JP2011035573A - Sound signal processing apparatus and sound signal processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress the deterioration of sound by adjusting output sound pressures for each of proper frequency bands in accordance with various environments for an input sound signal. <P>SOLUTION: A sound signal processing apparatus includes: a microphone 116 which collects the sound of pink noise, and converts the collected sound of pink noise into an electric signal; a characteristic deriving part 140 which derives a first transmission frequency characteristic by analyzing a sound pressure for each of a plurality of frequency bands of the electric signal; a difference calculation part 142 which calculates each difference of sound pressures of adjacent frequency bands in a plurality of frequency bands of the first transmission frequency characteristic; a boundary deriving part 144 which extracts, in order from a larger difference, one pair or more pairs of adjacent frequency bands in the first transmission frequency characteristic to derive boundary frequencies of adjacent frequency bands; an original sound division part which divides an original sound signal into signals for each of frequency bands divided by the boundary frequencies; a sound pressure adjustment part which adjusts sound pressures on the basis of the first transmission frequency characteristic; and a signal addition part 130 which adds signals of a plurality of frequency bands. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a sound signal processing device and a sound signal processing method for adjusting the balance of frequency characteristics of sound signals.

  In recent years, opportunities to listen to sounds (voices, music) output from audio devices in various environments such as in vehicles (automobiles), stores, and homes are increasing. In such an environment, sound transmission is affected by ambient noise and the shape and material of the vehicle body (wall). Therefore, even if the sound that the user listens to is balanced in frequency characteristics at the time of output, the balance of frequency characteristics may be lost at the time of reaching the user's ear depending on the environment.

  Therefore, for example, the sound signal is fixedly divided into two frequency bands, and the loudness characteristics and the vehicle speed corresponding to the running noise of the vehicle so that the dynamic range ratio between the frequency bands is the same as that of the live performance. A technique for adjusting the sound pressure for each of the two frequency bands according to the maximum listening level (volume) is disclosed (for example, Patent Document 1).

Japanese Patent Laid-Open No. 9-232896

  However, since the technology described in Patent Document 1 uniquely predicts ambient noise from the speed of the vehicle, it cannot be used in an environment other than the inside of the vehicle that does not have a speed parameter.

  In addition to the vehicle speed, the transmission of sound in the vehicle is not limited to the vehicle speed, the shape in the vehicle, the layout of the passenger seats, the number and position of the passengers, the mounted components, the engine configuration and displacement, the opening and closing of windows, the air conditioner It is affected by various factors such as the usage status and temperature. Since the frequency characteristics related to sound transmission (transmission frequency characteristics) fluctuate depending on the degree of influence of various factors, the error between a predetermined fixed transmission frequency characteristic and the actual transmission frequency characteristic may increase. There is.

  Furthermore, in the above-described prior art, the frequency that becomes the boundary for dividing the frequency band in the sound signal (hereinafter referred to as the boundary frequency) is fixed at, for example, 200 Hz, so that the actual transfer frequency characteristic changes. Regardless, uniform sound pressure adjustment is performed for each divided frequency band (hereinafter referred to as a divided frequency band). However, in practice, the position and number of boundary frequencies vary depending on the external environment, and the transmission frequency characteristic has a band that requires sound pressure adjustment and a band that does not require sound pressure adjustment. Therefore, the conventional uniform sound pressure adjustment performed in the fixed divided frequency band may break the balance of the frequency characteristics in the sound perceived by the user.

  In view of such a problem, the present invention adjusts the output sound pressure (sound pressure to be output) for each frequency band appropriately divided according to various environments with respect to the input sound signal. An object of the present invention is to provide a sound signal processing device and a sound signal processing method capable of suppressing deterioration.

  In order to solve the above problems, the sound signal processing device (100) of the present invention collects a sound having a known frequency characteristic output from a sound source at a predetermined listening point and converts the sound into an electric signal. A sound part (116), a characteristic deriving part (140) for deriving a first transfer frequency characteristic by analyzing sound pressures for a plurality of frequency bands of an electric signal, and a first transfer frequency characteristic in a plurality of frequency bands One or more sets of adjacent frequency bands in the first transfer frequency characteristic are extracted in descending order of the difference calculated by the difference calculation unit (142) that calculates the difference in sound pressure between adjacent frequency bands. A boundary deriving unit (144) for deriving a boundary frequency that is a boundary frequency between the extracted adjacent frequency bands, and an original sound that divides the input original sound signal into signals for each frequency band delimited by the boundary frequency A sound pressure adjusting unit that adjusts the sound pressure for each of signals in a plurality of frequency bands divided by the original sound dividing unit based on the first transmission frequency characteristic, and a plurality of frequency bands adjusted by the sound pressure adjusting unit And a signal adder (130) for adding signals and generating one sound signal.

  Here, the boundary derivation unit and the sound pressure adjustment unit perform the derivation of the appropriate boundary frequency and the signal adjustment based on the transmission frequency characteristic obtained by measuring the sound of the known frequency characteristic output from the sound source, The sound signal processing apparatus of the present invention can be effectively used in various environments regardless of the environment in which it is used. In the present invention, the boundary frequency is dynamically determined according to the transmission frequency characteristic, not a fixed value, so that the sound pressure of the transmission frequency characteristic approximates the frequency band divided by the boundary frequency. This is a summary of the current bandwidth. Therefore, sound pressure adjustment can be appropriately performed for each of a plurality of frequency bands divided by the boundary frequency, and output sound pressure for each frequency band can be adjusted to suppress sound deterioration.

  The characteristic deriving unit has a plurality of adjacent frequency bands of the first transfer frequency characteristic extracted by the boundary deriving unit, and a plurality of frequencies different from the plurality of frequency bands of the first transfer frequency characteristic in the electric signal. The second transfer frequency characteristic is derived by analyzing the sound pressure for each band, and the difference calculation unit calculates the difference between the sound pressures of adjacent frequency bands in the plurality of frequency bands of the second transfer frequency characteristic, and the boundary The derivation unit is configured so that the difference calculated by the difference calculation unit is the highest among the adjacent frequency bands in the plurality of frequency bands of the second transfer frequency characteristic for each of one or more adjacent frequency bands of the first transfer frequency characteristic. A large set may be extracted, and the extracted boundary frequency between adjacent frequency bands may be used as the boundary frequency.

  If the bandwidth for each of the plurality of frequency bands of the first transfer frequency characteristic is too wide, fine fluctuations are not properly reflected, and if it is too narrow, the processing load increases and the boundary frequency may be locally biased. There is. Here, first, the first transmission frequency characteristic divided into a plurality of frequency bands is derived, the adjacent frequency bands that are candidates for the boundary frequency are narrowed down, and only the narrowed adjacent frequency bands are described in detail. A second transfer frequency characteristic divided into a plurality of frequency bands different from the plurality of frequency bands of the one transfer frequency characteristic is derived. With this configuration, the boundary frequency can be appropriately determined without being locally biased by simple processing.

