functions and operations as the HRTF part 160 shown in FIGS. 1 to 3 and the HRTF part 270 shown in FIG. 4, respectively. Detailed description will be omitted.

The main signal unit 352 receives all Fourier coefficient values calculated by the FFT through independent channels, performs HRTF processing, and outputs the result to the summation unit 360. The HRTF processing unit (not shown) and the IFFT unit ( inverse fast fourier transform part (not shown).

The HRTF processing unit receives the Fourier coefficient values and performs HRTF processing on the frequency domain, and outputs the HRTF processed Fourier coefficient values to the IFFT unit.

The IFFT unit receives HRTF-processed Fourier coefficient values from the HRTF unit, performs IFFT on the input Fourier coefficient values, and adds a sound signal that is time domain data corresponding to the Fourier coefficient values. )

The adder 360 receives the HRTF-processed sound signals from the low frequency band unit 351, the main signal unit 352, and the high frequency band unit 353, sums them up, and then adds the sound signals to the sound reproducing apparatus such as a speaker 400. Output a sound signal.

The sound correction apparatus 300 performs low frequency band correction using the FFT unit 320 used for HRTF processing of the main sound signal without using a multi-band filter having a separate fine pass band for low frequency band correction. Thus, there is an advantage of low complexity and easy implementation.

9 is a flowchart illustrating a method for generating overtones for bass correction according to an embodiment of the present invention.

Referring to FIG. 9, the paging unit 110 of the bass correction apparatus 100 for bass correction receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, In order to lower the correlation between sound signals, the input stereo sound signal is phased and output to the FFT unit 120 through the left and right channels (S100).

Thereafter, the FFT unit 120 of the bass correction apparatus 100 for bass correction receives a paging sound signal from the paging unit 110 through the left and right channels, and outputs the sound signal of each channel. Fourier transform (FFT, fast fourier transform, hereinafter "FFT"), which is a discrete Fourier transform (DFT), is used for any frequency component of the sound signal, Preferably, the Fourier coefficient values for the low band frequency components are output to the sine wave generator 130 through independent channels (S110).

Subsequently, the sine wave generator 130 of the bass correction apparatus 100 for bass correction receives the Fourier coefficient values from the FFT unit 120 through the respective channels, and the magnitudes correspond to the Fourier coefficient values. Sine wave signals having a frequency and a frequency are generated and output to the harmonic generating unit 140 through separate channels, respectively (S120).

Thereafter, the aggregation unit 135 of the bass correction harmonic generation device 100 receives the generated sinusoidal signals from the sinusoidal wave generation unit 130 through the respective channels, and the sinusoidal signal of some of the sinusoidal signals. These signals are combined into one signal and output to the harmonic generator 140 through one channel, and the signals of the remaining channels are output to the harmonic generator 140 through the respective channels (S130).

Thereafter, the harmonic generating unit 140 of the bass correction harmonic generating device 100 receives the sinusoidal signals through the respective channels from the sinusoidal wave generator 130 and generates a harmonic signal for each sinusoidal signal. By outputting through the independent channel to the band pass filter unit 150 (S140).

Subsequently, the band pass filter unit 150 of the bass correction apparatus 100 for bass correction receives the harmonic signals through the respective channels from the harmonic generator 140 and sets a preset multiple of the harmonic signals of the respective channels. For example, by performing band pass filtering on the harmonic signals of the respective channels so that only first and second harmonic signals of the first and second multiples are output, the HRTF unit 160 is provided through separate channels. And outputs (S150).

Thereafter, the HRTF unit 160 of the bass correction apparatus 100 for bass correction receives the bandpass pilunged harmonic signals through the respective channels, and for the harmonic signals of the respective channels, a sound is heard in the sound source. A head related transfer function (HRTF) process is performed to reflect the change of the harmonic signals generated while propagating to the output unit 100 to output the sum unit 170 (S160).

