Patent Translate Powered by EPO and Google Notice This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate, complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or financial decisions, should not be based on machine-translation output. DESCRIPTION JPH07226992 [0001] The present invention compensates for the absence of bass components, ie, generates harmonic components based on bass components, for lack of bass response, due to poor loudspeakers of bass response. The present invention relates to a bass component compensation method that compensates for. [0002] BACKGROUND OF THE INVENTION To date, with the increasing popularity of strong, deep bass sounds, a great deal of effort has been made to increase the power of audio player bass outputs. Conventionally, this problem is addressed in the design of the speaker and / or the design of the amplification circuit. The frequency response of the speaker generally rolls off at 12 dB / octave for frequencies below the resonance point. Thus, the bass response of the loudspeaker can be improved by appropriately selecting a number of loudspeaker design parameters and materials to make the resonant frequency as low as possible. Also, by using a linear bass boost circuit, it is possible to amplify the bass signal and thereby drive the bass power from the amplifier to the loudspeaker more strongly. At this time, since the roll-off below the resonance point occurs rapidly, considerable amplification is required. [0003] 08-05-2019 1 However, although the method as described above is effective for generating strong bass, it may not be suitable. For example, due to space constraints, aesthetic reasons, cost, etc., it becomes difficult to design a low resonant frequency speaker. Also, the inclusion of a linear bass boost circuit alleviates this problem but is not sufficient and will sacrifice amplifier head room. That is, there is a problem that clipping occurs in the output signal of the amplifier when overdriven. [0004] The present invention takes into consideration such problems of the conventional bass reproduction, and obtains a sound that makes the powerful bass feel without being limited by the space, cost and the like of the speaker and without sacrificing the characteristics of the amplifier. It is an object of the present invention to provide a bass component compensation method capable of [0005] The present invention according to claim 1 is characterized in that a low frequency band component having a low sound pressure in predetermined speaker reproduction is extracted from an input signal, and the sound by the speaker is extracted based on the extracted low frequency band component. High frequency pressure harmonics, mixing the generated harmonics with the input signal, and utilizing the psychoacoustic effects of human auditory on complex tones to reproduce the bass component due to the reproduction of the speaker It is a bass region compensation method characterized by compensating. [0006] According to the present invention of claim 4, the bass component having a low sound pressure in predetermined speaker reproduction is extracted from the input signal, and the level of the extracted bass component is adjusted to a predetermined level determined in advance. Based on the bass component, the sound pressure generated by the speaker generates a high frequency harmonic, and the level of the generated harmonic is readjusted so as to correspond to the original bass component, Bass component compensation characterized in that the bass component due to the reproduction of the speaker is compensated by mixing the readjusted harmonics with the input signal and exploiting the psychoacoustic effects of the human auditory on complex tones It is a method. [0007] The present invention is applied to any audio player, for example, to generate a harmonic that can be heard well based on the attenuated bass component, and mixing the harmonic into the original signal, The lack of bass can be compensated for by making the listener able to sense the 08-05-2019 2 pitch of these lost or weak bass components in the reproduction of the loudspeaker. [0008] Also, the invention can be applied, for example, to any audio playback device, and can be economically implemented using analog circuitry or digital signal processors. [0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below based on the drawings showing the embodiments. [0010] FIG. 1 is a diagram showing the concept of the bass range compensation method of the present invention. Here, it is assumed that the frequency response of the loudspeaker has a lower resonant frequency of f0. As shown in FIG. 1, the loudspeaker response typically rolls off rapidly at 12 dB / octave for frequencies below f 0. If one piece of music contains bass components below the 6 dB cut-off frequency, these components will be significantly attenuated (low sound pressure). For example, the component of f1 shown in FIG. 1A is hardly reproduced by this speaker. To solve this problem, a low pass filter is used to filter out all components above frequency f0 and this component is extracted. Next, the extracted components are processed (details will be described later), and as shown in FIG. 1 (b), the