  The above characteristic derivation unit derives the first transmission frequency characteristic by performing one octave analysis, and ½ octave analysis, ３ octave analysis, ¼ octave analysis, ６ octave analysis, and オ ク タ octave analysis The second transfer frequency characteristic may be derived using either analysis or 1/24 octave analysis.

  With the configuration using the octave band in which the ratio between the upper limit frequency and the lower limit frequency of the frequency band is constant, the octave band analysis which is a fixed ratio analysis can be performed, and an analysis close to human hearing is possible. Further, the second transfer frequency characteristic is analyzed and derived in a band divided by a divisor of 24, which is twice the number of pitches included in one octave band, so that the band corresponding to the twelve scales included in one octave is obtained. You can make adjustments.

  The boundary deriving unit may not extract the frequency band of the adjacent first transmission frequency characteristics in which the absolute value of the difference in sound pressure is smaller than a predetermined threshold.

  With such a configuration, it is possible to avoid a situation where the absolute value of the difference in sound pressure is small and the frequency band that does not need to be adjusted is wasted, and the processing load can be reduced.

  The sound pressure adjusting unit may adjust each signal of a plurality of frequency bands so as to reduce the difference. With this configuration, the original sound signal can be adjusted so that the frequency band in which sound is difficult to be transmitted due to the influence of the surrounding environment and the frequency band in which sound is easy to be transmitted can be heard at the same level, thereby suppressing sound degradation. It becomes possible to do.

  The sound signal processing device generates a harmonic signal having a frequency component in a predetermined frequency band from the original sound signal so that the output sound pressure decreases as the order increases, and adds the generated harmonic signal to the original sound signal. It may further include a low-frequency range overtone generator.

  When the original sound signal is compressed by a predetermined audio compression method, the amount of information in the low sound range may be reduced. Even when the speaker is not compressed, when a small speaker is used, the output capability of a low-frequency sound signal is insufficient, and a sufficiently low-frequency sound is not output. In such a case, it is possible to add a harmonic overtone signal of an original sound signal in the low sound range to the original sound signal to generate a so-called missing fundamental, thereby emphasizing a low sound range component that is a fundamental sound of the overtone. However, if higher harmonics are added while maintaining the sound pressure, the output sound may be different from the original sound. Therefore, in the sound signal processing device of the present invention, the higher the order of the harmonic signal of the low-frequency component of the original sound signal, the lower the output sound pressure is added to the original sound signal. With this configuration, the added harmonics are hardly perceived in the middle and high frequencies, and the voice is not disturbed. In the low frequencies, the missing fundamentals can be emphasized by causing missing fundamentals.

  The low-frequency range overtone generating unit may include a high-pass filter (170).

  Unnecessary DC components may be generated in the process of generating a harmonic signal from the original sound signal. With the configuration using such a high-pass filter, it is possible to remove the DC component sound signal generated when the harmonic signal is generated and add only the harmonic component to the original sound signal.

  The sound signal processing apparatus further includes a high-frequency overtone generation unit (114) that generates a harmonic signal of a predetermined order of frequency components in a predetermined frequency band from the original sound signal, and adds the generated harmonic signal to the original sound signal. May be.

  The original sound signal may lack a high sound range by an audio compression method such as MP3 (MPEG Audio Layer-3), AAC (Advanced Audio Coding), Ogg Vorbis, or the like. In the present invention, the missing high sound region can be reproduced by adding the harmonics of a predetermined order of the high sound region component remaining in the original sound signal.

  The high-frequency range overtone generation unit may include a high-pass filter (184).

  Unnecessary DC components may be generated in the process of generating a harmonic signal from the original sound signal. With the configuration using such a high-pass filter, it is possible to remove the direct current component generated when the harmonic signal is generated and add only the harmonic component to the original sound signal.

  When the original sound signal is subjected to compression processing, the cut-off frequency of the high-pass filter may be the lower limit frequency of the frequency band cut off in the compression processing.

  Even in the high sound range of, for example, 16 kHz or higher that was discarded during compression, up to half the sampling frequency at the time of generating sound data, for example, up to 22 kHz is an effective region that affects sound quality. Therefore, the high-frequency range overtone generation unit generates secondary overtones from the original sound signal of 8 kHz to 11 kHz, which is approximately half of the missing 16 kHz to 22 kHz, and reproduces the missing high frequency range 16 kHz to 22 kHz. In this case, by setting the cut-off frequency to the lower limit frequency 16 kHz of the generated overtone, the DC component generated at the time of overtone generation can be reliably removed without affecting the generated overtone.

  In order to solve the above problems, the sound signal processing method of the present invention collects a sound having a known frequency characteristic output from a sound source at a predetermined listening point, converts the sound into an electric signal, The first transfer frequency characteristic is derived by analyzing the sound pressure for each of a plurality of frequency bands, and the difference between the sound pressures of adjacent frequency bands in the plurality of frequency bands of the first transfer frequency characteristic is calculated, and the calculated difference In order from the largest, one or more adjacent frequency bands in the first transfer frequency characteristic are extracted, a boundary frequency that is a boundary frequency between the extracted adjacent frequency bands is derived, and the input original sound signal is converted into the boundary frequency. Is divided into signals for each frequency band divided by, and based on the first transfer frequency characteristics, the sound pressure is adjusted for each of the divided multiple frequency band signals, and the adjusted multiple frequency band signals are added, And generating a sound signal.

  The above-described components based on the technical idea of the sound signal processing device and the description thereof can also be applied to the sound signal processing method.

  The sound signal processing apparatus of the present invention can suppress deterioration of sound by adjusting the output sound pressure for each appropriate frequency band according to various environments with respect to the input sound signal.

It is the functional block diagram which showed the electrical structure of the sound signal processing apparatus concerning this embodiment. It is the functional block diagram which showed the electrical structure of the low-tone overtone production | generation part. It is explanatory drawing which shows an equal loudness curve. It is the functional block diagram which showed the electrical structure of the high-range overtone production | generation part. It is explanatory drawing explaining the original sound signal with which the harmonic signal was added. It is explanatory drawing explaining the multiplication function concerning this embodiment. It is explanatory drawing which shows the sound pressure for every 1/3 octave band of the pink noise at the time of output. It is explanatory drawing which shows an example of the derivation | leading-out result of the transfer frequency characteristic using pink noise. It is explanatory drawing which shows an example of the result of the octave band analysis concerning this embodiment. It is explanatory drawing which shows an example of the result of the 1/3 octave band analysis concerning this embodiment. It is the flowchart which showed the specific process of the sound signal processing method concerning this embodiment.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The dimensions, materials, and other specific numerical values shown in the embodiment are merely examples for facilitating understanding of the invention, and do not limit the present invention unless otherwise specified. In the present specification and drawings, elements having substantially the same function and configuration are denoted by the same reference numerals, and redundant description is omitted, and elements not directly related to the present invention are not illustrated. To do.