Subsequently, the adder 170 of the bass correction harmonic generation device 100 sums the harmonic signals received through the respective channels from the HRTF unit 160 and outputs the summed harmonic signals through one channel (S170).

10 is a flowchart illustrating a method for generating overtones for bass correction according to another exemplary embodiment of the present invention.

Referring to FIG. 10, the paging unit 110 of the bass correction apparatus 100 for bass correction receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, In order to lower the correlation between sound signals, the input stereo sound signal is phased and output to the FFT unit 120 through the left and right channels (S200).

Thereafter, the low pass filter 125 of the bass correction apparatus 100 for bass correction receives a sound signal from which the paging is performed through the left and right channels, and the sound of each channel. Low-pass filtering is performed to output predetermined low sub-bass frequency signals for the signal to the FFT unit 120 through the left and right channels to the harmonic generation unit 140. Output (S210). The low pass filter 125 is preferably an infinite impulse response filter (SIR).

In addition, the FFT unit 120 of the apparatus for generating bass correction 100 receives the sound signals from which the paging unit 110 performs the paging signal from the paging unit 110 as shown in FIG. 1. The FFT is performed on the sound signal of each channel to calculate seven Fourier coefficient values for the low frequency components of the sound signal. Thereafter, the FFT unit 120 outputs Fourier coefficient values for mid-bass frequency components that do not belong to the low low band to the sine wave generator 130 through independent channels. (S220).

Subsequently, the sine wave generating unit 130 of the bass correction harmonic generation device 100 receives Fourier coefficient values from the FFT unit 120 through the independent channels, and the magnitudes corresponding to the Fourier coefficient values and Sine wave signals having a frequency are generated and output to the harmonic generator 140 through independent channels, respectively (S230).

Subsequently, the harmonic generation unit 140 of the bass correction apparatus 100 for bass correction uses one channel to output a sound signal from which the low pass filtering is performed through the left and right channels from the low frequency filter unit 135. Is received through, and generates the harmonic signals for the frequency components of the input sound signal and outputs through the one channel to the band pass filter unit 150. In addition, the harmonic generator 140 receives the sinusoidal signals through the respective channels from the sinusoidal generator 130 and generates harmonic signals for the sinusoidal signals of the respective channels, thereby generating the harmonic signals through the independent channels. Output to the pass filter unit 150 (S240).

Subsequently, the band pass filter unit 150 of the bass correction apparatus 100 for bass correction receives the harmonic signals through the respective channels from the harmonic generator 140 and sets a preset multiple of the harmonic signals of the respective channels. For example, by performing band pass filtering on the harmonic signals of the respective channels so that only first and second harmonic signals of the first and second multiples are output, the HRTF unit 160 is provided through separate channels. Output as S250.

Thereafter, the HRTF unit 160 of the bass correction harmonic generating device 100 receives the bandpass pilunged harmonic signals and generates sound when the sound is propagated from the sound source to the listener for the harmonic signals of the respective channels. A head related transfer function (HRTF) process reflecting the change of the harmonic signals is performed and output to the adder 170 (S260).

Thereafter, the summation unit 160 of the bass correction harmonic generation device 100 sums the harmonic signals received through the respective channels from the HRTF unit 160 and outputs the summed harmonic signals through one channel (S270).

11 is a flowchart illustrating a method for generating overtones for bass correction according to another embodiment of the present invention.

Referring to FIG. 11, the paging unit 210 of the treble correcting harmonic generating device 200 receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, In order to reduce correlation between sound signals, the input stereo sound signal is phased and output to the first band pass filter 220 through the left and right channels (S300).

Thereafter, the first band pass filter 220 of the treble correcting harmonic generating device 200 receives the sound signal subjected to the paging from the paging unit 210, and the frequency of the pass band preset for the sound signal. Band pass filtering is performed so that only components are output and output to the down sampling unit 230 (S310).