second, third, and the like 2f1, 3f1, and 4f1, respectively. Generate the fourth 08-05-2019 3 harmonic. These generated harmonics are added to the original music signal. Since these harmonics have a frequency exceeding f0, they can be reproduced by the speaker without much attenuation (high sound pressure). If the listener detects this mixed tone, the clear pitch is not an average of harmonic frequencies. Instead, it is the difference between successive harmonics, ie the pitch of the tone, which is equal to f1. That is, the low-frequency component of the fundamental sound is either lost or significantly attenuated (low sound pressure), but it is perceived by the listener by these psycho-physiological effects of the human auditory system. Ru. Therefore, with the psycho-psychological bass boost system (hereinafter referred to as PBBS), with a pitch of f1, it is strong, without accepting that the components of f1 would otherwise be inaudible or hardly audible. Low-range sound is created. [0011] At this time, not only the power of each harmonic related to the original fundamental sound but also the number of harmonics becomes an important point of design selection. Too much harmonics will cause irritating distortion, while too little will defeat this method. The above configuration should be determined by trial listening and will probably be different for each speaker to get the best harmony. According to the rule of thumb, the fundamental power and its generated harmonics are in descending order. In many cases, the fourth and higher harmonics can be ignored as they have less effect on the quality of the sound. Thus, the amplitude of the combined output signal should not be greater than the amplitude of the original signal. Furthermore, there must be no significant degradation of the amplifier head room. [0012] The harmonic components can be generated by applying mathematical operations to the fundamental components. By squaring the fundamental sound component, a second harmonic plus a constant term is obtained as shown in (Equation 1). Therefore, by squaring the fundamental sound and then subtracting the constant term, the pure second harmonic shown in equation (2) is obtained. [0014] (2) cos 2θ = 2 cos 2 θ−1, where θ = ωtω = the angular frequency of the fundamental sound signal, the third harmonic is obtained by subtracting the fitting amount of the fundamental sound 08-05-2019 4 from the cube of the fundamental sound by You can get it. Furthermore, the fourth harmonic can be obtained directly from the second harmonic or the fundamental sound by (Equation 4). [0015] [Equation 3] cos 3 θ = 4 cos 3 θ 3 cos θ [0016] Cos 4 θ = 2 cos 2 2 θ −1 = 8 cos 4 θ−8 cos 2 θ + 1 The present invention generates harmonics using the mathematical operations described above. 2 and 3 are block diagrams of the PBBS in the first embodiment for implementing this method. This device includes a DC remover 1, a bass region extractor 2, a normalizer 3, a second harmonic generator 4, a third harmonic generator 5, and, if necessary, a higher harmonic generator 6 etc. It consists of The functions of the respective units will be described below. a) The DC remover 1 input signal S0 often includes a DC component due to a DC offset from the front stage. In order to remove it, a DC remover 1 is required. Otherwise, the direct current component will result in undesirable by-products during processing in the subsequent stages. The DC remover 1 can only be realized by means of a high pass filter with a low 3 dB cut off frequency so as not to affect the audio signal as 10 Hz. b) Bass Region Extractor 2 The bass region extractor 2 is used to remove high frequencies from S1 and hold only the bass component that is the basic sound for generating harmonics. Since the low-frequency component in question is significantly attenuated by the speaker (low sound pressure), the low-band extractor 2 is ideally equal to the cut-off frequency of the speaker whose 3 dB cut-off frequency is to be compensated Low-pass filter. c) Normalizer 3 The normalizer 3 is used to set the amplitude of the extraction signal S2 to 1 after the extraction of the bass component. The harmonic generation process is non-linear, ie normalization is required as the amplitude of the generated harmonics is not proportional to the amplitude of the extracted signal. For example, if the amplitude of the input signal A cos θ is A = 0.5, then squaring this signal produces a second harmonic whose amplitude is proportional to A 2 = 0.25. However, if A = 0.05, then A2 = 0.0025, which results in a much smaller second harmonic compared to the fundamental sound. Normalization offers the advantage of processing the input signal independent of its original amplitude. [0017] 08-05-2019 5 The normalizer 3 comprises an amplitude extractor 8 for deriving the amplitude of the input signal, and a division function 37 for obtaining a normalized signal by dividing the input signal by the derived amplitude. In one example of implementation of the amplitude extractor 8, two input signals S2 are generated using a 90 ° phase difference network consisting of two specially designed all-pass filters 31, 32. A quadrature signal is generated, namely the Cos output S32 and the Sin output S31. The Cos output S32 draws out the Sin output S31 with a phase difference of 90 °. Another example of generating two quadrature components utilizes the Hilbert transform. By squaring these quadrature signals and then summing up, the square of the amplitude S35, ie equation (5), is obtained. [0018] Next, by applying the square root function 36, the amplitude S36 is derived. In one of the examples of calculating the square root, approximation by equation (6) is used. The calculated error is less than 1% if the range is 0.25 <x <1. [0019] If x x = 1.454895 x-1.34491 x 2 + 1. 