  In general, sound transmission is affected by ambient noise and the shape and material of the vehicle body (wall). Therefore, the sound that the user listens to may be out of balance in frequency characteristics when it reaches the user's ear. It was.

  An object of the present embodiment is to suppress the deterioration of sound by adjusting the output sound pressure for each frequency band appropriately divided according to various environments with respect to the sound signal. Hereinafter, the sound signal processing apparatus 100 capable of achieving such an object will be described, and then a series of operations of the sound signal processing method will be described. Note that the sound signal processing apparatus 100 that generates harmonics of the low sound range and the high sound range and can emphasize the low sound range and the high sound range will be described as an example.

(Sound signal processing apparatus 100)
FIG. 1 is a functional block diagram showing an electrical configuration of the sound signal processing apparatus 100 according to the present embodiment. The sound signal processing apparatus 100 includes a signal input unit 110, a low-frequency overtone generation unit 112, a high-frequency overtone generation unit 114, a microphone 116 as a sound collection unit, a sound pressure detection unit 118, and a central control unit 120. LPF (low pass filter) 122, BPF (band pass filter) 124 (BPF 124a, BPF 124b, BPF 124c, BPF 124d), HPF (high pass filter) 126, compressor 128 (compressor 128a, compressor 128b, compressor 128c, compressor 128d, A compressor 128e, a compressor 128f), an adder 130 as a signal adding unit, and a signal output unit 132. In the present embodiment, the LPF 122, the BPF 124, and the HPF 126 function as an original sound dividing unit. Further, the compressor 128 and a multiplication setting unit 146 described later function in cooperation with each other as a sound pressure adjusting unit.

  An original sound (audio) signal is input to the signal input unit 110 from an audio device such as a CD player, a DVD player, or an MP3 player (not shown) formed separately or integrally.

  The low-tone overtone generating unit 112 generates a harmonic signal having a predetermined frequency band, for example, a frequency component of 0 to 300 Hz, out of the original sound signal input to the signal input unit 110, and adds it to the input original sound signal.

  FIG. 2 is a functional block diagram showing an electrical configuration of the low-frequency range overtone generating unit 112. The low-frequency overtone generating unit 112 includes, for example, an LPF 160, a multiplier 162, a first adder 164, a first delay 166, an amplifier 168, an HPF (high pass filter) 170, and a second delay 172. , And a second adder 174.

  The LPF 160 passes a frequency component having a frequency equal to or lower than a predetermined frequency (in the present embodiment, 300 Hz that is a target of overtone generation) from the original sound signal transmitted from the signal input unit 110.

  In order to achieve the so-called missing fundamental effect, the user perceives the 5th to 6th harmonics (the sound pressure decreases as the order increases) up to a frequency component in the vicinity of 2 kHz where the minimum audible sound field is 0 dB. You can do it. Accordingly, since 2 kHz divided by 7 is approximately 289 Hz, in order to keep the 7th harmonic over 2 kHz or less, the frequency component of the harmonic generation target is set to about 300 Hz or lower. Here, missing fundamental refers to a phenomenon in which the pitch of the fundamental frequency is perceived as a fundamental tone if there is a continuous harmonic of the fundamental frequency even if there is no fundamental frequency.

  The multiplier 162 multiplies the original sound signal transmitted from the LPF 160 and the original sound signal transmitted from the amplifier 168 described later to generate a harmonic sound signal of the original sound signal.

  The first adder 164 adds the original sound signal transmitted from the LPF 160 and the harmonic sound signal transmitted from the multiplier 162.

  The first delay unit 166 delays the sound signal including the original sound signal and the harmonic signal transmitted from the first adder 164 by a predetermined sampling period.

  The amplifier 168 multiplies the sound signal transmitted from the first delay unit 166 by a predetermined gain of 1 or less. In this way, by repeating the processing from the multiplier 162 to the amplifier 168, a harmonic signal having an order of 2 or more is generated. Furthermore, when the amplifier 168 multiplies the sound signal by a gain of 1 or less, the sound pressure of the harmonic signal can be lowered as the higher-order harmonic signal is obtained.

  The amplifier 168 performs control so that the sound pressure of the frequency component of the original sound signal becomes low in the vicinity of 2 kHz where the minimum audible sound field is 0 dB as shown in FIG.

  The HPF 170 blocks a cutoff frequency, that is, a frequency component smaller than 50 Hz in the present embodiment, from the sound signal transmitted from the first adder 164.

  When a harmonic signal is generated from the original sound signal, an unnecessary DC component may be generated. However, with the configuration using the HPF 170, the DC component generated when generating the harmonic signal can be removed, and only the original sound signal and the harmonic signal can be added to the original sound signal. .

  The second delay unit 172 delays the original sound signal transmitted from the signal input unit 110 by a predetermined sampling period. The phase of the sound signal including the original sound signal and the harmonic signal transmitted from the HPF 170 and the original sound signal transmitted from the second delay device 172 at the time of addition in the second adder 174 by the second delay device 172. Can be combined.

  The second adder 174 adds the sound signal including the original sound signal and the harmonic signal transmitted from the HPF 170 and the original sound signal transmitted from the second delay 172.

  For example, when the input original sound signal is compressed by a predetermined audio compression method, the amount of information in the low sound range may be reduced. Even when the speaker is not compressed, when a small speaker is used, the output capability in the low range is insufficient, and the sound in the low range is not sufficiently output. In such a case, the above-described low-frequency overtone generation unit 112 adds the low-frequency original sound signal and the overtone signal to generate a so-called missing fundamental, and emphasizes the low-frequency component that is the fundamental tone of the overtone. Can do.

  However, if higher harmonic signals are added while maintaining the sound pressure, the output sound may be different from the input original sound. Hereinafter, this phenomenon will be described using an equal loudness curve.

  FIG. 3 is an explanatory diagram showing an equal loudness curve. The equal loudness curve decreases downward from 0 Hz to around 3 kHz. This indicates that even a sound having the same sound pressure can be heard by a user with a larger frequency. In the present embodiment, as shown in FIG. 3, a harmonic signal is generated for the original sound signal in the frequency band indicated by the double arrow 200.

  Due to the psychoacoustic characteristics indicated by this equal loudness curve, when a harmonic signal based on an original sound signal of 300 Hz or less is generated while maintaining the sound pressure, the harmonic signal is larger than the basic sound even if the fundamental sound and the sound pressure are the same. As a result, the output sound is disturbed.