Thereafter, the down sampling unit 230 of the treble correcting harmonic generation device 200 receives the down sampled harmonic signal from the down sampling unit 230 and performs frame repetition processing on the generated harmonic signal. The interpolation of the harmonic signal is performed and then output to the smoothing unit 250 (S320).

Thereafter, the frame repeater 240 of the treble correcting harmonic generating device 200 fills a blank portion of the input harmonic signal corresponding to the frame size generated by the downsampling with sound signal samples belonging to the previous sampled frame. Perform the operation (S330).

Thereafter, the smoothing unit 250 of the treble correcting harmonic generating device 200 receives an interpolated sound signal from the frame repeating unit 240, and a portion of the overtone signal in which the frame is repeated, that is, a frame edge boundary ( A smoothing operation for smoothing a frame edge boundary is performed on the interpolated harmonic signal and outputs it (S340).

Thereafter, the second band pass filter 260 of the treble correcting harmonic generation device 200 receives the interpolated harmonic signal from the smoothing unit 250, and preset harmonic signals among the interpolated harmonic signals. For example, band pass filtering is performed on the interpolated signal to output only a first harmonic signal and outputs the bandpass filtering to the HRTF unit 270 (S350). This is to remove frequency components other than the desired harmonic signal.

Thereafter, the HRTF unit 270 of the treble correcting harmonic generation device 200 receives the harmonic signals from the second passband filter 260, and for the input harmonic signals, sound is output from the sound source to the listener. In operation S360, a head related transfer function (HRTF) process reflecting changes in the harmonic signals generated while propagating is performed.

12 is a procedure of the treble correction method for generating harmonics according to an embodiment of the present invention.

Referring to FIG. 12, the paging unit 310 of the sound compensator 300 receives a multi-channel sound signal, for example, a stereo sound signal through left and right channels, and outputs a sound signal of each channel. In order to reduce the correlation between the signals, the stereo sound signal is phased and output to the first band pass filter 220 through the left and right channels (S400).

Thereafter, the FFT unit 320 of the sound compensator 300 receives the paging sound signal from the paging unit 310 through the left and right channels, and FFTs the sound signals of the respective channels. And outputs the Fourier coefficient values for the low band frequency components of the sound signal to the sine wave generator 130 through independent channels, and outputs all the Fourier coefficient values calculated by the FFT through the independent channels. Output to the HRTF unit 350 (S410).

Thereafter, the low frequency band correcting unit 330 of the sound calibrating apparatus 300 receives Fourier coefficient values for the low frequency components of the sound signal through the respective channels from the FFT unit 320 and receives the low frequency correcting signal. Generate and output (S420).

In addition, the high frequency band correction unit 340 of the sound compensator 300 receives the paging sound signal from the paging unit 210 and generates and outputs a high frequency correction signal (S430).

Thereafter, the HRTF unit 350 of the sound calibrating apparatus 300 is the harmonic generated as the sound propagates from the sound source to the listener with respect to the low frequency correction signal, the high frequency correction signal, and the Fourier coefficient values for the sound signal. The HRTF process reflecting the change of the signals is performed and output to the adder 360 (S440).

Thereafter, the adder 360 of the sound compensator 300 receives the HRTF-processed sound signals from the low frequency band unit 351, the main signal unit 352, and the high frequency band unit 353, adds the sound signals, and then adds the speaker ( The sound signal is output to a sound reproducing apparatus such as 400 (S450).

The 3D sound system is considered to be the next generation sound signal reproduction system that will be followed by mono and stereo systems. The three-dimensional corrected sound signal conveys an aesthetic and more realistic feeling to the listener.

The present invention can be applied to a digital signal processor (DSP) or a microprocessor of an embedded platform for an MP3 player, a mobile communication terminal, or the like. In addition, the present invention may be implemented as a sound correction plug-in or the like for a personal computer. In addition, the present invention can be used to generate more three-dimensional processed music files.

The present invention is not limited to the embodiments described above, and various modifications and changes can be made by those skilled in the art, which are included in the spirit and scope of the present invention as defined in the appended claims.