106 812 x 3-0.536 499 x 4 + 0.1 12 12 16 x 5 + 0.2 0 75 806 x is outside the range of 0.25 to 1, it must be scaled to a value within this range. Applying this calculation to the scaled values and then multiplying the result by the square root of the scaling value gives the square root of the original value. [0020] After the amplitude is derived, the normalized signal S3 can be obtained by dividing the input signal by the derived amplitude A. The division function 37 can be approximated by a long division method to the precision that the designer considers fit. When implemented using a digital signal processor, two modifications are recommended. [0021] First, it is desirable to obtain the numerator of the division not from the output S2 of the bass 08-05-2019 6 region extractor 2 but from the output S32 of the 90 ° phase difference network. Since the denominator of the division (derived A) is affected by the delay and transient of the phase difference network, more accurate output will be obtained if the numerator is also subjected to the same processing. [0022] Second, it is desirable to limit the value of the denominator to a predetermined minimum value. If the intensity of the low frequency signal in question is very low or zero, the numerator and denominator values will be very small. Dividing the two very small values results in unpredictable results and may result in undesirable audible effects. This problem can be solved by setting the denominator to this value whenever the denominator falls below a predetermined minimum value. The experimental results show that a minimum value of 0.002 is appropriate. d) Second harmonic generator 4 The second harmonic generator 4 generates the second harmonic by squaring the normalized input signal S3 and then subtracting the DC term according to (Equation 2) . Alternatively, the DC term may be removed using a DC remover 42 as shown in FIG. By multiplying the output signal S42 by the signal S36 derived from the amplitude extractor 8 of the normalizer and a predetermined user-defined constant k2, the level of the generated second harmonic S4 has a desired ratio with respect to the input signal S0. become. e) Third harmonic generator 5 and higher harmonics generator 6 The third harmonic generator 5 generates a normalized input from the cubed signal S52 of the normalized input according to (Equation 3) By subtracting a specific amount S53, a third harmonic is generated, which level is then adjusted to the desired ratio. Similarly, the fourth harmonic can be generated based on (Equation 4). Higher harmonics can also be generated using similar equations. [0023] The final output S7 of the PBBS is obtained by adding these generated harmonics, ie the signals S4, S5 and S6 (optional), to the original input signal S0. When two or more bass components are extracted, a square, a third power, and a higher order mathematical manipulation process provide a mixed output of harmonics + these bass components. The clear pitch of this composite signal is not well formed. Nevertheless, this method is still effective as it allows the speaker to produce a reproducible output instead of accepting an almost inaudible bass range. [0024] 08-05-2019 7 A poor bass speaker can not play the same sound as that played by a good bass response speaker, but using PBBS to enhance the feel of the bass sound It will be possible. The bass sound reproduced by the PBBS has the pitch of the bass component to be compensated. [0025] By utilizing the method described above, PBBS can be easily implemented using a digital signal processor. This method makes it easy to select harmonics and adjust the signal strength of each harmonic. The generated harmonics are always proportional to the input signal level within tolerance regardless of the absolute level of the input signal. [0026] As described above, the PBBS extracts bass components of frequencies lower than the cut-off frequency of the speaker from the input signal and generates harmonics of the extracted bass components without causing much attenuation. The speaker makes it possible to reproduce harmonics of frequencies higher than the cut-off frequency of the speaker. [0027] In order to extract the bass component and generate its harmonics, the method according to the present embodiment eliminates the DC remover 1 for removing the DC component of the input signal and the component whose frequency is higher than the cut-off frequency of the speaker A bass extractor 2; a normalizer 3 for normalizing the extracted bass signal so that it is not related to the original amplitude; a second harmonic generator 4 for generating harmonics from the normalized signal; It is comprised from the 3rd harmonics generator 5, the higher harmonics generator 6 grade ¦ etc.,. [0028] The conventional way to obtain strong bass sounds is to improve with the bass response of the speaker and / or linear bass boost circuit. However, in situations where these methods do not fit, PBBS provides an alternative solution to speakers with poor low frequency response. 