  Therefore, in the sound signal processing apparatus 100 of the present embodiment, the higher the order of the harmonic signal of the low-frequency component of the input original sound signal, the lower the output sound pressure is added to the original sound signal. With this configuration, the added harmonics are hardly perceived in the middle and high frequencies, and the voice is not disturbed. In the low frequencies, the missing fundamentals can be emphasized by causing missing fundamentals.

  The high-frequency overtone generation unit 114 generates a harmonic signal having a predetermined order of frequency components in a predetermined frequency band of the input original sound signal, up to the second order in the present embodiment, and adds the harmonic signal to the input original sound signal. To do.

  FIG. 4 is a functional block diagram showing an electrical configuration of the high-frequency range overtone generating unit 114. The high tone overtone generation unit 114 includes a BPF 180, a multiplier 182, an HPF (High Pass Filter) 184, a delay unit 186, and an adder 188.

  The BPF 180 passes only a frequency component of a predetermined lower limit frequency, which is higher than 8 kHz in the present embodiment, and a predetermined upper limit frequency, which is 11 kHz or less in the present embodiment, of the original sound signal transmitted from the low-frequency overtone generation unit 112. Here, the reason why the BPF 180 is set to 8 to 11 kHz will be described in detail later.

  The multiplier 182 multiplies the original sound signal transmitted from the BPF 180 by the original sound signal itself transmitted from the BPF 180 to generate a secondary harmonic signal.

  The HPF 184 blocks a frequency component lower than the cutoff frequency from the original sound signal transmitted from the multiplier 182.

  When an overtone signal is generated from the original sound signal, an unnecessary direct current component may be generated. However, the configuration using the HPF 184 may remove the direct current component generated at the time of generating the overtone signal and add only the overtone component to the original sound signal. it can.

  Further, the input original sound signal may lack a high sound range by an audio compression method such as MP3, AAC, Ogg Vorbis, etc., for example. When the MP3 audio compression method is used as the original sound signal of the present embodiment, the cutoff frequency of the HPF 184 may be the lower limit frequency of the missing frequency band, that is, 16 kHz.

  For example, a high frequency range of 16 kHz or more is an effective region that affects the sound quality up to half the sampling frequency at the time of generation of audio data, for example, up to 22 kHz. Should be composed of low to mid-range harmonics. Accordingly, the high-frequency overtone generation unit 114 generates a secondary overtone signal from the original sound signal of 8 kHz to 11 kHz (double arrow 202 shown in FIG. 3) corresponding to the half of the missing 16 kHz to 22 kHz (divided by 2) frequency. Then, the missing high frequency range 16 kHz to 22 kHz is reproduced.

  In this way, aliasing noise does not occur by generating a harmonic signal of a predetermined order for frequency components having a frequency obtained by dividing the lower limit frequency and the upper limit frequency of the missing frequency band by the predetermined order as the lower limit frequency and the upper limit frequency, respectively. The missing high frequency range can be reproduced without affecting the input original sound signal.

  Further, by setting the cutoff frequency of the HPF 184 to the lower limit frequency 16 kHz of the generated harmonic signal, the DC component generated when the harmonic signal is generated can be reliably removed without affecting the generated harmonic signal.

  The delay unit 186 delays the original sound signal transmitted from the low-frequency range overtone generating unit 112 by a predetermined sampling period. With this delay device 186, the phase of the harmonic signal transmitted from the HPF 184 and the original sound signal transmitted from the delay device 186 at the time of addition by the adder 188 can be matched.

  The adder 188 adds the overtone signal transmitted from the HPF 184 and the original sound signal transmitted from the delay unit 186 at the time of addition by the adder 188.

  FIG. 5 is an explanatory diagram for explaining an original sound signal to which a harmonic signal is added. In particular, FIG. 5B shows the frequency spectrum of the original sound signal after the low tone overtone generation unit 112 and the high tone overtone generation unit 114 add the overtone signal, and FIG. 5A shows the addition of the overtone signal. It is a frequency spectrum of the previous original sound signal.

  As shown in FIG. 5A, the original sound signal before adding the harmonic signal does not include the missing frequency component of 16 kHz or more indicated by the double arrow 210.

  When the high-frequency harmonic overtone generation unit 114 of the present embodiment generates and adds a secondary harmonic signal having a frequency component of 8 to 11 kHz, as shown in FIG. included. In general, since the high frequency range is composed of harmonics in the low to mid range, in this case, by adding the harmonic signal of a predetermined order of the high frequency component remaining in the original sound signal, the high frequency range that has been lost during data compression is added to some extent. It can be reproduced, and the sound output based on the original sound signal can be improved.

  Further, even if the low-frequency overtone generation unit 112 generates and adds a harmonic signal having a frequency component of 300 Hz or less, a change in the frequency component of 16 kHz or less can be seen as compared with FIGS. Is not often seen. In this way, the harmonic overtone signal generated by the low-frequency overtone generator 112 has a sound pressure that does not significantly disturb the original sound signal even when added to the original sound signal, but the missing fundamental is surely generated, and the harmonic fundamental and Can be emphasized bass.

  Next, returning to FIG. 1, the microphone 116 will be described. The microphone 116 collects a sound having a known frequency characteristic output from the sound source at a predetermined listening point. In the present embodiment, a sound having a known frequency characteristic is pink noise output from an external speaker based on a pink noise signal transmitted by a signal output unit 132 described later, and a predetermined listening point is a position where the microphone 116 is installed. .

  The microphone 116 converts the collected pink noise into an adjustment sound signal (hereinafter simply referred to as a measurement sound signal) that is an electrical signal. The position where the microphone 116 is installed is, for example, a headrest of a driver's seat in a car, a substantial center between seats, or a substantial center of a store in a store. It may be a position (or a substantially central position) where you want to hear sound with sound quality. Further, by using a microphone that can be attached to the ear and collecting pink noise while the user is wearing the ear, more accurate pink noise that reaches the user's ear can be collected.

  FIG. 7 is an explanatory diagram showing the sound pressure for each 1/3 octave band of pink noise during output. Pink noise is a sound in which the frequency of an arbitrary frequency component is inversely proportional to the sound pressure. Therefore, as shown in FIG. 7, the energy for each 1/3 octave band in which the ratio between the upper limit frequency and the lower limit frequency of the band is constant. Becomes uniform. This characteristic is the same for the octave band. The pink noise can be suitably used for an octave band analysis and a 1/3 octave band analysis, which will be described later, for deriving a transfer frequency characteristic (a first transfer frequency characteristic and a second transfer frequency characteristic, which will be described later).

  The sound pressure detection unit 118 detects the sound pressure at the listening point, that is, the installation position of the microphone 116. Specifically, the sound pressure detection unit 118 detects the sound pressure for each octave band and 1/3 octave band for the measurement sound signal obtained by converting the pink noise by the microphone 116 in accordance with an instruction from the characteristic deriving unit 140 described later.