According to the present invention as described above it is possible to provide a method for generating a harmonic correction harmonics and a device therefor that can be easily implemented with low complexity by using Fourier transform and sine wave generation instead of a multi-band pass filter having a high detail. There is.

In addition, according to the present invention as described above, instead of using a multiband pass filter having a high-definition, downsampling and frame repetition method using a low complexity complex to generate a harmonic correction method and apparatus for the same that can be easily implemented There is also an effect.

In addition, according to the present invention as described above there is an effect that can provide a sound correction device that provides a more three-dimensional sound by using the low tone correction harmonic generation device and high tone correction harmonic generation device.

Claims (66)

A Fourier transform unit performing a Fourier transform on the input sound signal to interpolate Fourier coefficient values of arbitrary frequency components of the sound signal by using a smoothing function; A sine wave generator for generating a sine wave signal having a magnitude and a frequency corresponding to the Fourier coefficient value; A harmonic generator for generating a harmonic signal of the sine wave signal; Harmonic generation apparatus using a Fourier transform, characterized in that it comprises a. The apparatus of claim 1, wherein the Fourier transform is a fast fourier transform. delete The apparatus of claim 1, wherein the harmonic generating unit is an asymmetrical soft compression system. The method according to claim 1, The input sound signal is a multi-channel sound signal, A paging unit configured to perform a paging on the input multi-channel sound signal to output a correlation to the Fourier transform unit to lower correlation between sound signals of each channel; An apparatus for generating overtones using a Fourier transform further comprising a. The apparatus for generating harmonics using a Fourier transform according to claim 5, wherein the paging is performed by Equations 1 and 2 below. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) The method according to claim 1, A filter unit configured to perform band pass filtering on the harmonic signal to output only the harmonic signals of a predetermined multiple of the generated harmonic signal; An apparatus for generating overtones using a Fourier transform further comprising a. The method according to claim 1, The Fourier transformer outputs Fourier coefficient values for a plurality of preset frequency components, And the sinusoidal wave generator outputs a sinusoidal signal having a magnitude and a frequency corresponding to each of the Fourier coefficient values. The method according to claim 8, An aggregate unit configured to sum some of the generated sinusoidal signals into one signal and output the sum; An apparatus for generating overtones using a Fourier transform further comprising a. The method according to claim 5, And a low pass filter configured to perform low pass filtering to output predetermined low sub-bass frequency signals to the sound signal input from the paging unit, and output the low pass filter to the harmonic generation unit. And the Fourier transform unit outputs Fourier coefficient values for low band frequency components not belonging to the low low band to the sine wave generator. The method according to claim 8, The low pass filter is a harmonic generation device using a Fourier transform, characterized in that the IIR filter (infinite impulse response filter). The method according to claim 1, An HRTF unit for outputting the generated harmonic signals by performing a head related transfer function (HRTF) process that reflects changes in the harmonic signals generated as a sound propagates from a sound source to a listener; An apparatus for generating overtones using a Fourier transform further comprising a. The harmonic generating device using the Fourier transform according to claim 1, wherein the frequency component is a frequency component of a low frequency band. Performing a Fourier transform on the received sound signal to interpolate Fourier coefficient values for arbitrary frequency components of the sound signal using a smoothing function; Generating a sinusoidal signal having a magnitude and a frequency corresponding to each of the calculated Fourier coefficients; A harmonic generation step of generating a harmonic signal of the generated sinusoidal signal; The harmonic generation method using a Fourier transform, characterized in that it comprises a. 15. The method of claim 14, wherein the Fourier transform is a fast Fourier transform. delete The method of claim 14, wherein the generated harmonics are generated by an asymmetrical soft compression system. The method according to claim 14, The input sound signal is a multi-channel sound signal, Before the Fourier transform step A paging step of outputting by performing phasing on the input multi-channel sound signal in order to lower correlation between sound signals of each channel; The harmonic generation method using a Fourier transform, characterized in that it further comprises. 