08-05-2019 8 In the case of PBBS, harmonics of the bass component are generated and added to the original signal. Those harmonics whose frequency exceeds the cut-off frequency of the loudspeaker are reproduced by the loudspeaker without causing significant attenuation. In addition, PBBS takes advantage of the psychoacoustic effects of human hearing to allow listeners to perceive the pitch of the missing fundamental. The PBBS concept is simple and can be implemented, for example, using a digital signal processor. The level of each harmonic is easily adjustable, and the ratio of each harmonic is constant within tolerance regardless of the level of the input signal. [0029] FIGS. 4 and 5 are block diagrams showing a PBBS for implementing the bass range compensation method according to the second embodiment of the present invention. That is, the PBBS removes the DC component from the input signal, the DC remover 1, the bass region extractor 2 for removing the high frequency component from the input signal and extracting the fundamental component for generating harmonics, and the fundamental component has a fixed value. Variable gain adjuster (abbreviated as AGC hereinafter) 103 for adjusting to, second harmonic generator 104 for producing second harmonic of fundamental sound component, and third harmonic generation for producing third harmonic of fundamental sound component , A higher order harmonic generator 106 (optional) that generates higher order harmonics if necessary, a summer 107 that adds the respective harmonics, and the level of the summed harmonics Inverse variable gain regulator (hereinafter abbreviated as VIG) 108 for readjustment, the amplifier 109 for adjusting the level of the readjusted harmonic to match the level of the input signal, and the harmonic for which the level is adjusted And a summer 100 which adds the input signal. [0030] Equations (2) and (3) above are applicable in certain cases where the amplitude of the fundamental signal is one. Equations (7) and (8) represent general mathematical equations for the signal of amplitude A. [0031] A 2 cos 2θ = 2 (A cos θ) 2 -A 2 08-05-2019 9 [0032] A3cos3θ = 4 (Acosθ) 3-3A3cosθ Here, as apparent from the amplitudes of the fundamental sound signal (several 7) and (several 8), the amplitudes of harmonics generated using the square and cubic method are It is not proportional to the amplitude of the fundamental sound of. Instead, the amplitudes of the second and third harmonics are proportional to the squared and cubed amplitudes of the fundamental sound, respectively. This condition is undesirable if the absolute level of the fundamental component varies across the musical material, as the volume of the harmonic component is not proportional to the volume of the fundamental component. In the method of the first embodiment described above, as shown in FIG. 2 and FIG. 3, the amplitude of the fundamental sound is converted to 1 by adopting the normalizer 3, and thereafter, by the equations (2) and (3) Generate harmonics. The levels of the generated harmonics are then scaled to the desired ratio to the fundamental signal. All pass filter, multiplication, division, and square root functions are utilized throughout this normalization process. This is complex and expensive when using analog circuitry. [0033] In this embodiment, AGC 103 and VIG 108 are used in place of normalizer 3. The AGC 103 adjusts the amplitude of the basic sound to a predetermined constant value before generating the harmonics according to (Equation 7) and (Equation 8). Since the amplitude A is constant, the amplitudes of the harmonics are known, so control is easy. [0034] Next, the operation of the harmonic generation method of the second embodiment will be described with reference to the drawings. [0035] First, the DC remover 1 is used to remove the DC component of the input signal S0 in order to prevent undesirable side effects during the processing of the subsequent stage. 08-05-2019 10 The DC remover 1 can be configured using a simpler high pass filter whose cut-off frequency is lower than the audible range. [0036] Next, the bass region extractor 2 removes high frequency components from the signal S1 from which direct current components have been removed, and holds only low frequency bass components that are the sources for generating harmonics. The bass range extractor 2 can be implemented using low-pass filters of matching order and cut-off frequency. [0037] Subsequently, the output S2 of the low frequency range extractor 2 is sent to the AGC 103. The AGC 103 comprises a variable gain amplifier 131 and a feedback control circuit 132. The feedback control circuit 132 detects the output S103 of the variable gain amplifier 131 to generate the output S131 (referred to as the AGC voltage) and controls the gain of the variable gain amplifier 