  The central control unit 120 includes a semiconductor integrated circuit including a central processing unit (CPU) and a signal processing unit (DSP: Digital Signal Processor), and manages and controls the entire sound signal processing unit 100 using a predetermined program.

  The central control unit 120 also functions as a characteristic deriving unit 140, a difference calculating unit 142, a boundary deriving unit 144, and a multiplication setting unit 146.

  The characteristic deriving unit 140 derives the first transfer frequency characteristic by analyzing the sound pressure for each of a plurality of frequency bands of the measurement sound signal. At this time, the characteristic deriving unit 140 causes the sound pressure detecting unit 118 described above to detect the sound pressure of the sound collected by the microphone 116 for each octave band and 1/3 octave band.

  In the present embodiment, the characteristic value of the transfer frequency characteristic in which sound is transmitted around is the ratio of the sound pressure of the sound when output from a predetermined sound source to the sound pressure of the sound when measured at the listening point. By comparing the ratio of the sound pressures for each frequency band, it is possible to grasp the ease of sound transmission (transmission frequency characteristic) for each frequency band. By using pink noise having the same sound pressure for each octave band and 1/3 octave band for derivation of the transmission frequency characteristic, the detected sound pressure (sound pressure level) can be used as a characteristic value as it is.

  FIG. 8 is an explanatory diagram illustrating an example of a derivation result of a transfer frequency characteristic using pink noise. In FIG. 8, unlike the present embodiment, when the first transmission frequency characteristic is derived by analyzing the sound pressure for each of the plurality of frequency bands of the measurement sound signal, the 1/3 octave band is used as the plurality of frequency bands. The results derived from the 1/3 octave band analysis. Such 1/3 octave band analysis and later-described octave band analysis are performed based on JIS C1513. 8 to 10, the vertical axis represents the sound pressure level, and the horizontal axis represents the nominal center frequency of the octave band or 1/3 octave band.

  Compared to the fact that the sound pressure level (characteristic value) for each 1/3 octave band of pink noise at the time of output shown in FIG. 7 is uniform, it is derived by detecting the sound pressure of the measured sound signal as shown in FIG. In the transfer frequency characteristic from the predetermined sound source to the predetermined listening point, the sound pressure level for each 1/3 octave band varies greatly due to the influence of the surrounding environment.

  If the original sound signal is output as it is without adjusting the output sound pressure in such an environment showing the transfer frequency characteristic, the balance of the frequency characteristic will be lost during transmission from the predetermined sound source to the predetermined listening point. End up. Therefore, if the sound signal processing apparatus 100 of the present embodiment is used as described later, it is possible to avoid such an imbalance of frequency characteristics.

  The difference calculation unit 142 calculates the difference in sound pressure between adjacent frequency bands in the plurality of frequency bands of the first transfer frequency characteristic.

  The boundary deriving unit 144 has a plurality of adjacent frequency bands (octave bands in the present embodiment) of the first transfer frequency characteristic for each of a plurality of frequency bands (in the present embodiment, 1 / Among the adjacent frequency bands in (3 octave bands), one set having the largest difference calculated by the difference calculating unit 142 is extracted, and the boundary frequency of the extracted adjacent frequency bands is set as the boundary frequency, and based on the boundary frequency. The LPF 122, the BPFs 124a to 124d, and the HPF 126 are controlled.

  Here, in the derivation of the first transmission frequency characteristic, if the bandwidths of the plurality of frequency bands of the first transmission frequency characteristic are too wide as in the octave band, fine fluctuations are not properly reflected, as shown in FIG. In addition, if it is too narrow as in the 1/3 octave band, the processing load increases and the boundary frequency may be locally biased.

  Therefore, the characteristic deriving unit 140 extracts a plurality of adjacent frequency bands of the first transfer frequency characteristic extracted by the boundary deriving unit 144 from the plurality of frequency bands of the first transfer frequency characteristic in the measurement sound signal. The second transmission frequency characteristic is derived by analyzing the sound pressure for each of a plurality of different frequency bands, in this embodiment, every 1/3 octave band.

  Specifically, the characteristic deriving unit 140 first analyzes the sound pressure for each octave band in the present embodiment based on the frequency characteristics of the sound collected by the microphone 116, in the present embodiment, A first transfer frequency characteristic is derived.

  And if the difference calculation part 142 calculates the difference of the sound pressure of the adjacent frequency band in the several frequency band (octave band) of a 1st transmission frequency characteristic, respectively, the characteristic calculation part 142 will calculate the difference calculation part 142 One or a plurality of adjacent octave bands of the first transmission frequency characteristic are extracted in the order of increasing difference, in the present embodiment, five.

  Subsequently, the characteristic deriving unit 140 analyzes a plurality of frequency bands different from the plurality of frequency bands of the first transmission frequency characteristic in the measurement sound signal, in this embodiment, the sound pressure for each 1/3 octave band. A second transfer frequency characteristic is derived. Both the first transmission frequency characteristic and the second transmission frequency characteristic are transmission frequency characteristics, and the rules for dividing the frequency band for deriving the characteristic value (sound pressure) described above are different.

  And the difference calculation part 142 calculates the difference of the sound pressure of the adjacent frequency band in the frequency band (1/3 octave band) of a 2nd transmission frequency characteristic, respectively.

  The boundary deriving unit 144 has adjacent frequencies in a plurality of frequency bands (1/3 octave band) of the second transfer frequency characteristic for each one or a plurality of adjacent frequency bands (octave bands) of the first transfer frequency characteristic. Among the bands, one set having the largest difference calculated by the difference calculation unit 142 is extracted, and the extracted boundary frequency between adjacent frequency bands is defined as a boundary frequency.

  The processes of the characteristic deriving unit 140, the difference calculating unit 142, and the boundary deriving unit 144 will be described in detail with reference to FIGS.

  FIG. 9 is an explanatory diagram showing an example of the result of octave band analysis according to the present embodiment. As shown in FIG. 9, the characteristic deriving unit 140 causes the sound pressure detecting unit 118 to detect the sound pressure level of the measurement sound signal for each octave band (octave band analysis), and derives the first transmission frequency characteristic. And the difference calculation part 142 calculates the difference of the sound pressure of the adjacent frequency band in the several frequency band of a 1st transmission frequency characteristic, respectively.

  In the present embodiment, assuming that the reproduction ability of the target speaker that outputs the sound signal is low and the sound pressure in the low and high sound ranges is low, 31.5 Hz at both ends indicated by hatching in FIG. In addition, the octave band having a nominal center frequency of 16 kHz is not a target for which the difference calculating unit 142 calculates the difference in sound pressure.