19. The method of claim 18, wherein the paging is performed by the following equation. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) The method according to claim 14, wherein after the harmonic generation step A filtering step of performing band pass filtering on the harmonic signal such that only the harmonic signals of a predetermined multiple of the generated harmonic signal are outputted: The harmonic generation method using a Fourier transform, characterized in that it further comprises. The method according to claim 14, The Fourier transform step calculates Fourier coefficient values for a plurality of preset frequency components, The sine wave generating step outputs a sinusoidal signal having a magnitude and a frequency corresponding to each of the calculated Fourier coefficients. The method according to claim 21, After the sine wave generating step An aggregation step of adding the sine wave signals of some of the generated sine wave signals into a single signal; The harmonic generation method using a Fourier transform, characterized in that it further comprises. The method according to claim 18, A low pass filtering step of performing low pass filtering to output predetermined low sub-bass frequency signals to the sound signal inputted from the paging step and outputting the low frequency band to the harmonic generator; And performing the Fourier transform step to calculate Fourier coefficient values for mid-bass frequency components that do not belong to the low low band. The method according to claim 23, The low pass filter is a harmonic generation method using a Fourier transform, characterized in that the IIR filter (infinite impulse response filter). The method according to claim 14, wherein after the harmonic generation step An HRTF step of outputting the generated harmonic signals by performing a head related transfer function (HRTF) process that reflects changes in the harmonic signals generated as sound propagates from a sound source to a listener; The harmonic generation method using a Fourier transform, characterized in that it further comprises. 15. The method of claim 14, wherein the frequency component is a frequency component of a low frequency band. A downsampling unit for downsampling the input sound signal by a preset multiple by a preset frame size to generate a harmonic signal of the sound signal; A frame repeater which performs frame repetition processing to interpolate the generated harmonic signal by using the sound signal with respect to the generated harmonic signal; The harmonic generation device by down sampling, characterized in that it comprises a. The method of claim 27, The input sound signal is a multi-channel sound signal, A paging unit configured to perform paging on the input multi-channel sound signal to output the downsampling unit to lower correlation between sound signals of each channel; The apparatus for generating overtones by down sampling, further comprising a. The apparatus of claim 28, wherein the paging is performed by the following equation. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) The method of claim 27, A first band pass filter performing band pass filtering on the input sound signal so that only frequency components of a predetermined band are outputted to the down sampling unit; The apparatus for generating overtones by down sampling, further comprising a. The method of claim 27, A smoothing unit for outputting the interpolated harmonic signal by performing a smoothing operation for smoothing a portion of the harmonic signal in which the frame is repeated; The apparatus for generating overtones by down sampling, further comprising a. 32. The apparatus of claim 31, wherein the smoothing unit performs a smoothing process by the following equation. B) (a + b + c) / 3 Where a is the last harmonic signal sample value of an arbitrary frame, b is the first harmonic signal sample value of the frame following the arbitrary frame, and c is the second harmonic signal sample value of the frame following the arbitrary frame. The method of claim 27, A second band pass filter configured to perform band pass filtering on the interpolated signal so as to output only a predetermined multiple of the harmonic signals among the interpolated harmonic signals; The apparatus for generating overtones by down sampling, further comprising a. 29. The apparatus of claim 27, further comprising: an HRTF unit for performing the output of the interpolated harmonic signals by performing a head related transfer function (HRTF) process that reflects a change in the harmonic signals generated as a sound propagates from a sound source to a listener; The apparatus for generating overtones by down sampling, further comprising a. The apparatus of claim 27, wherein the input sound signal is a high frequency sound signal. A downsampling step of downsampling the input sound signal by a preset multiple by a preset frame size to generate a harmonic signal of the sound signal; A frame repetition step of interpolating the harmonic signal by performing frame repetition processing on the generated harmonic signal; The harmonic generation method by down sampling, characterized in that it comprises a. The method of claim 36, The input sound signal is a multi-channel sound signal, Before the downsampling step A paging step of phasing the input multi-channel sound signal to output the downsampling unit to lower correlation between sound signals of each channel; The method for generating overtones by down sampling further comprising a. The method of claim 37, wherein the paging is performed by the following equation. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) The method of claim 36, wherein prior to the downsampling step A first band pass filtering step of performing band pass filtering to output only frequency components of a predetermined band with respect to an input sound signal; The method for generating overtones by down sampling further comprising a. The method according to claim 36, wherein after the frame repeat step A smoothing process of outputting the interpolated harmonic signal by performing a smoothing operation for smoothing a portion of the harmonic signal in which the frame is repeated; The method for generating overtones by down sampling further comprising a. 37. The method of claim 36, wherein the smoothing step is performed by the following equation. B) (a + b + c) / 3 Where a is the last harmonic signal sample value of an arbitrary frame, b is the first harmonic signal sample value of the frame following the arbitrary frame, and c is the second harmonic signal sample value of the frame following the arbitrary frame. The method according to claim 36, wherein after the frame repeat step A second band pass filtering step of performing band pass filtering on the interpolated signal to output only the predetermined multiples of the harmonic signals among the interpolated harmonic signals; The method for generating overtones by down sampling further comprising a. The method according to claim 36, wherein after the frame repeat step An HRTF processing step of outputting the interpolated harmonic signals by performing a head related transfer function (HRTF) process that reflects changes in the harmonic signals generated as sound propagates from a sound source to a listener; The method for generating overtones by down sampling further comprising a. The method of claim 36, wherein the input sound signal is a high frequency sound signal. A Fourier transform unit for performing Fourier transform on the input sound signal and outputting Fourier coefficient values for the sound signal; A low frequency band correcting unit generating a sinusoidal signal having a magnitude and a frequency corresponding to a Fourier coefficient value of an arbitrary frequency component of the sound signal, and generating and outputting a low frequency correction signal that is a harmonic signal of the sinusoidal signal;  High-pass filtering the received sound signal, down-sampling to a preset multiple by a preset frame size, generates a harmonic signal of the sound signal, and performs a frame repetition process on the generated harmonic signal to view the harmonic signal. A high frequency band correction unit for outputting a high frequency correction signal generated by a simple operation; Outputting the summed signal by performing HRTF processing on the low frequency correction signal, the high frequency correction signal, and the Fourier coefficient values for the sound signal to reflect the change of the harmonic signals generated as the sound propagates from the sound source to the listener HRTF unit; Sound correction device comprising a. 46. The sound compensator of claim 45, wherein the low frequency band corrector generates a harmonic signal using an asymmetrical soft compression system. The method of claim 45, wherein the input sound signal is a multi-channel sound signal, A paging unit configured to perform a paging on the input multi-channel sound signal to output a correlation to the Fourier transform unit and the high frequency band correction unit to reduce correlation between sound signals of each channel; Sound correction device further comprises. 48. The apparatus of claim 47, wherein the paging is performed by the following formula. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) The method of claim 45, wherein the low frequency band correction unit And a band pass filtering is performed on the harmonic signal to output only the harmonic signals of a predetermined multiple of the generated harmonic signal. The method of claim 45, wherein the low frequency band correction unit And outputting Fourier coefficient values for a plurality of preset frequency components, and outputting a sine wave signal having a magnitude and a frequency corresponding to each of the Fourier coefficient values. The method of claim 45, wherein the low frequency band correction unit And sine wave signals of some of the generated sine wave signals into one signal and output the sum. The method of claim 45, wherein the high frequency band correction unit And down-sampling after performing band pass filtering so that only frequency components of a predetermined band are output to the received sound signal. The method of claim 45, wherein the high frequency band correction unit And a smoothing operation for smoothing a portion in which the frame is repeated