131 so that the signal S103 spans a wide dynamic range. , A fairly constant amplitude (indicated by B in FIGS. 4 and 5). With such a constant output, the harmonics generated in the subsequent stages have known and controllable magnitudes. [0038] Next, the second harmonic generator 104 generates a second harmonic according to (Equation 7). The input S103 is squared by the multiplier 141 and then amplified by the amplifier 142 whose gain is 1 / B. Note that 1 / B is a known value. It becomes clear from the electrical characteristics of the AGC 103. For convenience, the second harmonic signal S104 is obtained by removing the DC term of the signal S142 using a DC remover 143 (a simple high pass filter). [0039] Also, the third harmonic generator 105 generates the third harmonic according to (Equation 8). The multiplier 151 multiplies the signal S103 by the output S142 of the 1 / B amplifier 142 to obtain the signal S151. Signal S 151 is subjected to level adjustment by amplifier 152 to produce 08-05-2019 11 signal S 152, and signal S 103 is subjected to level adjustment by amplifier 153 to produce signal S 153. Next, subtracting the signal S153 from the signal S152 produces a third harmonic signal S154, which is further subjected to level adjustment by the amplifier 155 (so that the gain is equal to k3). A third harmonic signal S105 adjusted to a desired ratio with respect to the signal S104 is generated. [0040] Furthermore, the higher order harmonics generator 106 generates the higher order harmonics component S106 according to the matching mathematical rules. This generator is optional as it adds to the complexity of implementation and can not add significant differences to the acoustics. [0041] Next, the summer 107 sums all the harmonic components S104, S105, and S106 to generate a composite signal S107. [0042] The amplitude of the composite signal S107, which is fairly constant due to the AGC effect, is adjusted by the VIG 108. The output S108 swings in proportion to the input signal S0. VIG 108 is controlled by AGC voltage S 131, and its gain is the inverse of the gain of AGC 103. [0043] The signal S108 is further level-adjusted by the amplifier 109 and becomes the output signal S109, and the second, third, and higher harmonics have a desired ratio with respect to the input signal S0. Thereafter, the output signal S109 and the input signal S0 are added by the summing unit 100 to generate the final output S100. 08-05-2019 12 [0044] Match AGC 103 and VIG 108 to make sure that the relationship between their gains is completely reversed. An example for realizing it is shown in FIG. The AGC 103 includes a variable gain amplifier 131 and a feedback control circuit 132. The variable gain amplifier 131 is a class A amplifier with a gain determined by the resistance of the voltage controlled resistor VCR1. VCR1 is realized by a JFET transistor. The feedback control circuit 132 detects the amplitude of the signal S103 in both the positive cycle and the negative cycle using an envelope detector (consisting of D3, D4, R4, C4 and an inverter A3). The amplifier A4 is used to generate an AGC voltage S131 to control the resistance of VCR1. VIG 108 is comprised of a variable gain amplifier whose gain is determined by VCR2. VCR2 is controlled by the same AGC voltage S131. [0045] The gains of AGC 103 and VIG 108 are obtained by the following Equations 9 and 10 respectively: [0046] AGC gain = S103 / S2 = -R2 / (R1‖Rds1) [0047] In the case of VIG gain = S108 / S107 =-(R1dsRds2) / R2 Rds1 = Rds2, that is, when the electrical characteristics of the two JFET transistors are identical, the former is the reverse of the latter. Thus, in this design, it is essential to use a well-matched JFET pair. [0048] As described above, instead of the normalizer 3 used in the method of the first embodiment, by using the AGC 103 and the VIG 108 whose gain is opposite to the gain of the AGC 103, the method of the first embodiment and Similar sound effects can be realized more easily and at lower cost, and a PBBS that is extremely advantageous for use in consumer electronics where cost is a major concern can be realized. 08-05-2019 13 [0049] Although the AGC 103 and the VIG 108 are constituted by circuits as shown in FIG. 6 in the above embodiment, the present invention is not limited to this as long as they have circuit configurations having similar functions. [0050] As is apparent from the foregoing, the present invention is to obtain a sound that gives a strong bass without feeling limited by the space and cost of the speaker and without sacrificing the characteristics of the amplifier. Has the advantage of being able to [0051] Further, according to the present invention, the level of the extracted bass component is adjusted to a predetermined level, and based on the adjusted bass component, the attenuation in reproduction by the speaker is small (the sound pressure is high). A simple configuration by generating harmonics, re-adjusting the levels of the generated harmonics so as to correspond to the original bass component, and mixing the re-adjusted harmonics with the input signal There is an advantage that a low cost bass component compensation method can be realized. 08-05-2019 14

1/--страниц