  The boundary deriving unit 144 has five (differences b, c, d) in descending order of difference in sound pressure between adjacent octave bands (differences g, c, d, b, e, a, f in FIG. 9). , E, and g), adjacent octave bands B (125 Hz and 250 Hz), C (250 Hz and 500 Hz), D (500 Hz and 1 kHz), E (1 kHz and 2 kHz), and G (4 kHz and 8 kHz) are extracted. .

  FIG. 10 is an explanatory diagram showing an example of a result of 1/3 octave band analysis according to the present embodiment. As shown in FIG. 10, the characteristic deriving unit 140 sets the sound pressure level of the measured sound signal for each ３ octave band for the adjacent octave bands B, C, D, E, and G extracted in the description of FIG. The sound pressure detection unit 118 detects (1/3 octave band analysis). Here, the characteristic deriving unit 140 has already detected the sound pressure level (characteristic value) for each 1/3 octave band having a nominal center frequency of 100 Hz to 10 kHz, and overlaps with the extracted adjacent octave band. Even if there are three octave bands, the sound pressure level is detected only once.

  The difference calculation unit 142, among the 1/3 octave band groups (indicated by B, C, D, E, and G in FIG. 10) extracted by the characteristic deriving unit 140 and corresponding to adjacent octave bands, The difference between the frequency bands of adjacent 1/3 octave bands corresponding to the 3 octave bands is calculated.

  The boundary deriving unit 144 includes, for every five sets of adjacent frequency bands (octave bands) of the first transfer frequency characteristic, among the adjacent frequency bands in the plurality of frequency bands (1/3 octave band) of the second transfer frequency characteristic. Then, one set having the largest difference calculated by the difference calculating unit 142 is extracted, and the extracted boundary frequency between adjacent 1/3 octave bands is set as the boundary frequency.

  In FIG. 10, the nominal center frequency is a boundary frequency of 160 Hz and 200 Hz, 400 Hz and 500 Hz, 630 Hz and 800 Hz, 1.6 kHz and 2 kHz, and 1/3 octave band boundaries of 8 kHz and 10 kHz.

  As described above, the characteristic deriving unit 140 first derives the first transfer frequency characteristic divided for each octave band, narrows down adjacent frequency bands including frequencies that are candidates for the boundary frequency, and reduces the frequency band to the narrowed frequency band. Only the second transfer frequency characteristic divided for each 1/3 octave band is derived in detail only. With the configuration in which detection and derivation are performed in two stages, the boundary frequency can be appropriately determined by a simple process without locally biasing.

  Further, the boundary deriving unit 144 does not extract the adjacent first transmission frequency characteristic frequency band, that is, the adjacent octave band, in which the absolute value of the sound pressure difference is smaller than the predetermined threshold. With such a configuration, it is possible to avoid a situation where the absolute value of the difference in sound pressure is small and the frequency band that does not need to be adjusted is wasted, and the processing load can be reduced.

  As described above, in the present embodiment, the characteristic deriving unit 140 derives the first transfer frequency characteristic by performing 1 octave analysis, and derives the second transfer frequency characteristic by performing 1/3 octave analysis. However, the present invention is not limited to this, and any one of 1/2 octave analysis, 1/3 octave analysis, 1/4 octave analysis, 1/6 octave analysis, 1/12 octave analysis, 1/24 octave analysis is used. Thus, the second transfer frequency characteristic may be derived.

  With the configuration using such an octave band, octave band analysis, which is a fixed ratio analysis for obtaining sound pressure for each frequency band in which the ratio between the upper limit frequency and the lower limit frequency of the band is constant, is close to human hearing. Analysis becomes possible. Also, the second transfer frequency characteristic is analyzed and derived in a band divided by a divisor of 24, which is a multiple of the pitch included in one octave band, thereby obtaining a band in accordance with the 12 scales included in one octave. You can make adjustments. In particular, in view of the processing load cost and the effect of higher sound quality, it is appropriate to derive the second transfer frequency characteristic using a 1/3 octave band analysis using 3 of the divisors of 24. .

  Further, when more detailed detection and derivation is performed, for example, when the first transfer frequency characteristic is derived using a 1/3 octave band and the second transfer frequency characteristic is derived using a 1/12 octave band, boundary frequency candidates are derived. The boundary frequency can be derived more quickly by narrowing the frequency band including the frequency to be in two stages. Further, even when the number of compressors 128 is smaller than that of the present embodiment, the number of adjacent octave bands to be extracted is reduced, so that the number of processes until the boundary frequency is derived can be reduced, and the boundary frequency can be quickly increased. Can be derived.

  The derivation of the transfer frequency characteristic due to the pink noise described above is performed at predetermined time intervals in the present embodiment, but is not limited to such a case, and may be performed according to a user input, or when used in a vehicle. It may be performed when the vehicle speed exceeds or falls below a predetermined value, or may be performed when passing through a toll gate on the highway.

  The multiplication setting unit 146 sets a multiplication function for each frequency band divided by the boundary frequency based on the transfer frequency characteristic. Such a multiplication function includes a threshold value at which the multiplication rate changes as a parameter and an inclination indicating inclination, and a value that reduces the difference in sound pressure between adjacent octave bands, that is, the sound pressure level for each octave band is It is derived as a function calculated so as to approach the average value of the sound pressure level of the entire frequency band for which the transfer frequency characteristic is derived, and is used to adjust the input original sound signal. That is, the multiplication setting unit 146 has a sound pressure level lower than a frequency band having a lower sound pressure level in a frequency band divided by a boundary frequency of the input original sound signal (precisely, the original sound signal to which the harmonic signal is added). The multiplication function is set so that the frequency band with a higher level is attenuated more. In the present embodiment, the sound pressure adjustment unit performs processing such that the output level is lower than the input level, but may perform processing that amplifies the sound pressure level in a frequency band where the sound pressure level is small. good.

  In the following description of the present embodiment, the multiplication rate is a percentage of the ratio of the sound pressure (output level) of the original sound signal to be output to the sound pressure (input level) of the input original sound signal.

…（数式１）
ここで、Ｘ_１はコンプレッサ１２８の乗算率が変化する開始点（閾値）、ｍは乗算率の変化推移の傾斜を示す傾きである。 FIG. 6 is an explanatory diagram illustrating a multiplication function according to the present embodiment. In the present embodiment, the multiplication function shown in FIG. 6A used by the compressor 128 is expressed, for example, by Equation 1 below, where x is the input level and y is the output level.

... (Formula 1)
Wherein, X ₁ is the starting point multiplication rate of the compressor 128 is changed (threshold), m is the slope indicating the inclination of the multiplication factor of the change trend.

…（数式２） Further, when it is necessary to greatly attenuate the input level than FIG. 6 (a), the if m <1, multiplying setting unit 146, as shown in FIG. 6 (b), the threshold value X ₁ is smaller than the starting point setting the threshold _{X 2.} Then, the multiplication function is expressed by the following formula 2, for example.