among the harmonic signals is output to the interpolated harmonic signal. The apparatus of claim 53, wherein the high frequency band correction unit A sound correction device, characterized in that the smoothing process is performed by the following equation. B) (a + b + c) / 3 Where a is the last harmonic signal sample value of an arbitrary frame, b is the first harmonic signal sample value of the frame following the arbitrary frame, and c is the second harmonic signal sample value of the frame following the arbitrary frame. The method of claim 45, wherein the high frequency band correction unit And a band pass filtering is performed on the interpolated signal so that only the predetermined multiples of the harmonic signals are output from the interpolated harmonic signal. A Fourier transform step of performing Fourier transform on the input sound signal and outputting Fourier coefficient values for the sound signal; A first low frequency band correction step of generating a sinusoidal signal having a magnitude and a frequency corresponding to a Fourier coefficient value for any frequency component of the sound signal; A second low frequency band correction step of generating and outputting a low frequency correction signal that is a harmonic signal of the sine wave signal; A first high frequency band correction step of high pass filtering the received sound signal; A second high frequency band correction step of generating a harmonic signal of the sound signal by down sampling a preset multiple by a preset frame size;  A third high frequency band correction step of performing a frame repetition process on the generated harmonic signal to output a high frequency correction signal generated by interpolating the harmonic signal; Outputting the summed signal by performing HRTF processing on the low frequency correction signal, the high frequency correction signal, and the Fourier coefficient values for the sound signal to reflect the change of the harmonic signals generated as the sound propagates from the sound source to the listener HRTF processing step; Sound correction method comprising a. 59. The method of claim 56, wherein the second low frequency band correction step is A method of sound correction, characterized by generating a harmonic signal using an asymmetrical soft compression system. The method of claim 56, wherein The input sound signal is a multi-channel sound signal, And a paging step of outputting a phasing on the input multi-channel sound signal in order to lower correlation between sound signals of each channel before the Fourier transform step. And the input sound signal of the first high frequency band correcting step is a sound signal on which the paging is performed. 59. The method of claim 58, wherein the paging is performed by the following formula. Equation 1) L2 = L1-aR1 Equation 2) R2 = R1-aL1 (L1 and R1 are the stereo sound signals of the left and right channels respectively input, L2 and R2 are the stereo sound signals of the left and right channels respectively, and a is a paging coefficient) 57. The method of claim 56, wherein after the second low frequency band correction step A band pass filtering step of performing a band pass filtering on the harmonic signal to output only the harmonic signals of a predetermined multiple of the generated harmonic signal; Sound correction method further comprises. 59. The method of claim 56, wherein the second low frequency band correcting step outputs Fourier coefficient values for a plurality of preset frequency components, and outputs a sine wave signal having a magnitude and frequency corresponding to each of the Fourier coefficient values. Sound correction method to use. 62. The method of claim 61, wherein after the second low frequency band correction step Summing and outputting some of the generated sinusoidal signals into one signal; Sound correction method further comprises. 59. The method of claim 56, prior to the first high frequency band correction step. A first band pass filtering step of performing band pass filtering to output only the frequency components of a predetermined band with respect to the input sound signal; Sound correction method further comprises. The method of claim 56, wherein after the third high frequency band correction step A frame repetition step of outputting the interpolated harmonic signal by performing a smoothing operation for smoothing a portion of the harmonic signal in which the frame is repeated; Sound correction method further comprises. The sound correction method according to claim 64, wherein the smoothing process is performed by the following equation. B) (a + b + c) / 3 Where a is the last harmonic signal sample value of an arbitrary frame, b is the first harmonic signal sample value of the frame following the arbitrary frame, and c is the second harmonic signal sample value of the frame following the arbitrary frame. The method of claim 56, wherein after the third high frequency band correction step A second band pass filtering step of performing band pass filtering on the interpolated signal to output only the predetermined multiples of the harmonic signals among the interpolated harmonic signals; Sound correction method further comprises.

Images