... (Formula 2)

…（数式３） Similarly, when it is necessary to attenuate the input level more than that in FIG. 6A, the multiplication setting unit 146 sets a slope n smaller than the slope m as shown in FIG. 6C. Then, for example, the multiplication function is expressed by Equation 3 below.

... (Formula 3)

  As described above, the multiplication function includes the starting point (threshold value) at which the multiplication rate changes and the slope indicating the slope of the change transition as parameters. In addition, by providing an upper limit (lower limit) for each of the threshold at which the multiplication rate changes and the slope indicating the slope, it is possible to avoid a phenomenon in which the threshold or slope becomes an extreme value and the sound signal is distorted.

  The multiplication function is not limited to a function that reduces the difference in sound pressure between adjacent octave bands. For example, the multiplication setting unit 146 emphasizes a low frequency range or emphasizes a high frequency range. A multiplication function may be set as described above.

  With the configuration using the multiplication setting unit 146, the original sound signal is adjusted so that the frequency band in which sound is difficult to be transmitted due to the influence of the surrounding environment and the frequency band in which sound is easily transmitted can be heard at the same level. It is possible to suppress the deterioration of sound by adjusting the output sound pressure for each band.

  The boundary deriving unit 144 controls the LPF 122, the BPFs 124a to 124d, and the HPF 126 based on the boundary frequency. The original sound signal to which the harmonic signal is added is divided into signals for each frequency band delimited by the boundary frequency derived by the boundary deriving unit 144. Specifically, the LPF 122, the BPF 124a to 124d, and the HPF 126 each have a frequency component in a frequency band having a lowest boundary frequency as a lower limit frequency, and a frequency component in a frequency band excluding a frequency band having a maximum and minimum boundary frequency as a lower limit frequency. Only the frequency component of the frequency band having the boundary frequency having the highest frequency as the lower limit frequency is allowed to pass.

  More specifically, regarding the BPF 124, only the frequency component of the frequency band having the lower boundary frequency as the lower boundary frequency is passed through the BPF 124a, the BPF 124b third, the BPF 124c fourth, the BPF 124d fifth.

  The compressor 128 (compressors 128a to 128f) is based on the first transmission frequency characteristic, and in the present embodiment, the multiplication function set by the multiplication setting unit 146 according to the first transmission frequency characteristic (and the second transmission frequency characteristic). The sound pressure is adjusted for each of signals in a plurality of frequency bands divided by the LPF 122, the BPFs 124a to 124d, and the HPF 126. Specifically, the sound pressures of the signals in the plurality of frequency bands transmitted from the LPF 122, the BPF 124a to 124d, and the HPF 126 are adjusted in the compressors 128a to 128f based on the multiplication functions set for the compressors 128a to 128f, respectively. .

  The adder 130 adds the signals in the plurality of frequency bands transmitted from the compressors 128a to 128f and adjusted by the compressors 128a to 128f to generate one sound signal. The signal output unit 132 outputs one sound signal transmitted from the adder 130 to an audio device such as an external amplifier or a speaker (not shown). Thus, the original sound signal transmits the surrounding environment as sound.

  Further, in the present embodiment, the signal output unit 132 is a reference sound signal (frequency characteristic measurement signal) that is output in order to derive the transmission frequency characteristic of the sound of the surrounding environment in response to an instruction from the characteristic deriving unit 140. In this embodiment, a pink noise signal that is a sound signal of pink noise is output.

  As described above, the boundary deriving unit 144 and the compressor 128 derive the appropriate boundary frequency and signal based on the transfer frequency characteristic obtained by measuring the sound in the predetermined environment with the known frequency characteristic output from the sound source. Therefore, the sound signal processing apparatus 100 can be used effectively in various environments regardless of the environment to be used. In the present embodiment, the boundary frequency is dynamically determined according to the transmission frequency characteristic, not a fixed value, and therefore, a plurality of frequency bands delimited by the boundary frequency is the magnitude of the sound pressure level of the transmission frequency characteristic. Is a summary of the bands that are approximated. Therefore, sound pressure adjustment can be appropriately performed for each of a plurality of frequency bands divided by the boundary frequency, and output sound pressure for each frequency band can be adjusted to suppress sound deterioration.

(Sound signal processing method)
Next, a sound signal processing method using the sound signal processing apparatus 100 described above will be described. FIG. 11 is a flowchart showing specific processing of the sound signal processing method according to the present embodiment. The sound signal processing method is repeatedly executed at predetermined sampling periods while the original sound signal is being input to the sound signal processing apparatus 100.

  When the original sound signal is input, the characteristic deriving unit 140 determines whether or not a predetermined time has elapsed since the first transmission frequency characteristic was derived last time (S300). When the predetermined time has elapsed (YES in S300), the characteristic deriving unit 140 causes the signal output unit 132 to output a pink noise signal (S302).

  The characteristic deriving unit 140 performs octave band analysis by analyzing the sound pressure for each of a plurality of frequency bands of the measurement sound signal obtained by converting the pink noise collected by the microphone 116 (S304). Then, the difference calculation unit 142 calculates the difference between the sound pressures in adjacent frequency bands in the plurality of frequency bands (octave bands) of the first transfer frequency characteristic derived by the octave band analysis (S306). Subsequently, the boundary deriving unit 144 extracts a predetermined number (for example, 5) and adjacent frequency bands (octave bands) in descending order of the sound pressure difference (S308), and the characteristic deriving unit 140 selects each adjacent frequency. For the band, 1/3 octave band analysis is performed (S310).

  Subsequently, the difference calculation unit 142 calculates the difference in sound pressure between adjacent frequency bands in a plurality of frequency bands (1/3 octave bands) of the first transfer frequency characteristic derived by the 1/3 octave band analysis. (S312). The boundary deriving unit 144 has adjacent frequencies in a plurality of frequency bands (1/3 octave band) of the second transfer frequency characteristic for each one or a plurality of adjacent frequency bands (octave bands) of the first transfer frequency characteristic. Among the bands, one set having the largest difference calculated by the difference calculation unit 142 is extracted (S314), and a boundary frequency which is a frequency that becomes a boundary between adjacent 1/3 octave bands is derived (S316).

  Then, the multiplication setting unit 146 sets a multiplication function for each frequency band divided by the boundary frequency based on the transfer frequency characteristic (S318).

  Subsequently, the low-frequency overtone generation unit 112 outputs a harmonic signal having a predetermined frequency band, for example, a frequency component of 0 to 300 Hz, from the original sound signal input to the signal input unit 110, as the harmonic order increases. The pressure is generated to be small and added to the input original sound signal (S320).

  Then, the high-frequency overtone generation unit 114 has a predetermined order (for example, secondary) of frequency components greater than 8 kHz and less than or equal to 11 kHz among the original sound signal added with the overtone signal transmitted from the low-frequency range harmonic generation unit 112. A harmonic signal is generated, and the harmonic signal is added to the original sound signal added with the harmonic signal transmitted from the low-frequency range harmonic generation unit 112 (S322).

  The LPF 122, the BPF 124, and the HPF 126 are frequency bands in which the original sound signal added with the harmonic overtone signal transmitted from the high frequency overtone generation unit 114 according to the boundary frequency derived by the boundary deriving unit 144 is divided by the boundary frequency. Each signal is divided (S324).

  Subsequently, the compressor 128 adjusts the sound pressure for each signal divided by the LPF 122, the BPF 124, and the HPF 126 based on the multiplication function set by the multiplication setting unit 146 (S326), and the adder 130 performs adjustment processing. The respective signals thus added are added together to generate one sound signal (S328). The signal output unit 132 outputs the original sound signal transmitted from the compressor 128 to an audio device such as an external amplifier or a speaker (not shown) (S330).

  When the sound signal processing method of this embodiment is used, the boundary frequency is not a fixed value but dynamically determined according to the transfer frequency characteristic, and the divided frequency band is a band where the sound pressure level of the transfer frequency characteristic is close. Since they are summarized, adjustment processing can be appropriately performed for each divided frequency band, and output sound pressure for each frequency band can be adjusted to suppress sound deterioration.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to this embodiment. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

  Note that each step in the sound signal processing method of the present specification does not necessarily have to be processed in time series in the order described in the flowchart, and may include parallel or subroutine processing.

  The present invention can be used for a sound signal processing device and a sound signal processing method for adjusting the balance of frequency characteristics of sound signals.

DESCRIPTION OF SYMBOLS 100 ... Sound signal processing apparatus 112 ... Low-range overtone production | generation part 114 ... High-range overtone production | generation part 116 ... Microphone (sound collection part)
122 ... LPF (original sound division)
124 ... BPF (original sound division unit)
126 ... HPF (original sound division unit)
128 ... Compressor (Sound pressure adjustment unit)
130 ... adder (signal adder)
140 ... division derivation unit 142 ... difference calculation unit 144 ... boundary derivation unit 146 ... multiplication setting unit (sound pressure adjustment unit)
170 ... HPF (High Pass Filter)
184 ... HPF (High Pass Filter)

Claims (11)

A sound collection unit that collects sound of a known frequency characteristic output from a sound source at a predetermined listening point, and converts the sound into an electrical signal;
A characteristic deriving unit for deriving a first transfer frequency characteristic by analyzing sound pressure for each of a plurality of frequency bands of the electrical signal;
A difference calculating unit for calculating a difference in sound pressure between adjacent frequency bands in the plurality of frequency bands of the first transfer frequency characteristic;
One or more adjacent frequency bands in the first transmission frequency characteristic are extracted in descending order of the difference calculated by the difference calculation unit, and a boundary frequency that is a boundary frequency between the extracted adjacent frequency bands is derived. A boundary derivation unit;
An original sound dividing unit that divides the input original sound signal into signals for each frequency band divided by the boundary frequency;
A sound pressure adjusting unit that adjusts a sound pressure for each of signals in a plurality of frequency bands divided by the original sound dividing unit based on the first transmission frequency characteristic;
A signal adding unit that adds signals of a plurality of frequency bands adjusted by the sound pressure adjusting unit to generate one sound signal;
A sound signal processing apparatus comprising: The characteristic deriving unit includes a plurality of adjacent frequency bands of the first transmission frequency characteristic extracted by the boundary deriving unit, and a plurality of frequency bands of the first transmission frequency characteristic in the electrical signal. Derives the second transfer frequency characteristic by analyzing the sound pressure for each of a plurality of different frequency bands,
The difference calculation unit calculates a difference between sound pressures in adjacent frequency bands in a plurality of frequency bands of the second transfer frequency characteristic,
The boundary deriving unit is configured such that, for each of one or a plurality of adjacent frequency bands of the first transmission frequency characteristic, the difference calculation unit includes, among the adjacent frequency bands in the plurality of frequency bands of the second transmission frequency characteristic, the difference calculating unit 2. The sound signal processing apparatus according to claim 1, wherein one set having the largest calculated difference is extracted, and the extracted frequency at the boundary between adjacent frequency bands is set as a boundary frequency.   The characteristic deriving unit derives the first transmission frequency characteristic by performing one octave analysis, and performs 1/2 octave analysis, 1/3 octave analysis, 1/4 octave analysis, 1/6 octave analysis, 1/12 The sound signal processing apparatus according to claim 2, wherein the second transfer frequency characteristic is derived using either octave analysis or 1/24 octave analysis.   4. The sound signal according to claim 2, wherein the boundary deriving unit does not extract a frequency band of the adjacent first transmission frequency characteristics whose absolute value of the difference between the sound pressures is smaller than a predetermined threshold value. 5. Processing equipment.   5. The sound signal processing apparatus according to claim 1, wherein the sound pressure adjustment unit adjusts for each signal in a plurality of frequency bands so as to reduce the difference. 6.   Of the original sound signal, a harmonic overtone signal having a frequency component in a predetermined frequency band is generated such that the output sound pressure decreases as the order increases, and the generated overtone signal is added to the original sound signal. The sound signal processing apparatus according to claim 1, further comprising:   The sound signal processing apparatus according to claim 6, wherein the low-frequency range overtone generation unit includes a high-pass filter.   2. A high-frequency overtone generation unit that generates a harmonic signal of a predetermined order of a frequency component in a predetermined frequency band from the original sound signal, and adds the generated harmonic signal to the original sound signal. 8. The sound signal processing device according to any one of 1 to 7.   The sound signal processing apparatus according to claim 8, wherein the high-frequency range overtone generation unit includes a high-pass filter.   10. The sound signal processing according to claim 9, wherein when the original sound signal is subjected to compression processing, a cutoff frequency of the high-pass filter is a lower limit frequency of a frequency band discarded in the compression processing. apparatus. A sound having a known frequency characteristic output from a sound source is collected at a predetermined listening point, and the sound is converted into an electric signal.
Deriving the first transfer frequency characteristic by analyzing the sound pressure for each of a plurality of frequency bands of the electrical signal,
Calculating the difference between the sound pressures of adjacent frequency bands in the plurality of frequency bands of the first transmission frequency characteristic,
In order from the calculated difference, extract one or a plurality of adjacent frequency bands in the first transmission frequency characteristic, and derive a boundary frequency that is a boundary frequency of the extracted adjacent frequency bands;
Divide the input original sound signal into signals for each frequency band divided by the boundary frequency,
Based on the first transmission frequency characteristic, the sound pressure is adjusted for each of the signals of the plurality of divided frequency bands,
A sound signal processing method comprising: adding a plurality of adjusted frequency band signals to generate one sound signal.

Images